Engineer System
ZDLRA, Manual clone for Tape
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

ZDLRA, Manual clone for Tape

In my previous post, I showed how the clone to tape occurs for ZDLRA. But as explained, the clones occur through the scheduler and follow some rules. For full backup, as an example, it clones the last available.
But sometimes,  it is needed to call the clone for some specific backup, maybe to do long-term storage to follow some regimentation/law. And if we leverage this for the clone, jobs can maybe take to long, or clone more that you need.

 

Manual Clone

It is possible to clone one backup piece directly using CLI at ZDLRA. But we need to specify backup piece per backup piece since the parameter is the bp_key. But before clone, I recommend identifying some important things.
The first is the database key inside the rman catalog. A simple list db_unique_name show this:

 

RMAN> list db_unique_name of database ORCL18C;







List of Databases

DB Key  DB Name  DB ID            Database Role    Db_unique_name

------- ------- ----------------- ---------------  ------------------

15993   ORCL18C  558466555        PRIMARY          ORCL18C




RMAN>

 

The second is to identify the backupset and the backup piece that you need to clone. In this case, I picked up one from SYSTEM tablespace:

 

RMAN> list backupset 16512;







List of Backup Sets

===================







BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16512   Incr 0  330.35M    SBT_TAPE    00:03:03     02/02/2020 16:12:04

        BP Key: 16513   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB-INC0

        Handle: VB$_1891149551_16507I   Media:

  List of Datafiles in backup set 16512

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1909028    02/02/2020 16:09:01              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




RMAN>

 

Other are related inside ZDLRA database and are the protection policy linked with the database doing select over ra_database table:

 

SQL> select policy_name, db_key, db_unique_name from ra_database where db_unique_name = 'ORCL18C';




POLICY_NAME                                            DB_KEY DB_UNIQUE_NAME

-------------------------------------------------- ---------- --------------------------------

ZDLRA_ONLY_TAPE                                         15993 ORCL18C




SQL>

 

And to finish it needed to discover the template it is associated with the protection policy:

 

SQL> set linesize 250

SQL> col LIB_NAME format a30

SQL> col TEMPLATE_NAME format a50

SQL> col LAST_SCHEDULE_TIME format a35

SQL> col POLICY_NAME format a50

SQL> select TEMPLATE_NAME, BACKUP_TYPE, LAST_SCHEDULE_TIME, LIB_NAME, DB_UNIQUE_NAME, POLICY_NAME FROM RA_SBT_JOB WHERE LIB_NAME = 'OSBSBT' and POLICY_NAME = 'ZDLRA_ONLY_TAPE' ;




TEMPLATE_NAME                                      BACKUP_TYPE      LAST_SCHEDULE_TIME                  LIB_NAME                       DB_UNIQUE_NAME                 POLICY_NAME

-------------------------------------------------- ---------------- ----------------------------------- ------------------------------ ------------------------------ --------------------------------------------------

ONLYTAPE_FULL_ARCH                                 FULL, ARCH       02-FEB-20 04.40.37.077769 PM +01:00 OSBSBT                                                        ZDLRA_ONLY_TAPE

ONLYTAPE_ARCH                                      ARCH                                                 OSBSBT                                                        ZDLRA_ONLY_TAPE




SQL>

 

COPY_BACKUP_PIECE

The information collected before it is important because it is needed to use the DBMS_RA.COPY_BACKUP_PIECE procedure.
The parameter for the usage:
  • BP_KEY: Backup piece key that will be cloned.
  • FORMAT: Format to create the name for the backup. It follows the same format as RMAN (including the same wildcards).
  • TEMPLATE_NAME: The template that will be used to clone to tape. This is how ZDLRA discover the library and attributes for the clone.
So, to clone we just need to execute at ZDLRA:

 

SQL> BEGIN

  2      DBMS_RA.COPY_BACKUP_PIECE(

  3          BP_KEY => 16513,

  4          FORMAT => 'RA_SBT_ORCL18C_16512_16513_%U',

  5          TEMPLATE_NAME => 'ONLYTAPE_FULL_ARCH'

  6      );

  7  END;

  8  /




PL/SQL procedure successfully completed.




SQL>

 

Important there is the FORMAT. Look that that I used RA_SBT in the beginning to follow the same used by the normal clone to the tape of ZDLRA, after I added the DB_UNIQUE_NAME, with <BS_KEY>_<BP_KEY>, and %U to generate a unique name.
If we look internally, the backup was scheduled:

 

SQL> SELECT TASK_ID, STATE, DB_UNIQUE_NAME, ERROR_TEXT, BS_KEY, PIECE# FROM RA_SBT_TASK WHERE LIB_NAME = 'OSBSBT' AND  DB_UNIQUE_NAME = 'ORCL18C' and BS_KEY = 16512;




   TASK_ID STATE                DB_UNIQUE_NAME                 ERROR_TEXT                                             BS_KEY     PIECE#

---------- -------------------- ------------------------------ -------------------------------------------------- ---------- ----------

     33857 EXECUTABLE           ORCL18C                                                                                16512          1




SQL>

 

And OSB executed it:

 

[root@zdlras1n1 ~]# obtool lsjob

Job ID           Sched time  Contents                       State

---------------- ----------- ------------------------------ ---------------------------------------

oracle/42        none        database ORCL18C (dbid=558466555) processed; Oracle job(s) scheduled

oracle/42.1      none        datafile backup                running since 2020/02/02.20:48

[root@zdlras1n1 ~]#

[root@zdlras1n1 ~]# cat /usr/etc/ob/xcr/[email protected]

[4100000001]2020/02/02.20:48:35 ______________________________________________________________________

[4100000001]2020/02/02.20:48:35

[4100000002]2020/02/02.20:48:35         Transcript for job oracle/42.1 running on zdlras1n1

[4100000002]2020/02/02.20:48:35

[1100000003]2020/02/02.20:48:35 (amh)  qdv__automount_in_mh entered

[1100000004]2020/02/02.20:48:35 (amh)  qdv__automount_in_mh tape03 at 2020/02/02.20:48:35, flags 0x100

[1100000005]2020/02/02.20:48:35 (amh)  mount volume options list contains:

[1100000006]2020/02/02.20:48:35 (amh)     vtype 3 (app), vid (null), vs_create 0, family zdlras1-osbmf, retain (null), size 0, mediainfo 2, scratch 0

[1100000007]2020/02/02.20:48:36 (amh)  don't preserve previous mh automount state

[1100000008]2020/02/02.20:48:36 (gep)  getting reservation for element 0x3 (dte)

[1100000009]2020/02/02.20:48:36 (una)  unload_anywhere entered

[110000000A]2020/02/02.20:48:36 (amh) attempting automount with lm based volume selection

[110000000B]2020/02/02.20:48:36 (ulg)  1,(lm_amh)  beginning pass 1, mediainfo pass 1 at 2020/02/02.20:48:36

[110000000C]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    1 no vid

[110000000D]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    2 no vid

[110000000E]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    3 no vid

[110000000F]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    4 no vid

[1100000010]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  need next volume after oid 146 (vid zdlras1-osbmf-000001, tag E01005L4, create_time 1577658912, mediainfo 0) for append; its oid 155

[1100000011]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    5 oid 146 doesn't meet mount requirements - the next volume of this set is needed (OB device mgr)

[1100000012]2020/02/02.20:48:36 (ulg)  1,(lm_amh) oid 146 is added to state

[1100000013]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  need next volume after oid 148 (vid zdlras1-osbmf-000002, tag E01006L4, create_time 1577658918, mediainfo 4) for append; its oid 157

[1100000014]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    6 oid 148 doesn't meet mount requirements - the next volume of this set is needed (OB device mgr)

[1100000015]2020/02/02.20:48:36 (ulg)  1,(lm_amh) oid 148 is added to state

[1100000016]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  need next volume after oid 150 (vid zdlras1-osbmf-000003, tag E01007L4, create_time 1577658926, mediainfo 4) for append; its oid 153

[1100000017]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    7 oid 150 doesn't meet mount requirements - the next volume of this set is needed (OB device mgr)

[1100000018]2020/02/02.20:48:36 (ulg)  1,(lm_amh) oid 150 is added to state

[1100000019]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  family zdlras1-osbmf != RMAN-DEFAULT

[110000001A]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  oid 151 (vid RMAN-DEFAULT-000001, tag E01008L4, create_time 1577034094, mediainfo 0) fails mount criteria

[110000001B]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    8 oid 151 doesn't meet mount requirements - volume doesn't meet mount criteria (OB device mgr)

[110000001C]2020/02/02.20:48:36 (ulg)  1,(lm_amh) oid 151 is added to state

[110000001D]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  oid 153 (vid zdlras1-osbmf-000004, tag E01009L4, create_time 1577658926, mediainfo 4) passes criteria

[110000001E]2020/02/02.20:48:36 (ulg)  1,(lm_amh)    9 appendable volume found

[110000001F]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  oid 155 (vid zdlras1-osbmf-000005, tag E01010L4, create_time 1577658912, mediainfo 0) passes criteria

[1100000020]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   10 better appendable volume found

[1100000021]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   10 available space 159744,   9                                               available space 233472

[1100000022]2020/02/02.20:48:36 (ulg)  1,(lm_mmr)  oid 157 (vid zdlras1-osbmf-000006, tag E01011L4, create_time 1577658918, mediainfo 0) passes criteria

[1100000023]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   11 better appendable volume found

[1100000024]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   11 available space 18432,  10                                               available space 159744

[1100000025]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   12 no vid

[1100000026]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   13 no vid

[1100000027]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   14 no vid

[1100000028]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   15 no vid

[1100000029]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   16 no vid

[110000002A]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   17 no vid

[110000002B]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   18 no vid

[110000002C]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   19 no vid

[110000002D]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   20 no vid

[110000002E]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   21 vacant

[110000002F]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   22 no vid

[1100000030]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   23 no vid

[1100000031]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   24 vacant

[1100000032]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   25 vacant

[1100000033]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   26 vacant

[1100000034]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   27 vacant

[1100000035]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   28 vacant

[1100000036]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   29 vacant

[1100000037]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   30 no vid

[1100000038]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   31 no vid

[1100000039]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   32 no vid

[110000003A]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   33 no vid

[110000003B]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   34 no vid

[110000003C]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   35 no vid

[110000003D]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   36 no vid

[110000003E]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   37 no vid

[110000003F]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   38 no vid

[1100000040]2020/02/02.20:48:36 (ulg)  1,(lm_amh)   39 no vid

[1100000041]2020/02/02.20:48:36 (ulg)  1,(lm_amh)  end of pass 1 at 2020/02/02.20:48:362020/02/02.20:48:36

[1100000042]2020/02/02.20:48:36 (amh)  qlm__select_volume returned se state for vid = zdlras1-osbmf-000006 last_se = 0

[1100000043]2020/02/02.20:48:36 (amh)  state.pass = 1

[1100000044]2020/02/02.20:48:36 (amh)  state.last_se_checked = 11

[1100000045]2020/02/02.20:48:36 (amh)  state.mediainfo_pass = 1

[1100000046]2020/02/02.20:48:36 (amh)  state.mediainfo_loops = 1

[1100000047]2020/02/02.20:48:36 (amh)  state.rls_eltype = se

[1100000048]2020/02/02.20:48:36 (amh)  state.rls_elnum = 11

[1100000049]2020/02/02.20:48:36 (amh)   11 loading

[110000004A]2020/02/02.20:48:38 (atv)  qdv__automount_this_vol entered

[110000004B]2020/02/02.20:48:38 (atv)  calling qdv__mount

[110000004C]2020/02/02.20:48:38 (mt)   qdv__read_mount_db() succeeded, found vol_oid 0

[110000004D]2020/02/02.20:48:38 (mt)   qdv__read_label() succeeded; read 65536 bytes

[110000004E]2020/02/02.20:48:38 (mt)   exp time obtained from label

[110000004F]2020/02/02.20:48:38 (mt)   qdb__label_event() returned vol_oid 157

[1100000050]2020/02/02.20:48:38 (mt)   setting vol_oid in mount_info to 157

[1100000051]2020/02/02.20:48:38 (mt)   updated volume close time from db

[1100000052]2020/02/02.20:48:38 (atv)  qdv__mount succeeded

[1100000053]2020/02/02.20:48:38 (atv)  automount worked

[1100000054]2020/02/02.20:48:38 (atv)  qdv__automount_this_vol exited

[1100000055]2020/02/02.20:48:38 (gep)  getting reservation for element 0x3 (dte)

[1100000056]2020/02/02.20:48:38 (amh)   11 automount worked - returning

[1100000057]2020/02/02.20:48:38 (amh)  end of automount at 2020/02/02.20:48:38 (0x0)

[1100000058]2020/02/02.20:48:38 (amh)  returning from qdv__automount_in_mh

[3100000059]2020/02/02.20:48:38 Info: volume in tape03 is usable for this operation.

[110000005A]2020/02/02.20:48:38 (pvfw) at BOT

[110000005B]2020/02/02.20:48:38 (pvfw) previous state is invalid

[110000005C]2020/02/02.20:48:38 (alv)  backup image label is valid, file 19, section 2

[110000005D]2020/02/02.20:48:38 (pvfw) invalidating tape position in mount db

[110000005E]2020/02/02.20:48:38 (ial)  invalidate backup image label (was valid)

[110000005F]2020/02/02.20:48:38 (pvfw) space to EOD

[1100000060]2020/02/02.20:48:38 (pvfw) inspect_recs BSR: rtypes [0] = filemark

[1100000061]2020/02/02.20:48:38 (pvfw) inspect_recs BSR: rtypes [1] = filemark

[1100000062]2020/02/02.20:48:38 (pvfw) inspect_recs BSR: rtypes [2] = data

[1100000063]2020/02/02.20:48:38 (pvfw) inspect_recs BSR: rtypes [3] = filemark

[1100000064]2020/02/02.20:48:38 (pvfw) inspect_recs FSF

[1100000065]2020/02/02.20:48:38 (pvfw) inspect_recs ready to mount

[1100000066]2020/02/02.20:48:38 (pvfw) mounting at inspect_rec's request in rw_mode 2

[1100000067]2020/02/02.20:48:38 (pvfw) mounted ok

[1100000068]2020/02/02.20:48:38 (pvfw) at OB EOD, returning (2)

[1100000069]2020/02/02.20:48:38 (pvfw) pos_vol_cleanup not returning pstate

[010000006A]2020/02/02.20:48:38 (dmap) tape03 success

[410000006B]Volume label:

[410000006C]    Volume tag:             E01011L4

[410000006D]    Volume UUID:            a1f8db28-109b-1038-847b-080027b2a93d

[410000006E]    Volume ID:              zdlras1-osbmf-000006

[410000006F]    Volume set ID:          zdlras1-osbmf-000002

[4100000070]    Previous volume:        zdlras1-osbmf-000002

[4100000071]    Volume sequence:        2

[4100000072]    Volume set owner:       root

[4100000073]    Volume set created:     Sun Dec 29 23:35:18 2019

[4100000074]    Volume set closes:      Wed Oct 19 00:35:18 2022 (no writes after this time)

[4000000075]    Media family:           zdlras1-osbmf

[4100000076]

[4100000077]    Volume set expires: never; content manages reuse

[4100000078]

[4100000079]Archive label:

[410000007A]    File number:            24

[410000007B]    File section:           1

[410000007C]    Owner:                  root

[410000007D]    Client host:            zdlras1n1

[410000007E]    Backup level:           0

[410000007F]    S/w compression:        no

[4100000080]    Archive created:        Sun Feb 02 20:48:38 2020

[4100000081]    Archive owner:          oracle (UUID bbad0004-01a9-1038-abb2-080027b2a93d)

[4100000082]    Owner class:            oracle (UUID a43aa638-01a9-1038-8809-080027b2a93d)

[4100000083]    Backup piece name:      RA_SBT_ORCL18C_16512_16513_2tunhkot_1_2_0_qEdAn

[4100000084]    Backup db name:         ORCL18C

[4100000085]    Backup db id:           558466555

[4100000086]    Backup copy number:     not applicable

[4100000087]    Backup content:         full

[4100000088]    Encryption:             off

[4100000089]    Catalog data:           no

[410000008A]    Backup image UUID:      8cb63f94-281a-1038-86d5-080027b2a93d

[410000008B]    Backup instance UUID:   8cb63f9e-281a-1038-86d5-080027b2a93d

[410000008C]    Backup instance created:Sun Feb 02 20:48:38 2020

[410000008D]

[410000008E]

[410000008E]End of tape has been reached.  Please wait while I rewind and unload the tape.

[410000008F]The Volume ID of the next tape to be written is zdlras1-osbmf-000007.

[4100000090]The tape has been unloaded.

[4100000090]

[4100000091]Volume label:

[4100000092]    Volume tag:             E01012L4

[4100000093]    Volume UUID:            93cb6548-281a-1038-86d5-080027b2a93d

[4100000094]    Volume ID:              zdlras1-osbmf-000007

[4100000095]    Volume set ID:          zdlras1-osbmf-000002

[4100000096]    Previous volume:        zdlras1-osbmf-000006

[4100000097]    Volume sequence:        3

[4100000098]    Volume set owner:       root

[4100000099]    Volume set created:     Sun Dec 29 23:35:18 2019

[410000009A]    Volume set closes:      Wed Oct 19 00:35:18 2022 (no writes after this time)

[400000009B]    Media family:           zdlras1-osbmf

[410000009C]

[410000009D]    Volume set expires: never; content manages reuse

[410000009E]

[410000009F]Archive label:

[41000000A0]    File number:            24

[41000000A1]    File section:           2

[41000000A2]    Owner:                  root

[41000000A3]    Client host:            zdlras1n1

[41000000A4]    Backup level:           0

[41000000A5]    S/w compression:        no

[41000000A6]    Archive created:        Sun Feb 02 20:48:38 2020

[41000000A7]    Archive owner:          oracle (UUID bbad0004-01a9-1038-abb2-080027b2a93d)

[41000000A8]    Owner class:            oracle (UUID a43aa638-01a9-1038-8809-080027b2a93d)

[41000000A9]    Backup piece name:      RA_SBT_ORCL18C_16512_16513_2tunhkot_1_2_0_qEdAn

[41000000AA]    Backup db name:         ORCL18C

[41000000AB]    Backup db id:           558466555

[41000000AC]    Backup copy number:     not applicable

[41000000AD]    Backup content:         full

[41000000AE]    Encryption:             off

[41000000AF]    Catalog data:           no

[41000000B0]    Backup image UUID:      8cb63f94-281a-1038-86d5-080027b2a93d

[41000000B1]    Backup instance UUID:   8cb63f9e-281a-1038-86d5-080027b2a93d

[41000000B2]    Backup instance created:Sun Feb 02 20:48:38 2020

[41000000B3]

[41000000B4]

[41000000B4]Backup statistics:

[41000000B5]status 0

[40000000B6]devices

[40000000B7] tape03

[41000000B8]

[40000000B9]volumes

[40000000BA] zdlras1-osbmf-000006

[41000000BB]

[40000000BC]voltags

[40000000BD] E01011L4

[41000000BE]

[41000000BF]file 24

[41000000C0]host zdlras1n1

[41000000C1]encryption off

[41000000C2]compression no

[41000000C3]start_time  Sun Feb 02 2020 at 20:48:38 (1580672918)

[41000000C4]end_time    Sun Feb 02 2020 at 20:49:36 (1580672976)

[41000000C5]backup_time Sun Feb 02 2020 at 20:48:38 (1580672918)

[41000000C6]dev_kbytes 769938

[41000000C7]dev_iosecs 53

[41000000C8]dev_iorate 14.9 MB/S

[41000000C9]wrt_iosecs 58

[41000000CA]wrt_iorate 13.6 MB/S

[41000000CB]physical_blks_written 376

[41000000CC]write_errors 0

[41000000CD]physical_blks_read 0

[41000000CE]read_errors 0

[41000000CF]error_rate 0%

[41000000D0]piece RA_SBT_ORCL18C_16512_16513_2tunhkot_1_2_0_qEdAn completed, status 0

[root@zdlras1n1 ~]#

 

Look that above you can see that one tape became full and OSB used another to finish.

 

After the Clone

 

Returning to rman to check the same list we have:

 

RMAN> list backupset 16512;







List of Backup Sets

===================







BS Key  Type LV Size

------- ---- -- ----------

16512   Incr 0  330.35M

  List of Datafiles in backup set 16512

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1909028    02/02/2020 16:09:01              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




  Backup Set Copy #1 of backup set 16512

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    03:40:36     02/02/2020 16:12:04 YES        BKP-DB-INC0




    List of Backup Pieces for backup set 16512 Copy #1

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    16513   1   AVAILABLE                           VB$_1891149551_16507I




  Backup Set Copy #2 of backup set 16512

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    03:40:36     02/02/2020 19:49:37 NO         BKP-DB-INC0




    List of Backup Pieces for backup set 16512 Copy #2

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    16921   1   AVAILABLE   zdlras1-osbmf-000006,zdlras1-osbmf-000007 RA_SBT_ORCL18C_16512_16513_2tunhkot_1_2_0_qEdAn




RMAN>

 

As you can see, the backupset 16512 (of backup piece 16513) has two copies. One in ZDLRA database and another at OSB (look the media because of lack of space in one).

 

QUEUE_SBT_BACKUP_TASK

Another way to call manually the clone is to use the DBMS_RA.QUEUE_SBT_BACKUP_TASK. And in this case, everything that is linked with the template is cloned to tape. This a good way to force one clone one database entirely.
Look below that I have some archivedlogs that was not cloned yet:

 

RMAN> list copy of archivelog all;




List of Archived Log Copies for database with db_unique_name ORCL18C

=====================================================================




Key     Thrd Seq     S Low Time

------- ---- ------- - -------------------

16943   1    22      A 02/02/2020 16:28:28

        Name: /u01/app/oracle/oradata/ORCL18C/archivelog/2020_02_02/o1_mf_1_22_h3gbs5fk_.arc




16965   1    23      A 02/02/2020 21:03:49

        Name: /u01/app/oracle/oradata/ORCL18C/archivelog/2020_02_02/o1_mf_1_23_h3gcfwc8_.arc







RMAN> BACKUP DEVICE TYPE SBT ARCHIVELOG ALL DELETE ALL INPUT TAG 'BKP-ARCH';




Starting backup at 02/02/2020 21:15:09

current log archived

using channel ORA_SBT_TAPE_1

channel ORA_SBT_TAPE_1: starting archived log backup set

channel ORA_SBT_TAPE_1: specifying archived log(s) in backup set

input archived log thread=1 sequence=22 RECID=16 STAMP=1031346234

input archived log thread=1 sequence=23 RECID=17 STAMP=1031346892

input archived log thread=1 sequence=24 RECID=18 STAMP=1031346910

channel ORA_SBT_TAPE_1: starting piece 1 at 02/02/2020 21:15:11

channel ORA_SBT_TAPE_1: finished piece 1 at 02/02/2020 21:15:36

piece handle=ORCL18C_6nuni6mv_1_1 tag=BKP-ARCH comment=API Version 2.0,MMS Version 12.2.0.2

channel ORA_SBT_TAPE_1: backup set complete, elapsed time: 00:00:25

channel ORA_SBT_TAPE_1: deleting archived log(s)

archived log file name=/u01/app/oracle/oradata/ORCL18C/archivelog/2020_02_02/o1_mf_1_22_h3gbs5fk_.arc RECID=16 STAMP=1031346234

archived log file name=/u01/app/oracle/oradata/ORCL18C/archivelog/2020_02_02/o1_mf_1_23_h3gcfwc8_.arc RECID=17 STAMP=1031346892

archived log file name=/u01/app/oracle/oradata/ORCL18C/archivelog/2020_02_02/o1_mf_1_24_h3gcgg15_.arc RECID=18 STAMP=1031346910

Finished backup at 02/02/2020 21:15:36




Starting Control File and SPFILE Autobackup at 02/02/2020 21:15:37

piece handle=c-558466555-20200202-03 comment=API Version 2.0,MMS Version 12.2.0.2

Finished Control File and SPFILE Autobackup at 02/02/2020 21:15:54




RMAN> list backup of archivelog sequence between 22 and 24;







List of Backup Sets

===================







BS Key  Size       Device Type Elapsed Time Completion Time

------- ---------- ----------- ------------ -------------------

16991   25.25M     SBT_TAPE    00:00:20     02/02/2020 21:15:31

        BP Key: 16992   Status: AVAILABLE  Compressed: NO  Tag: BKP-ARCH

        Handle: ORCL18C_6nuni6mv_1_1   Media: Recovery Appliance (ZDLRAS1)




  List of Archived Logs in backup set 16991

  Thrd Seq     Low SCN    Low Time            Next SCN   Next Time

  ---- ------- ---------- ------------------- ---------- ---------

  1    22      1910746    02/02/2020 16:28:28 1923481    02/02/2020 21:03:49

  1    23      1923481    02/02/2020 21:03:49 1924479    02/02/2020 21:14:52

  1    24      1924479    02/02/2020 21:14:52 1924521    02/02/2020 21:15:09




RMAN>

 

And I called the template backup that I defined just for archivelogs (see my previous post) manually:

 

SQL> BEGIN

  2      DBMS_RA.QUEUE_SBT_BACKUP_TASK('ONLYTAPE_ARCH');

  3  END;

  4  /




PL/SQL procedure successfully completed.




SQL>

 

After that we can see the list backup again:

 

RMAN> list backup of archivelog sequence between 22 and 24;







List of Backup Sets

===================







BS Key  Size

------- ----------

16991   25.25M




  List of Archived Logs in backup set 16991

  Thrd Seq     Low SCN    Low Time            Next SCN   Next Time

  ---- ------- ---------- ------------------- ---------- ---------

  1    22      1910746    02/02/2020 16:28:28 1923481    02/02/2020 21:03:49

  1    23      1923481    02/02/2020 21:03:49 1924479    02/02/2020 21:14:52

  1    24      1924479    02/02/2020 21:14:52 1924521    02/02/2020 21:15:09




  Backup Set Copy #1 of backup set 16991

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:29:22     02/02/2020 21:15:35 NO         BKP-ARCH




    List of Backup Pieces for backup set 16991 Copy #1

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    16992   1   AVAILABLE   Recovery Appliance (ZDLRAS1) ORCL18C_6nuni6mv_1_1




  Backup Set Copy #2 of backup set 16991

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:29:22     02/02/2020 20:45:49 NO         BKP-ARCH




    List of Backup Pieces for backup set 16991 Copy #2

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    17065   1   AVAILABLE   zdlras1-osbmf-000005    RA_SBT_ORCL18C_558466555_16850_6nuni6mv_1_2_16991




RMAN>

 

And now we can see that we have a second copy for the backupset 16991.

 

Conclusion

Sometimes it is needed to manually clone the backups for tape. Remember that for full backups the ZDLRA only clones the last full backup. If you need, for some reason, clone other backups for long-term retention you can call manually the clone to tape. The options allow one to one clone, or the entire template.
 
 
Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”

ZDLRA, OSB and Clone to Tape
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

ZDLRA, OSB and Clone to Tape

As you saw in my last post, the configuration to enable clone to tape for ZDLRA it is not complicated, but you need to take care of some details to avoid errors. Besides that, ZDLRA relies on OSB to do that (when configured with native tape support) and this has some details that you need to be aware of.

 

In this post, I will show how the clone to tape works for ZDLRA. And how you can check some details about OSB.

 

Basic Config

Below you can see how I configured the clone to tape templates:

 

SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_JOB_TEMPLATE (

  3          template_name => 'ONLYTAPE_FULL_ARCH',

  4          protection_policy_name => 'ZDLRA_ONLY_TAPE',

  5          attribute_set_name => 'bronzerepfull',

  6          backup_type => 'FULL,ARCH',

  7          priority => DBMS_RA.SBT_PRIORITY_LOW,

  8          window => INTERVAL '4' HOUR

  9      );

 10  END;

 11  /




PL/SQL procedure successfully completed.




SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_JOB_TEMPLATE (

  3          template_name => 'ONLYTAPE_ARCH',

  4          protection_policy_name => 'ZDLRA_ONLY_TAPE',

  5          attribute_set_name => 'bronzerepfull',

  6          backup_type => 'ARCH',

  7          priority => DBMS_RA.SBT_PRIORITY_LOW,

  8          window => INTERVAL '4' HOUR

  9      );

 10  END;

 11  /




PL/SQL procedure successfully completed.




SQL>

 

Basic configuration for the protection policy ZDLRA_ONLY_TAPE and with FULL and ARCHIVELOGS clones. For more details you can check my previous post.
You can check the table RASYS.RA_SBT_JOB to check all the templates already configured:

 

SQL> set linesize 250

SQL> col LIB_NAME format a30

SQL> col TEMPLATE_NAME format a50

SQL> col LAST_SCHEDULE_TIME format a35

SQL> col POLICY_NAME format a50

SQL> select TEMPLATE_NAME, BACKUP_TYPE, LAST_SCHEDULE_TIME, LIB_NAME, DB_UNIQUE_NAME, POLICY_NAME FROM RA_SBT_JOB WHERE LIB_NAME = 'OSBSBT' ;




TEMPLATE_NAME                                      BACKUP_TYPE      LAST_SCHEDULE_TIME                  LIB_NAME                       DB_UNIQUE_NAME                 POLICY_NAME

-------------------------------------------------- ---------------- ----------------------------------- ------------------------------ ------------------------------ --------------------------------------------------

BRONZE_ARCH                                        ARCH             29-DEC-19 11.59.59.513917 PM +01:00 OSBSBT                                                        ZDLRA_BRONZE

BRONZE_FULL_ARCH                                   FULL, ARCH       03-JAN-20 11.10.27.411320 PM +01:00 OSBSBT                                                        ZDLRA_BRONZE

GOLDEN_ARCH                                        ARCH                                                 OSBSBT                                                        ZDLRA_GOLDEN

GOLDEN_FULL_ARCH                                   FULL, ARCH       01-JAN-20 11.10.47.987397 PM +01:00 OSBSBT                                                        ZDLRA_GOLDEN

ONLYTAPE_FULL_ARCH                                 FULL, ARCH                                           OSBSBT                                                        ZDLRA_ONLY_TAPE

ONLYTAPE_ARCH                                      ARCH                                                 OSBSBT                                                        ZDLRA_ONLY_TAPE




6 rows selected.




SQL>

 

Check that if the clone for this template was called you can see the information in the column LAST_SCHEDULE_TIME. And in this case, since I have not created the scheduler for this, it is empty.
To do that, just created the scheduler for that:

 

SQL> BEGIN

  2      DBMS_SCHEDULER.CREATE_JOB(

  3          job_name => 'SBTJOB_ONLYTAPE_FULL_ARCH',

  4          job_type => 'PLSQL_BLOCK',

  5          job_action => 'dbms_ra.queue_sbt_backup_task(''ONLYTAPE_FULL_ARCH'');',

  6          start_date => SYSDATE+(1/1440),

  7          enabled => TRUE,

  8          auto_drop => TRUE,

  9          repeat_interval => 'freq=WEEKLY; BYDAY=SUN; BYHOUR=20'

 10      );

 11  END;

 12  /




PL/SQL procedure successfully completed.




SQL> BEGIN

  2      DBMS_SCHEDULER.CREATE_JOB(

  3          job_name => 'SBTJOB_ONLYTAPE_ARCH',

  4          job_type => 'PLSQL_BLOCK',

  5          job_action => 'dbms_ra.queue_sbt_backup_task(''ONLYTAPE_ARCH'');',

  6          start_date => SYSDATE+(1/1440),

  7          enabled => TRUE,

  8          auto_drop => TRUE,

  9          repeat_interval => 'freq=HOURLY; BYMINUTE=40'

 10      );

 11  END;

 12  /




PL/SQL procedure successfully completed.




SQL>

 

Database Backup

For ZDLRA I have the database ORCL18C linked with this protection policy and I have some backups for this database. You can see below the backups for datafile 1:

 

RMAN> list backup of datafile 1;







List of Backup Sets

===================







BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16512   Incr 0  330.35M    SBT_TAPE    00:03:03     02/02/2020 16:12:04

        BP Key: 16513   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB-INC0

        Handle: VB$_1891149551_16507I   Media:

  List of Datafiles in backup set 16512

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1909028    02/02/2020 16:09:01              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16673   Incr 1  96.00K     SBT_TAPE    00:00:04     02/02/2020 16:27:01

        BP Key: 16674   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_1891149551_16669I   Media:

  List of Datafiles in backup set 16673

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    1  Incr 1910623    02/02/2020 16:26:57              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16701   Incr 0  329.22M    SBT_TAPE    00:00:04     02/02/2020 16:27:01

        BP Key: 16702   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_1891149551_16669_1   Media:

  List of Datafiles in backup set 16701

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1910623    02/02/2020 16:26:57              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




RMAN>

 

Check that I have the backupsets 16512 and 16701 that are the full backup. And the same for archivelogs.

 

During the Clone

When the scheduler trigger the clones, we can follow it inside of ZDLRA in several ways. But the point is that (as usual), for ZDLRA, everything is a task. So, we can see information in the tables RA_TASK and in RA_SBT_TASK.
You can create your sql’s to check what it is running, but look this example:

 

SQL> col state format a20

SQL> col ERROR_TEXT format a50

SQL> SELECT TASK_ID, STATE, DB_UNIQUE_NAME, ERROR_TEXT, BS_KEY, PIECE# FROM RA_SBT_TASK WHERE LIB_NAME = 'OSBSBT' AND  DB_UNIQUE_NAME = 'ORCL18C' ORDER BY DB_UNIQUE_NAME, BS_KEY, PIECE#;




   TASK_ID STATE                DB_UNIQUE_NAME                 ERROR_TEXT                                             BS_KEY     PIECE#

---------- -------------------- ------------------------------ -------------------------------------------------- ---------- ----------

     33698 EXECUTABLE           ORCL18C                                                                                16542          1

     33700 EXECUTABLE           ORCL18C                                                                                16565          1

     33702 EXECUTABLE           ORCL18C                                                                                16582          1

     33721 EXECUTABLE           ORCL18C                                                                                16618          1

     33704 EXECUTABLE           ORCL18C                                                                                16701          1

     33722 EXECUTABLE           ORCL18C                                                                                16706          1

     33706 EXECUTABLE           ORCL18C                                                                                16725          1

     33708 EXECUTABLE           ORCL18C                                                                                16734          1

     33710 EXECUTABLE           ORCL18C                                                                                16743          1

     33712 EXECUTABLE           ORCL18C                                                                                16752          1

     33714 EXECUTABLE           ORCL18C                                                                                16761          1




   TASK_ID STATE                DB_UNIQUE_NAME                 ERROR_TEXT                                             BS_KEY     PIECE#

---------- -------------------- ------------------------------ -------------------------------------------------- ---------- ----------

     33716 EXECUTABLE           ORCL18C                                                                                16782          1

     33718 EXECUTABLE           ORCL18C                                                                                16791          1

     33723 EXECUTABLE           ORCL18C                                                                                16795          1

     33720 EXECUTABLE           ORCL18C                                                                                16825          1




15 rows selected.




SQL>

 

The output above I picked up exactly when the job SBTJOB_ONLYTAPE_FULL_ARCH was running. You can see that ZDLRA put all the backupsets that was not cloned yet in the queue. At this moment, ZDLRA is creating the backupsets (since they exist jus in RA.VBDF table) and sending them (one to one) to OSB.
You can do the same for RA_TASK:

 

SQL> select TASK_ID, TASK_TYPE, STATE, WAITING_ON, DB_KEY, DB_UNIQUE_NAME, CREATION_TIME, ERROR_COUNT, INTERRUPT_COUNT, BP_KEY,BS_KEY,DF_KEY,VB_KEY from rasys.ra_task where DB_UNIQUE_NAME = 'ORCL18C' and task_type LIKE '%SBT' order by 5,2,7,10,11,12,13;




   TASK_ID TASK_TYPE                      STATE                     WAITING_ON     DB_KEY DB_UNIQUE_NAME                 CREATION_TIME                       ERROR_COUNT INTERRUPT_COUNT     BP_KEY     BS_KEY     DF_KEY     VB_KEY

---------- ------------------------------ ------------------------- ---------- ---------- ------------------------------ ----------------------------------- ----------- --------------- ---------- ---------- ---------- ----------

     33698 BACKUP_SBT                     RUNNING                                   15993 ORCL18C                        02-FEB-20 04.40.36.631549 PM +01:00           0               0                 16542

     33700 BACKUP_SBT                     RUNNING                                   15993 ORCL18C                        02-FEB-20 04.40.36.833354 PM +01:00           0               0                 16565

     33702 BACKUP_SBT                     RUNNING                                   15993 ORCL18C                        02-FEB-20 04.40.36.844599 PM +01:00           0               0                 16582

     33704 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.861902 PM +01:00           0               0                 16701

     33706 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.883175 PM +01:00           0               0                 16725

     33708 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.898718 PM +01:00           0               0                 16734

     33710 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.905336 PM +01:00           0               0                 16743

     33712 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.924090 PM +01:00           0               0                 16752

     33714 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.939241 PM +01:00           0               0                 16761

     33716 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.961104 PM +01:00           0               0                 16782

     33718 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.980517 PM +01:00           0               0                 16791




   TASK_ID TASK_TYPE                      STATE                     WAITING_ON     DB_KEY DB_UNIQUE_NAME                 CREATION_TIME                       ERROR_COUNT INTERRUPT_COUNT     BP_KEY     BS_KEY     DF_KEY     VB_KEY

---------- ------------------------------ ------------------------- ---------- ---------- ------------------------------ ----------------------------------- ----------- --------------- ---------- ---------- ---------- ----------

     33720 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.992346 PM +01:00           0               0                 16825

     33721 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.37.059282 PM +01:00           0               0                 16618

     33722 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.37.061375 PM +01:00           0               0                 16706

     33723 BACKUP_SBT                     EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.37.063679 PM +01:00           0               0                 16795

     33699 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.783661 PM +01:00           0               0                 16542      16014      16538

     33701 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.834255 PM +01:00           0               0                 16565      16012      16561

     33703 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.845368 PM +01:00           0               0                 16582      16018      16577

     33705 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.862820 PM +01:00           0               0                 16701      16006      16669

     33707 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.883803 PM +01:00           0               0                 16725      16008      16710

     33709 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.899324 PM +01:00           0               0                 16734      16022      16729

     33711 PLAN_SBT                       COMPLETED                                 15993 ORCL18C                        02-FEB-20 04.40.36.906037 PM +01:00           0               0                 16743      16010      16738




   TASK_ID TASK_TYPE                      STATE                     WAITING_ON     DB_KEY DB_UNIQUE_NAME                 CREATION_TIME                       ERROR_COUNT INTERRUPT_COUNT     BP_KEY     BS_KEY     DF_KEY     VB_KEY

---------- ------------------------------ ------------------------- ---------- ---------- ------------------------------ ----------------------------------- ----------- --------------- ---------- ---------- ---------- ----------

     33713 PLAN_SBT                       EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.924844 PM +01:00           0               0                 16752      16020      16747

     33715 PLAN_SBT                       EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.939997 PM +01:00           0               0                 16761      16024      16756

     33717 PLAN_SBT                       EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.961701 PM +01:00           0               0                 16782      16016      16777

     33719 PLAN_SBT                       EXECUTABLE                                15993 ORCL18C                        02-FEB-20 04.40.36.981277 PM +01:00           0               0                 16791      16026      16786




26 rows selected.




SQL>

 

Inside OSB

At OSB side we can check some information about how it is processing the clones. For OSB, everything is one job, and we can check in with obtool.

To check the jobs we execute obtool lsjob:

 

[root@zdlras1n1 ~]# obtool lsjob

Job ID           Sched time  Contents                       State

---------------- ----------- ------------------------------ ---------------------------------------

oracle/38        none        database ORCL18C (dbid=558466555) processed; Oracle job(s) scheduled

oracle/38.2      none        datafile backup                running since 2020/02/02.16:43

oracle/38.4      none        datafile backup                running since 2020/02/02.16:44

[root@zdlras1n1 ~]#

 

Some important information here. For every scheduler that was triggered by ZDLRA, we can see the job for each database, in this case, oracle/38. This is the master job, just to control what is happening. If you have more than one database in the same job being cloned, you will see more.
For each backupset that it is needed to be cloned, you will see child jobs (derived from the master). And for this, we can see logs from what is happening. All the logs for OSB are in the folder /usr/etc/ob/xcr/. Doing a cat from one finished job we can see useful information:

 

[root@zdlras1n1 ~]# cat /usr/etc/ob/xcr/[email protected]

[4100000001]2020/02/02.16:43:15 ______________________________________________________________________

[4100000001]2020/02/02.16:43:15

[4100000002]2020/02/02.16:43:15         Transcript for job oracle/38.2 running on zdlras1n1

[4100000002]2020/02/02.16:43:15

[3100000003]2020/02/02.16:43:15 Info: mount data verified.

[3100000004]2020/02/02.16:43:15 Info: volume in tape04 is usable for this operation.

[1100000005]2020/02/02.16:43:15 (pvfw) not at BOT

[1100000006]2020/02/02.16:43:15 (pvfw) previous state is invalid

[1100000007]2020/02/02.16:43:15 (alv)  backup image label is valid, file 21, section 1

[1100000008]2020/02/02.16:43:15 (pvfw) at OB EOD, returning (1)

[1100000009]2020/02/02.16:43:15 (pvfw) pos_vol_cleanup not returning pstate

[010000000A]2020/02/02.16:43:15 (dmap) tape04 success

[410000000B]Volume label:

[410000000C]    Volume tag:             E01011L4

[410000000D]    Volume UUID:            a1f8db28-109b-1038-847b-080027b2a93d

[410000000E]    Volume ID:              zdlras1-osbmf-000006

[410000000F]    Volume set ID:          zdlras1-osbmf-000002

[4100000010]    Previous volume:        zdlras1-osbmf-000002

[4100000011]    Volume sequence:        2

[4100000012]    Volume set owner:       root

[4100000013]    Volume set created:     Sun Dec 29 23:35:18 2019

[4100000014]    Volume set closes:      Wed Oct 19 00:35:18 2022 (no writes after this time)

[4000000015]    Media family:           zdlras1-osbmf

[4100000016]

[4100000017]    Volume set expires: never; content manages reuse

[4100000018]

[4100000019]Archive label:

[410000001A]    File number:            21

[410000001B]    File section:           1

[410000001C]    Owner:                  root

[410000001D]    Client host:            zdlras1n1

[410000001E]    Backup level:           0

[410000001F]    S/w compression:        no

[4100000020]    Archive created:        Sun Feb 02 16:43:15 2020

[4100000021]    Archive owner:          oracle (UUID bbad0004-01a9-1038-abb2-080027b2a93d)

[4100000022]    Owner class:            oracle (UUID a43aa638-01a9-1038-8809-080027b2a93d)

[4100000023]    Backup piece name:      RA_SBT_ORCL18C_558466555_16849_rpunhlqh_1_2_16701

[4100000024]    Backup db name:         ORCL18C

[4100000025]    Backup db id:           558466555

[4100000026]    Backup copy number:     not applicable

[4100000027]    Backup content:         full

[4100000028]    Encryption:             off

[4100000029]    Catalog data:           no

[410000002A]    Backup image UUID:      4599459c-27f8-1038-adc6-080027b2a93d

[410000002B]    Backup instance UUID:   459945b0-27f8-1038-adc6-080027b2a93d

[410000002C]    Backup instance created:Sun Feb 02 16:43:15 2020

[410000002D]

[410000002E]

[410000002E]Backup statistics:

[410000002F]status 0

[4000000030]devices

[4000000031] tape04

[4100000032]

[4000000033]volumes

[4000000034] zdlras1-osbmf-000006

[4100000035]

[4000000036]voltags

[4000000037] E01011L4

[4100000038]

[4100000039]file 21

[410000003A]host zdlras1n1

[410000003B]encryption off

[410000003C]compression no

[410000003D]start_time  Sun Feb 02 2020 at 16:43:15 (1580658195)

[410000003E]end_time    Sun Feb 02 2020 at 16:44:20 (1580658260)

[410000003F]backup_time Sun Feb 02 2020 at 16:43:15 (1580658195)

[4100000040]dev_kbytes 753036

[4100000041]dev_iosecs 60

[4100000042]dev_iorate 12.9 MB/S

[4100000043]wrt_iosecs 65

[4100000044]wrt_iorate 11.9 MB/S

[4100000045]physical_blks_written 368

[4100000046]write_errors 0

[4100000047]physical_blks_read 0

[4100000048]read_errors 0

[4100000049]error_rate 0%

[410000004A]piece RA_SBT_ORCL18C_558466555_16849_rpunhlqh_1_2_16701 completed, status 0

[root@zdlras1n1 ~]#

 

Some important details from there:
  • Volume tag: This is the tag for your tape. This can from the barcode read by OSB
  • Backup piece name: The name that was created and identify the file inside OSB
  • wrt_iorate: This is the speed that OSB wrote into the tape
  • write_errors: Errors that occurred while writing.
If that you can create some scripts to verify details from your OSBD clones:

 

[root@zdlras1n1 ~]# cat /usr/etc/ob/xcr/oracle@38.* |grep "Volume tag"

[410000000C]    Volume tag:             E01010L4

[410000000C]    Volume tag:             E01011L4

[410000000C]    Volume tag:             E01009L4

[410000000C]    Volume tag:             E01010L4

[410000000C]    Volume tag:             E01010L4

[410000000C]    Volume tag:             E01011L4

[410000000C]    Volume tag:             E01009L4

[root@zdlras1n1 ~]#

 

As told before (in my previous post), the integration between OSB and ZDLRA is not the best (to be honest). For ZDLRA OSB is one external tool. As an example, from ZDLRA library (cli or in cloud control) you can’t check OSB jobs and other details (like if the library has free tapes). If the clone fails, ZDLRA does not provide a good report and you need to go to CLI and use obtool. Look below that OSB is a separate box (image took from ZDLRA Administrator’s Guide).

Useful OSB commands

You can see more information about the devices using the obtool lsdev. Here I have 4 drivers:

 

[root@zdlras1n1 ~]# obtool lsdev

library             lib01            in service

  drive 1           tape01           in service

  drive 2           tape02           in service

  drive 3           tape03           in service

  drive 4           tape04           in service

[root@zdlras1n1 ~]#

 

And about the library (listing tapes) with obtool lsvol. To verify if you have free tapes inside of OSB library, as an example:

 

[root@zdlras1n1 ~]# obtool lsvol -L lib01

Inventory of library lib01:

    in    1:             unlabeled, barcode E01001L4, readonly

    in    2:             unlabeled, barcode E01002L4, readonly

    in    3:             unlabeled, barcode E01003L4, readonly

    in    4:             unlabeled, barcode E01004L4, readonly

    in    5:             volume zdlras1-osbmf-000001, barcode E01005L4, full, content manages reuse

    in    6:             volume zdlras1-osbmf-000002, barcode E01006L4, full, content manages reuse, mediainfo hw encryptable

    in    7:             volume zdlras1-osbmf-000003, barcode E01007L4, full, content manages reuse, mediainfo hw encryptable

    in    8:             volume RMAN-DEFAULT-000001, barcode E01008L4, content manages reuse

    in    12:            barcode E01012L4

    in    13:            barcode E01013L4

    in    14:            barcode E01014L4

    in    15:            barcode E01015L4

    in    16:            barcode E01016L4

    in    17:            barcode E01017L4

    in    18:            barcode E01018L4

    in    19:            barcode E01019L4

    in    20:            barcode E01020L4

    in    22:            barcode CLN101L4

    in    23:            barcode CLN102L5

    in    30:            barcode F01030L5

    in    31:            barcode F01031L5

    in    32:            barcode F01032L5

    in    33:            barcode F01033L5

    in    34:            barcode F01034L5

    in    35:            barcode F01035L5

    in    36:            barcode F01036L5

    in    37:            barcode F01037L5

    in    38:            barcode F01038L5

    in    39:            barcode F01039L5

    in    dte1:          volume zdlras1-osbmf-000004, barcode E01009L4, 233472 kb remaining, content manages reuse, lastse 9, mediainfo hw encryptable

    in    dte3:          volume zdlras1-osbmf-000005, barcode E01010L4, 159744 kb remaining, content manages reuse, lastse 10

    in    dte4:          volume zdlras1-osbmf-000006, barcode E01011L4, 18432 kb remaining, content manages reuse, lastse 11

[root@zdlras1n1 ~]#

 

And if you want to check for some volume obtool lsvol -b can be used. This is important to see everything that is inside one tape in specific:

 

[root@zdlras1n1 ~]# obtool lsvol -b E01011L4

    VOID    OOID Seq Volume ID          Barcode     Family      Created     Attributes

     157     157   2 zdlras1-osbmf-000006 E01011L4    zdlras1-osbmf 12/29.23:35 open; closes 2022/10/18; content manages reuse

[root@zdlras1n1 ~]#

 

And for the OSB volume that it inside of this tape we can use obtool lsbkup:

 

[root@zdlras1n1 ~]# obtool lsbkup -l -c zdlras1-osbmf-000006

Backup image name:    zdlras1n1-20200103-221130

    Type:                   Oracle database

    Client:                 zdlras1n1

    Backup piece name:      RA_SBT_ORCL19_13806_13805_nrul3953_1_2_0_MMNYx

    Database:               ORCL19

    Content:                full

    Size:                   768.2 MB

    Backup owner:           oracle

    Owner class:            oracle

    Backup date and time:   2020/01/03.23:11

    Created by job:         oracle/36.1

    UUID:                   89a815e8-109b-1038-847b-080027b2a93d

Backup image name:    zdlras1n1-20200202-154208

    Type:                   Oracle database

    Client:                 zdlras1n1

    Backup piece name:      RA_SBT_ORCL18C_558466555_16849_33unhlfr_1_2_16565

    Database:               ORCL18C

    Content:                full

    Size:                   240.8 MB

    Backup owner:           oracle

    Owner class:            oracle

    Backup date and time:   2020/02/02.16:42

    Created by job:         oracle/37.2

    UUID:                   1ef6b2ee-27f8-1038-995e-080027b2a93d

Backup image name:    zdlras1n1-20200202-154315

    Type:                   Oracle database

    Client:                 zdlras1n1

    Backup piece name:      RA_SBT_ORCL18C_558466555_16849_rpunhlqh_1_2_16701

    Database:               ORCL18C

    Content:                full

    Size:                   735.4 MB

    Backup owner:           oracle

    Owner class:            oracle

    Backup date and time:   2020/02/02.16:43

    Created by job:         oracle/38.2

    UUID:                   4599459c-27f8-1038-adc6-080027b2a93d

Backup image name:    zdlras1n1-20200202-154431

    Type:                   Oracle database

    Client:                 zdlras1n1

    Backup piece name:      RA_SBT_ORCL18C_558466555_16849_f2unhlqu_1_2_16782

    Database:               ORCL18C

    Content:                full

    Size:                   1.2 MB

    Backup owner:           oracle

    Owner class:            oracle

    Backup date and time:   2020/02/02.16:44

    Created by job:         oracle/38.6

    UUID:                   72577360-27f8-1038-8987-080027b2a93d

Backup image name:    zdlras1n1-20200202-154504

    Type:                   Oracle database

    Client:                 zdlras1n1

    Backup piece name:      RA_SBT_c-558466555-20200202-00

    Database:               ORCL18C

    Content:                full

    Size:                   18.1 MB

    Backup owner:           oracle

    Owner class:            oracle

    Backup date and time:   2020/02/02.16:45

    Created by job:         oracle/39.2

    UUID:                   85c228c8-27f8-1038-bcf9-080027b2a93d

[root@zdlras1n1 ~]#

 

As you can see we can use these commands to check what a lot of details about the volumes and tapes. In the command above you can see that inside of the same volume store backups from different databases. This occurs because of the share the same attribute inside ZDLRA, and there, I specified only one volume. 
If you want to see some details about one database you can use obtool lspiece:

 

[root@zdlras1n1 ~]# obtool lspiece -d ORCL18C

    POID Database   Content    Copy Created      Host             Piece name

     190 ORCL18C    full          0 02/02.16:42  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_34unhlje_1_2_16582

     191 ORCL18C    full          0 02/02.16:42  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_33unhlfr_1_2_16565

     192 ORCL18C    full          0 02/02.16:42  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_30unhl75_1_2_16542

     193 ORCL18C    full          0 02/02.16:43  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_cqunhlqk_1_2_16725

     194 ORCL18C    full          0 02/02.16:43  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_obunhlqn_1_2_16734

     195 ORCL18C    full          0 02/02.16:44  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_5dunhlqo_1_2_16743

     196 ORCL18C    full          0 02/02.16:43  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_rpunhlqh_1_2_16701

     197 ORCL18C    full          0 02/02.16:44  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_vuunhlqp_1_2_16752

     198 ORCL18C    full          0 02/02.16:44  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_f2unhlqu_1_2_16782

     199 ORCL18C    full          0 02/02.16:44  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_h1unhlqt_1_2_16761

     200 ORCL18C    full          0 02/02.16:44  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_9junhlqv_1_2_16791

     201 ORCL18C    full          0 02/02.16:45  zdlras1n1        RA_SBT_c-558466555-20200202-00

     202 ORCL18C    full          0 02/02.16:45  zdlras1n1        RA_SBT_c-558466555-20200202-02

     203 ORCL18C    full          0 02/02.16:45  zdlras1n1        RA_SBT_c-558466555-20200202-01

     204 ORCL18C    full          0 02/02.16:45  zdlras1n1        RA_SBT_ORCL18C_558466555_16849_6lunhltm_1_2_16795

[root@zdlras1n1 ~]#

 

For job list you can use the obtool lsjob -A:

 

[root@zdlras1n1 ~]# obtool lsjob -A

Job ID           Sched time  Contents                       State

---------------- ----------- ------------------------------ ---------------------------------------

oracle/36        none        database ORCL19 (dbid=323177095) completed successfully at 2020/01/03.23:12

oracle/36.1      none        datafile backup                completed successfully at 2020/01/03.23:12

oracle/37        none        database ORCL18C (dbid=558466555) completed successfully at 2020/02/02.16:42

oracle/37.1      none        datafile backup                completed successfully at 2020/02/02.16:42

oracle/37.2      none        datafile backup                completed successfully at 2020/02/02.16:42

oracle/37.3      none        datafile backup                completed successfully at 2020/02/02.16:42

oracle/38        none        database ORCL18C (dbid=558466555) completed successfully at 2020/02/02.16:44

oracle/38.1      none        datafile backup                completed successfully at 2020/02/02.16:43

oracle/38.2      none        datafile backup                completed successfully at 2020/02/02.16:44

oracle/38.3      none        datafile backup                completed successfully at 2020/02/02.16:43

oracle/38.4      none        datafile backup                completed successfully at 2020/02/02.16:44

oracle/38.5      none        datafile backup                completed successfully at 2020/02/02.16:44

oracle/38.6      none        datafile backup                completed successfully at 2020/02/02.16:44

oracle/38.7      none        datafile backup                completed successfully at 2020/02/02.16:44

oracle/39        none        database ORCL18C (dbid=558466555) completed successfully at 2020/02/02.16:45

oracle/39.1      none        datafile backup                completed successfully at 2020/02/02.16:45

oracle/39.2      none        datafile backup                completed successfully at 2020/02/02.16:45

oracle/39.3      none        datafile backup                completed successfully at 2020/02/02.16:45

oracle/40        none        database ORCL18C (dbid=558466555) completed successfully at 2020/02/02.16:45

oracle/40.1      none        datafile backup                completed successfully at 2020/02/02.16:45

oracle/41        none        database ORCL18C (dbid=558466555) completed successfully at 2020/02/02.16:45

oracle/41.1      none        datafile backup                completed successfully at 2020/02/02.16:45

[root@zdlras1n1 ~]#

 

As told, with these commands you can verify allot of useful info about the health of your clone to tapes at OSB side. And unfortunately, it is the only way to do that since it can be done using ZDLRA CLI.

 

After clone

After the clone job finish you can see that a new copy for the backupset appears in the rman catalog:

 

RMAN> list backup of datafile 1;







List of Backup Sets

===================







BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16512   Incr 0  330.35M    SBT_TAPE    00:03:03     02/02/2020 16:12:04

        BP Key: 16513   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB-INC0

        Handle: VB$_1891149551_16507I   Media:

  List of Datafiles in backup set 16512

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1909028    02/02/2020 16:09:01              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

16673   Incr 1  96.00K     SBT_TAPE    00:00:04     02/02/2020 16:27:01

        BP Key: 16674   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_1891149551_16669I   Media:

  List of Datafiles in backup set 16673

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    1  Incr 1910623    02/02/2020 16:26:57              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




BS Key  Type LV Size

------- ---- -- ----------

16701   Incr 0  329.22M

  List of Datafiles in backup set 16701

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 1910623    02/02/2020 16:26:57              NO    /u01/app/oracle/oradata/ORCL18C/system01.dbf




  Backup Set Copy #1 of backup set 16701

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:42:37     02/02/2020 16:27:01 YES        BKP-DB




    List of Backup Pieces for backup set 16701 Copy #1

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    16702   1   AVAILABLE                           VB$_1891149551_16669_1




  Backup Set Copy #2 of backup set 16701

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:42:37     02/02/2020 15:44:20 NO         BKP-DB




    List of Backup Pieces for backup set 16701 Copy #2

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    16876   1   AVAILABLE   zdlras1-osbmf-000006    RA_SBT_ORCL18C_558466555_16849_rpunhlqh_1_2_16701




RMAN>

 

Look that now the backupset 16701 has two copies. One inside ZDLRA database and others in the tape. And as you can see above, just the last full was cloned to tape. This is the default behavior for ZDLRA, just the last full is cloned to tape. I explained more in the previous post.
For archivelogs the process is a little different, but basically clones everything that was not yet. And if you have real-time redo you will have a lot of clones (since each one has a different backupset).
The management of the rman catalog is done automatically by ZDLRA. If your backup passes over the retention for recovery_window_goal (that manages just what is inside ZDLRA database), the copy will disappear from the list. The same occurs for cloned backups that follow the recovery_window_sbt parameter for your policy. The same is done when the clone is made, the new copy appears automatically.
 

 

 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”

 


ZDLRA, Configuring Tape Library
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

ZDLRA, Configuring Tape Library

With ZDLRA you can clone your backups to tape using two ways. The first is using third-party software and the second is using Oracle Secure Backup (OSB). This integration from ZDLRA and OSB is the native way to do that.
Cloning backups to tape will help to offload backups from ZDLRA (reducing the space usage if you need to sustain long recovery windows), and add another layer of protection (since you can put tapes in a third site).
Here I will show how easy is to configure the OSB backup and how to integrate it into your backup policy.

 

TAPE and ZDLRA

Tape backups are recommended for long-term storage, going to years and decades of storage. For ZDLRA itself, when using OSB, it turns itself as Media Manager too. This means that all the operations are realized by ZDLRA, tape load, rewind, cleans. Everything.
Using OSB as a tape library for ZDLRA, the communication with tape is done trough SAN, with direct access. This means that every database node of ZDLRA have SAN card, and you can connect to tape directly, or using SAN switch. All the cards have two links (4 as total) and you can configure as failover access (in each node).

 

 

All the tasks for tape are done with OSB (load, unload, overwrite), so, we can use commands like obtool lsdev, obtool lsvol and others to check the library and tapes.
Since OSB is external and exists outside the ZDLRA world there are some details that need be explained. As explained in previous posts (ZDLRA Internals, Virtual Full Backup), ZDLRA deconstructs the ingested backup and this database internal tables are not compatible with OSB. OSB not connect inside of the ZDLRA database and copy the backups.
So (copied directly from documentation), when Recovery Appliance executes a copy-to-tape job for a virtual full backup, it constructs the physical backup sets, and copies them to tape, and then writes the metadata to the recovery catalog. If desired, the Recovery Appliance can also copy successive incremental backups and archived redo log file backups to tape. Whereas the backup on the Recovery Appliance is virtual, the backup on tape is a non-virtual, full physical backup. The Recovery Appliance automatically handles requests to restore backups from tape, with no need for administrator intervention.
The info above is clear, but basically this means that ZDLRA constructs a new backupset and sends it to OSB to copy to tape. If you are sending a full backup to tape (and your backup has 14TB as an example) be aware of the time consumption to do that. Another point is that ZDLRA will automatically update the rman catalog for your database when the copy is finished at the tape, the backupset will appears inside the rman catalog.
One important point here it is there OSB manages totally the tape communication. If you receive one backupset to clone, you can’t go to ZDLRA and pause the copy for this backupset (you can pause the job that triggered the copy inside ZDLRA, but no the copy that already started). Consider that OSB as external software, where you send something to copy and the result will be OK or ERROR.
We can use some commands using obtool, but since it is integrated/managed by ZDLRA, it is not allowed to change the configuration. Commands like lsjob, lsdev, lsinventory, unlabelvol can be used.
Be careful when managing tapes because their communication is only from ZDLRA to OSB. If you clean one tape at OSB side, ZDLRA will not be aware of that and (if you have some not expired backupset there) you can lose some data.
Third-party Libraries
If you are not using OSB you can use external/third-party libraries to offload to tape. In this case, it needs to be compatible with RMAN. Basically, when you create the library inside ZDLRA, you point to the operational system client library (installed in each database node) and follow the normal procedure as OSB.
Be aware that you need to allow in your third-party software, that the client can send the backups to do the server. The start of the clone always will start at the ZDLRA side (and not called by third-party software).

 

CREATE_SBT_LIBRARY

In this post I will use OSB as a clone library, so, some details will be relevant just for OSB. But the process for creating the library will be similar (just adjust some parameters).
The first step is to create the library using the DBMS_RA.CREATE_SBT_LIBRARY:

 

SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_LIBRARY(

  3          lib_name => 'osbsbt',

  4          drives => 4,

  5          restore_drives => 1,

  6          parms => 'SBT_LIBRARY=libobk.so'

  7      );

  8  END;

  9  /




PL/SQL procedure successfully completed.




SQL>

 

The parameters are:
  • lib_name: Name that identify the library inside ZDRLA. It is unique name
  • drivers: Number of drivers that are available for backup usage
  • restore_drivers: Number of the driver used for restore. Will be always a subset of the drivers If not specified, ZDLRA can use all drivers to do the restore.
  • parms: here you can specify some parameters that are passed when to call the library.
Remember that the tape library needs to be compatible with RMAN? So, the parms and send attributes for the procedure are used as parameters for ALLOCATE CHANNEL when ZDLRA calls the tape library. If you need to define some specific parameters (like tapes groups, client names, or whatever), you need to define it here. This follows the same format as PARMS in rman ALLOCATE CHANNEL.
After that, you can create attributes for your tape using the CREATE_SBT_ATTRIBUTE_SET. These attributes are important if you want to segregate some backups. As an example, you can create one attribute to send some backups for specific media families. So, if you have tapes to store backups after some application (or company milestones) you can create specific attributes.

 

SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_ATTRIBUTE_SET(

  3          lib_name => 'osbsbt',

  4          attribute_set_name => 'bronzerepfull',

  5          streams => 4,

  6          parms => 'ENV=(OB_MEDIA_FAMILY=zdlras1-osbmf)'

  7      );

  8  END;

  9  /




PL/SQL procedure successfully completed.




SQL>

 

Here, I defined that the attribute bronzerepfull for my library osbsbt. That means (for OSB) that will use the media family zdlras1-osbmf to store the backups. Again, the parms and send use the format for rman PARMS at ALLOCATE CHANNEL in rman.
One detail important it is that streams attribute needs to be less or equal than drivers value that you specified during CREATE_SBT_LIBRARY commands.
If you check the RASYS.RA_SBT_LIBRARY table the library is there:

 

SQL> col LIB_NAME format a30

SQL> col LAST_ERROR_TEXT format a30

SQL> SELECT LIB_NAME, LAST_ERROR_TEXT, STATUS FROM RA_SBT_LIBRARY;




LIB_NAME                       LAST_ERROR_TEXT                STATU

------------------------------ ------------------------------ -----

OSBSBT                                                        READY

REP$LIB_9043_ZDLRAS2_REP                                      READY




SQL>

 

If you are starting the ZDLRA and it already will come with OSB integration the steps above are not needed. Will be configured directly during the deployment phase of ZDLRA.

 

Clone to Tape Jobs

The clone to tape jobs inside of ZDLRA is managed entirely with DBMS_SCHEDULER and for protection policies or db_unique_name as a parameter. This means that unfortunately, you need to take care of the overlaps of the clone to tape jobs
If, as an example, you defined the clone for BRONZE policies to start from 10:00 until 11:00 and the SILVER to start at 11:00. And your backup for BRONZE took more time to execute and pass over 11:00, you will see some contentions.
 

DBMS_RA.CREATE_SBT_JOB_TEMPLATE

 

The first step to clone to tape is to define what and how to do that. For ZDLRA we can clone full, incremental, and archivelogs backups. Besides that, we can define the templates for protection policies or database unique names.

 

SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_JOB_TEMPLATE (

  3          template_name => 'BRONZE_FULL_ARCH',

  4          protection_policy_name => 'ZDLRA_BRONZE',

  5          attribute_set_name => 'bronzerepfull',

  6          backup_type => 'FULL,ARCH',

  7          priority => DBMS_RA.SBT_PRIORITY_LOW,

  8          window => INTERVAL '4' HOUR

  9      );

 10  END;

 11  /




PL/SQL procedure successfully completed.




SQL> BEGIN

  2      DBMS_RA.CREATE_SBT_JOB_TEMPLATE (

  3          template_name => 'BRONZE_ARCH',

  4          protection_policy_name => 'ZDLRA_BRONZE',

  5          attribute_set_name => 'bronzerepfull',

  6          backup_type => 'ARCH',

  7          priority => DBMS_RA.SBT_PRIORITY_LOW,

  8          window => INTERVAL '4' HOUR

  9      );

 10  END;

 11  /




PL/SQL procedure successfully completed.




SQL>

 

Basically, with CREATE_SBT_JOB_TEMPLATE, we define what we want to backup (database or policy) or which type of backup we want (full, archive).
Another important definition it is the priority between the templates. So, in case of your backup took more time than expected, the priority helps ZDLRA to automatically handle the overlaps of clones times. You can use the window to specify the maximum time that clone can runs, after that the backupset will put in the queue to be copy in the next run.
In the example above you can see that:
  • BRONZE_FULL_ARCH: Template that defines that FULL and ARCHIVELOGS backups will be copied at the same time to tape. We will use the SBT attribute bronzerepfull that we defined previously. And have 4 hours to maximum runs and have LOW priority.
  • BRONZE_ARCH: Template that defines that just ARCHIVELOGS backups will be copied at the same time to tape. We will use the SBT attribute bronzerepfull that we defined previously. And have 4 hours to maximum runs and have LOW priority.
Important to remember that ZDLRA always copies the last backupset available. This means that for a full backup, it will be always the last that was not copied since the last execution of your job. For archivelogs, it will be from the last clone to tape (even if the last copied just the archivelogs). One example is that you execute every day one copy per day for your database at 22:00, but you do two incremental backups (one 11:00 and other 19:00), just the full backup generated because the incremental backup at 19:00 will be cloned to tape.
For archivelogs, the rule is copy all since the last clone to tape. This guarantees that you always have the archivelogs needed to recover the database. One point that needs attention is when using real-time redo, in this case, each backup of archivelog is unique (different than we execute filesperset 1000 – which will have one backupset with 1000 archivelogs inside). This means that if you generate 2000 archivelogs per day, you will need to clone these files one to one, and for tape (depending the tape type), can took long time because will be (inside tape) 2000 different files (think about the process for tape, where each file need a mark where start and end – this took a lot of time consumption for tape process).
It is possible to create a template with TAG’s too. In this case, just backups with specific TAGS (from rman) will be cloned to tape. This is useful when you need to clone specific backups for long retention schemas.

 

DBMS_SCHEDULER.CREATE_JOB

 

After creating the templates it is possible to schedule it using the database scheduler. Is used the normal scheduler, setting the desired template and the desired time that will be called.
As an example:

 

SQL> BEGIN

  2      DBMS_SCHEDULER.CREATE_JOB(

  3          job_name => 'SBTJOB_BRONZE_FULL_ARCH',

  4          job_type => 'PLSQL_BLOCK',

  5          job_action => 'dbms_ra.queue_sbt_backup_task(''BRONZE_FULL_ARCH'');',

  6          start_date => SYSDATE+(1/1440),

  7          enabled => TRUE,

  8          auto_drop => TRUE,

  9          repeat_interval => 'freq=WEEKLY; BYDAY=SUN; BYHOUR=20'

 10      );

 11  END;

 12  /




PL/SQL procedure successfully completed.




SQL>

 

We can see that it is a normal schedule that calls one PLSQL_BLOCK, and the block it is dbms_ra.queue_sbt_backup_task that will inform the internal ZDLRA library to generate the backupset and send to OSB. As you can see, the parameter for the function is the template name created before.
The start_date and repeat_internal are defined as is for the normal scheduler. In this particular example, the backup will be every week, Sunday at 20:00.
And translating everything this means that I will have, in tape, one full backup (plus archivelog) for each database that it is part of BRONZE protection policy. And this will occurs every Sunday at 20:00. The library will be OSB and the media family to store the backups will be zdlras1-osbmf.
Another example:

 

SQL> BEGIN

  2      DBMS_SCHEDULER.CREATE_JOB(

  3          job_name => 'SBTJOB_BRONZE_ARCH',

  4          job_type => 'PLSQL_BLOCK',

  5          job_action => 'dbms_ra.queue_sbt_backup_task(''BRONZE_ARCH'');',

  6          start_date => SYSDATE+(1/1440),

  7          enabled => TRUE,

  8          auto_drop => TRUE,

  9          repeat_interval => 'freq=HOURLY; BYMINUTE=40'

 10      );

 11  END;

 12  /




PL/SQL procedure successfully completed.




SQL>

 

And some examples of repeat intervals:
  • repeat_interval => ‘freq=HOURLY; BYMINUTE=0’: Every hour as *:00
  • repeat_interval => ‘FREQ=DAILY; INTERNAL=1; BYHOUR=20; BYMINUTE=15’: Every day at 20:15
  • repeat_interval => ‘FREQ=DAILY; BYDAY=MON,TUE,WED,THU,FRI; BYHOUR=7,8,9,10,11,12,13,14,15,16,17,18; BYMINUTE=05’: Monday to Friday, from 08-18 at 05 minute.
  • repeat_interval => ‘freq=DAILY; BYHOUR=22; BYDAY=MON,TUE,WED,THU,FRI,SAT,SUN’: Another way to specify every day at 22:00
  • repeat_interval => ‘freq=WEEKLY; BYDAY=SAT; BYHOUR=23’: Every week at 20:00

Cloned Backups

 

After configure the scheduler, and it is called we can see this in rman:

 

RMAN> list backup of archivelog sequence 92;







List of Backup Sets

===================







BS Key  Size

------- ----------

10082   2.80M




  List of Archived Logs in backup set 10082

  Thrd Seq     Low SCN    Low Time            Next SCN   Next Time

  ---- ------- ---------- ------------------- ---------- ---------

  1    92      2520099    01/01/2020 22:31:51 2530992    01/01/2020 23:29:08




  Backup Set Copy #1 of backup set 10082

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:43:14     01/01/2020 23:29:24 YES        TAG20200101T232923




    List of Backup Pieces for backup set 10082 Copy #1

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    10083   1   AVAILABLE                           $RSCN_1920977_RTIM_1028557385_THRD_1_SEQ_92_CTKEY_10033_BACKUP




  Backup Set Copy #2 of backup set 10082

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    00:43:14     01/01/2020 22:46:09 YES        TAG20200101T232923




    List of Backup Pieces for backup set 10082 Copy #2

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    10338   1   AVAILABLE   zdlras2-osbmf-000003    RA_SBT_OR19DG_41954437_10086_66uku2ij_1_2_10082




RMAN>

 

And for datafile

 

RMAN> list backup of datafile 1;







List of Backup Sets

===================







BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

9067    Incr 0  337.52M    SBT_TAPE    00:00:57     01/01/2020 19:07:12

        BP Key: 9068   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB-INC0

        Handle: VB$_2127575003_9062I   Media:

  List of Datafiles in backup set 9067

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 2438840    01/01/2020 19:03:19 2440207      NO    +DATA/OR19DGS/DATAFILE/system.265.1028562437




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

9210    Incr 1  136.00K    SBT_TAPE    00:00:10     01/01/2020 19:16:43

        BP Key: 9211   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_2127575003_9206I   Media:

  List of Datafiles in backup set 9210

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    1  Incr 2441418    01/01/2020 19:15:20              NO    +DATA/OR19DGS/DATAFILE/system.265.1028562437




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

9218    Incr 0  336.36M    SBT_TAPE    00:00:10     01/01/2020 19:16:43

        BP Key: 9219   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_2127575003_9206_1   Media:

  List of Datafiles in backup set 9218

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 2441418    01/01/2020 19:15:20              NO    +DATA/OR19DGS/DATAFILE/system.265.1028562437




BS Key  Type LV Size       Device Type Elapsed Time Completion Time

------- ---- -- ---------- ----------- ------------ -------------------

9496    Incr 1  168.00K    SBT_TAPE    00:00:10     01/01/2020 20:26:41

        BP Key: 9497   Status: AVAILABLE  Compressed: YES  Tag: BKP-DB

        Handle: VB$_2127575003_9492I   Media:

  List of Datafiles in backup set 9496

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    1  Incr 2452089    01/01/2020 20:24:22              NO    +DATA/OR19DGS/DATAFILE/system.265.1028562437




BS Key  Type LV Size

------- ---- -- ----------

9500    Incr 0  336.36M

  List of Datafiles in backup set 9500

  File LV Type Ckp SCN    Ckp Time            Abs Fuz SCN Sparse Name

  ---- -- ---- ---------- ------------------- ----------- ------ ----

  1    0  Incr 2452089    01/01/2020 20:24:22              NO    +DATA/OR19DGS/DATAFILE/system.265.1028562437




  Backup Set Copy #1 of backup set 9500

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    02:14:00     01/01/2020 20:26:41 YES        BKP-DB




    List of Backup Pieces for backup set 9500 Copy #1

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    9501    1   AVAILABLE                           VB$_2127575003_9492_1




  Backup Set Copy #2 of backup set 9500

  Device Type Elapsed Time Completion Time     Compressed Tag

  ----------- ------------ ------------------- ---------- ---

  SBT_TAPE    02:14:00     01/01/2020 22:40:31 NO         BKP-DB




    List of Backup Pieces for backup set 9500 Copy #2

    BP Key  Pc# Status      Media                   Piece Name

    ------- --- ----------- ----------------------- ----------

    10125   1   AVAILABLE   zdlras2-osbmf-000002,zdlras2-osbmf-000005 RA_SBT_OR19DG_41954437_10086_qjuktnrn_1_2_9500




RMAN>

 

For both cases look that we have the Backup Set Copy #2 in tape. And specifically for datafile, look that just the last full backup was cloned to tape.
These are just examples, in other posts, I will provide more examples about clone to tape.

 

Conclusion

Configurate ZDLRA to do clone to tapes and offload the backup is not complicated. But will require some attention because of the details. The most important it is related with the job scheduler. It is DBA duties correctly schedule it and avoid some overlap between the jobs.
Unfortunately, there is no easy way to do that. And can be more complicate because you can add more database to your protection policies and this will increate the clone time.
Beside that we have the fact that OSB (or third-part library) are handle as “external” by ZDLRA. This means that are not much control beside start and stop coming from ZDLRA. No easy error identification, report and solution besides “try again”.
Clone to tape backup is one of activity that demand more attention for ZDLRA, and need to take care a lot of details to have a good clone to tape plan that will lead less errors in the future.
 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”

 


Solving MGMTDB errors during 18c GI RU apply
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

Solving MGMTDB errors during 18c GI RU apply

Recently I executed the upgrade of Oracle GI to 19c version, from 18.6.0.0 to 19.5.0.0 version. But one step that was not showed there was that, because of requirements, the GI was upgraded from 18.2.0.0 to 18.6.0.0. This upgrade is a just Release Update (RU) apply and opatchauto command.
But during this upgrade, from 18.2 to 18.6, I faced (more than one time – 5 to be precise) errors during the update because of the MGMTDB errors. I got these errors:
  • ORA-12514, TNS: Listener does not currently know of service requested in connect descriptor
  • ORA-01017: invalid username/password; logon denied
  • MGTCA-1005 : Could not connect to the GIMR.
  • CRS-10407: (:CLSCRED1079:)Credential domain does not exist.
Here I will show how to solve these errors, how to identify if everything was fine and if you can continue. Be careful that it is an example, always open a support SR to identify the source of the error.

 

RU and Patch Process

To apply 18c RU over GI it is simple, basically, it is needed to call opatchauto for every node. One example of correct execution is:

 

[root@exa01vm01 ~]# opatchauto apply /u01/patches/grid/29301682 -oh /u01/app/18.0.0/grid




OPatchauto session is initiated at Mon Jan 20 12:48:25 2020




System initialization log file is /u01/app/18.0.0/grid/cfgtoollogs/opatchautodb/systemconfig2020-01-20_12-48-48PM.log.




Session log file is /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/opatchauto2020-01-20_12-49-44PM.log

The id for this session is RQ3F




Executing OPatch prereq operations to verify patch applicability on home /u01/app/18.0.0/grid

Patch applicability verified successfully on home /u01/app/18.0.0/grid







Bringing down CRS service on home /u01/app/18.0.0/grid

CRS service brought down successfully on home /u01/app/18.0.0/grid







Start applying binary patch on home /u01/app/18.0.0/grid

Binary patch applied successfully on home /u01/app/18.0.0/grid







Starting CRS service on home /u01/app/18.0.0/grid

CRS service started successfully on home /u01/app/18.0.0/grid




OPatchAuto successful.




--------------------------------Summary--------------------------------




Patching is completed successfully. Please find the summary as follows:




Host:exa01vm01

CRS Home:/u01/app/18.0.0/grid

Version:18.0.0.0.0

Summary:




==Following patches were SUCCESSFULLY applied:




Patch: /u01/patches/grid/29301682/28435192

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-20_12-51-42PM_1.log




Patch: /u01/patches/grid/29301682/28547619

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-20_12-51-42PM_1.log




Patch: /u01/patches/grid/29301682/29301631

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-20_12-51-42PM_1.log




Patch: /u01/patches/grid/29301682/29301643

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-20_12-51-42PM_1.log




Patch: /u01/patches/grid/29301682/29302264

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-20_12-51-42PM_1.log










OPatchauto session completed at Mon Jan 20 13:00:01 2020

Time taken to complete the session 11 minutes, 36 seconds

[root@exa01vm01 ~]#










####################################################

#Execute in node 2

####################################################







[root@exa02vm01 ~]# opatchauto apply /u01/patches/grid/29301682 -oh /u01/app/18.0.0/grid




OPatchauto session is initiated at Tue Jan 21 13:12:37 2020




System initialization log file is /u01/app/18.0.0/grid/cfgtoollogs/opatchautodb/systemconfig2020-01-21_01-12-51PM.log.




Session log file is /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/opatchauto2020-01-21_01-13-37PM.log

The id for this session is NFXL




Executing OPatch prereq operations to verify patch applicability on home /u01/app/18.0.0/grid

Patch applicability verified successfully on home /u01/app/18.0.0/grid







Bringing down CRS service on home /u01/app/18.0.0/grid

CRS service brought down successfully on home /u01/app/18.0.0/grid







Start applying binary patch on home /u01/app/18.0.0/grid

Binary patch applied successfully on home /u01/app/18.0.0/grid







Starting CRS service on home /u01/app/18.0.0/grid

CRS service started successfully on home /u01/app/18.0.0/grid




OPatchAuto successful.




--------------------------------Summary--------------------------------




Patching is completed successfully. Please find the summary as follows:




Host:exa02vm01

CRS Home:/u01/app/18.0.0/grid

Version:18.0.0.0.0

Summary:




==Following patches were SUCCESSFULLY applied:




Patch: /u01/patches/grid/29301682/28435192

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-21_13-15-58PM_1.log




Patch: /u01/patches/grid/29301682/28547619

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-21_13-15-58PM_1.log




Patch: /u01/patches/grid/29301682/29301631

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-21_13-15-58PM_1.log




Patch: /u01/patches/grid/29301682/29301643

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-21_13-15-58PM_1.log




Patch: /u01/patches/grid/29301682/29302264

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-21_13-15-58PM_1.log










OPatchauto session completed at Tue Jan 21 13:35:21 2020

Time taken to complete the session 22 minutes, 44 seconds

[root@exa02vm01 ~]#

 

As you can see, call in one node first, and after the second node. Basically during this process, the opatch will apply the binary updates in GI home, restart GI (to open with the new version), and apply needed SQL patches over MGMTDB and ASM.

 

MGMTDB and errors

But some errors can occur during the apply, mainly for MGMTDB. These errors are usually related to the communication between MGMTDB and CRS resources (listener and others). Usually, these errors occur when applying the patch in the last node of the cluster because the MGMTDB is swapped from other nodes (until the last one) during the patch application.
Some errors examples that I faced:
ORA-12514, TNS:listener does not currently know of service requested in connect descriptor

 

...

CRS-4123: Oracle High Availability Services has been started.

Oracle Clusterware active version on the cluster is [18.0.0.0.0]. The cluster upgrade state is [NORMAL]. The cluster active patch level is [4127601284].

SQL Patching tool version 18.0.0.0.0 Production on Wed Jan 15 13:57:08 2020

Copyright (c) 2012, 2019, Oracle.  All rights reserved.




Log file for this invocation: /u01/app/grid/cfgtoollogs/sqlpatch/sqlpatch_360667_2020_01_15_13_57_08/sqlpatch_invocation.log




Connecting to database...OK

Gathering database info...done




Note:  Datapatch will only apply or rollback SQL fixes for PDBs

       that are in an open state, no patches will be applied to closed PDBs.

       Please refer to Note: Datapatch: Database 12c Post Patch SQL Automation

       (Doc ID 1585822.1)




Bootstrapping registry and package to current versions...done

Determining current state...done




Current state of interim SQL patches:

  No interim patches found




Current state of release update SQL patches:

  Binary registry:

    18.6.0.0.0 Release_Update 1903190102: Installed

  PDB CDB$ROOT:

    Applied 18.2.0.0.0 Release_Update 1804041635 successfully on 02-OCT-18 03.11.47.950696 PM

  PDB PDB$SEED:

    Applied 18.2.0.0.0 Release_Update 1804041635 successfully on 02-OCT-18 03.16.34.680672 PM




Adding patches to installation queue and performing prereq checks...done

Installation queue:

  For the following PDBs: CDB$ROOT PDB$SEED

    No interim patches need to be rolled back

    Patch 29301631 (Database Release Update : 18.6.0.0.190416 (29301631)):

      Apply from 18.2.0.0.0 Release_Update 1804041635 to 18.6.0.0.0 Release_Update 1903190102

    No interim patches need to be applied




Installing patches...

Patch installation complete.  Total patches installed: 2




Validating logfiles...done

Patch 29301631 apply (pdb CDB$ROOT): SUCCESS

  logfile: /u01/app/grid/cfgtoollogs/sqlpatch/29301631/22832106/29301631_apply__MGMTDB_CDBROOT_2020Jan15_13_58_14.log (no errors)

Patch 29301631 apply (pdb PDB$SEED): SUCCESS

  logfile: /u01/app/grid/cfgtoollogs/sqlpatch/29301631/22832106/29301631_apply__MGMTDB_PDBSEED_2020Jan15_14_01_22.log (no errors)

SQL Patching tool complete on Wed Jan 15 14:02:41 2020

MGTCA-1005 : Could not connect to the GIMR.




Listener refused the connection with the following error:




ORA-12514, TNS:listener does not currently know of service requested in connect descriptor










2020/01/15 14:02:49 CLSRSC-180: An error occurred while executing the command '/u01/app/18.0.0/grid/bin/mgmtca applysql'




After fixing the cause of failure Run opatchauto resume




]

OPATCHAUTO-68061: The orchestration engine failed.

OPATCHAUTO-68061: The orchestration engine failed with return code 1

OPATCHAUTO-68061: Check the log for more details.

OPatchAuto failed.




OPatchauto session completed at Wed Jan 15 14:02:50 2020

Time taken to complete the session 28 minutes, 34 seconds




 opatchauto failed with error code 42

[root@exa03vm02 ~]#

 

ORA-01017: invalid username/password; logon denied

 

CRS-4123: Oracle High Availability Services has been started.

Oracle Clusterware active version on the cluster is [18.0.0.0.0]. The cluster upgrade state is [NORMAL]. The cluster active patch level is [4127601284].

SQL Patching tool version 18.0.0.0.0 Production on Fri Nov 29 10:32:41 2019

Copyright (c) 2012, 2019, Oracle.  All rights reserved.




Log file for this invocation: /u01/app/grid/cfgtoollogs/sqlpatch/sqlpatch_128123_2019_11_29_10_32_41/sqlpatch_invocation.log




Connecting to database...OK

Gathering database info...done




Note:  Datapatch will only apply or rollback SQL fixes for PDBs

       that are in an open state, no patches will be applied to closed PDBs.

       Please refer to Note: Datapatch: Database 12c Post Patch SQL Automation

       (Doc ID 1585822.1)




Bootstrapping registry and package to current versions...done

Determining current state...done




Current state of interim SQL patches:

  No interim patches found




Current state of release update SQL patches:

  Binary registry:

    18.6.0.0.0 Release_Update 1903190102: Installed

  PDB CDB$ROOT:

    Applied 18.2.0.0.0 Release_Update 1804041635 successfully on 02-OCT-18 01.32.26.561113 PM

  PDB GIMR_DSCREP_10:

    Applied 18.2.0.0.0 Release_Update 1804041635 successfully on 02-OCT-18 01.37.12.290147 PM

  PDB PDB$SEED:

    Applied 18.2.0.0.0 Release_Update 1804041635 successfully on 02-OCT-18 01.37.12.290147 PM




Adding patches to installation queue and performing prereq checks...done

Installation queue:

  For the following PDBs: CDB$ROOT PDB$SEED GIMR_DSCREP_10

    No interim patches need to be rolled back

    Patch 29301631 (Database Release Update : 18.6.0.0.190416 (29301631)):

      Apply from 18.2.0.0.0 Release_Update 1804041635 to 18.6.0.0.0 Release_Update 1903190102

    No interim patches need to be applied




Installing patches...

Patch installation complete.  Total patches installed: 3




Validating logfiles...done

Patch 29301631 apply (pdb CDB$ROOT): SUCCESS

  logfile: /u01/app/grid/cfgtoollogs/sqlpatch/29301631/22832106/29301631_apply__MGMTDB_CDBROOT_2019Nov29_10_33_45.log (no errors)

Patch 29301631 apply (pdb PDB$SEED): SUCCESS

  logfile: /u01/app/grid/cfgtoollogs/sqlpatch/29301631/22832106/29301631_apply__MGMTDB_PDBSEED_2019Nov29_10_36_02.log (no errors)

Patch 29301631 apply (pdb GIMR_DSCREP_10): SUCCESS

  logfile: /u01/app/grid/cfgtoollogs/sqlpatch/29301631/22832106/29301631_apply__MGMTDB_GIMR_DSCREP_10_2019Nov29_10_36_02.log (no errors)

SQL Patching tool complete on Fri Nov 29 10:38:04 2019

MGTCA-1005 : Could not connect to the GIMR.




ORA-01017: invalid username/password; logon denied










2019/11/29 10:38:13 CLSRSC-180: An error occurred while executing the command '/u01/app/18.0.0/grid/bin/mgmtca applysql'




After fixing the cause of failure Run opatchauto resume




]

OPATCHAUTO-68061: The orchestration engine failed.

OPATCHAUTO-68061: The orchestration engine failed with return code 1

OPATCHAUTO-68061: Check the log for more details.

OPatchAuto failed.




OPatchauto session completed at Fri Nov 29 10:38:15 2019

Time taken to complete the session 26 minutes, 48 seconds




 opatchauto failed with error code 42

[root@exa01vm02 ~]#

 

And inside of the opatch log you can reach/see something like:

 

SQL Patching tool complete on Fri Nov 29 10:44:09 2019

MGTCA-1005 : Could not connect to the GIMR.




ORA-01017: invalid username/password; logon denied










2019/11/29 10:44:18 CLSRSC-180: An error occurred while executing the command '/u01/app/18.0.0/grid/bin/mgmtca applysql'




After fixing the cause of failure Run opatchauto resume




]]. Failures:

OPATCHAUTO-68067: Check the details to determine the cause of the failure.

        at com.oracle.glcm.patch.auto.action.PatchActionExecutor.execute(PatchActionExecutor.java:172)

        at com.oracle.glcm.patch.auto.wizard.silent.tasks.PatchActionTask.execute(PatchActionTask.java:102)

        ... 2 more

2019-11-29 10:44:19,660 INFO  [1] com.oracle.glcm.patch.auto.db.integration.model.productsupport.DBBaseProductSupport - Space available after session: 29898 MB

2019-11-29 10:44:19,728 SEVERE [1] com.oracle.glcm.patch.auto.OPatchAuto - OPatchAuto failed.

com.oracle.glcm.patch.auto.OPatchAutoException: OPATCHAUTO-68061: The orchestration engine failed.

OPATCHAUTO-68061: The orchestration engine failed with return code 1

OPATCHAUTO-68061: Check the log for more details.

        at com.oracle.glcm.patch.auto.OrchestrationEngineImpl.orchestrate(OrchestrationEngineImpl.java:40)

        at com.oracle.glcm.patch.auto.OPatchAuto.orchestrate(OPatchAuto.java:858)

        at com.oracle.glcm.patch.auto.OPatchAuto.orchestrate(OPatchAuto.java:398)

        at com.oracle.glcm.patch.auto.OPatchAuto.orchestrate(OPatchAuto.java:344)

        at com.oracle.glcm.patch.auto.OPatchAuto.main(OPatchAuto.java:212)

2019-11-29 10:44:19,729 INFO  [1] com.oracle.cie.common.util.reporting.CommonReporter - Reporting console output : Message{id='null', message='OPATCHAUTO-68061: The orchestration engine failed.

OPATCHAUTO-68061: The orchestration engine failed with return code 1

OPATCHAUTO-68061: Check the log for more details.'}

2019-11-29 10:44:19,729 INFO  [1] com.oracle.cie.common.util.reporting.CommonReporter - Reporting console output : Message{id='null', message='OPatchAuto failed.'}

^C

[root@exa01vm02 ~]#

 

But the common in all cases occurred during the “/u01/app/18.0.0/grid/bin/mgmtca applysql” call because the opatch detected that is need to apply the patch over MGMTDB, and tried to apply (and reached the error).

 

Solving the issue

After some dig in the web about the error and MOS/Metalink search, nothing was found. No error, hint or something related. Even after one SR opened the report was always the same: inconclusive and not relevant to build a fix. But for all one solution/workaround can be applied and worked every time.
Before, just to show that if you look inside the MGMTDB itself you can see that the patch was applied correctly:

 

[grid@exa01vm02 -MGMTDB]$ sqlplus / as sysdba




SQL*Plus: Release 18.0.0.0.0 - Production on Thu Nov 14 09:19:05 2019

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle.  All rights reserved.







Connected to:

Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0




SQL> show pdbs




    CON_ID CON_NAME                       OPEN MODE  RESTRICTED

---------- ------------------------------ ---------- ----------

         2 PDB$SEED                       READ ONLY  NO

SQL> select instance_name from v$instance;




INSTANCE_NAME

----------------

-MGMTDB




SQL> exit

Disconnected from Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$ cd $ORACLE_HOME/OPatch

[grid@exa01vm02 -MGMTDB]$ ./datapatch -verbose

SQL Patching tool version 18.0.0.0.0 Production on Thu Nov 14 09:21:33 2019

Copyright (c) 2012, 2019, Oracle.  All rights reserved.




Log file for this invocation: /u01/app/grid/cfgtoollogs/sqlpatch/sqlpatch_242627_2019_11_14_09_21_33/sqlpatch_invocation.log




Connecting to database...OK

Gathering database info...done




Note:  Datapatch will only apply or rollback SQL fixes for PDBs

       that are in an open state, no patches will be applied to closed PDBs.

       Please refer to Note: Datapatch: Database 12c Post Patch SQL Automation

       (Doc ID 1585822.1)




Bootstrapping registry and package to current versions...done

Determining current state...done




Current state of interim SQL patches:

  No interim patches found




Current state of release update SQL patches:

  Binary registry:

    18.6.0.0.0 Release_Update 1903190102: Installed

  PDB CDB$ROOT:

    Applied 18.6.0.0.0 Release_Update 1903190102 successfully on 13-NOV-19 05.17.48.183101 PM

  PDB PDB$SEED:

    Applied 18.6.0.0.0 Release_Update 1903190102 successfully on 13-NOV-19 05.17.50.152110 PM




Adding patches to installation queue and performing prereq checks...done

Installation queue:

  For the following PDBs: CDB$ROOT PDB$SEED

    No interim patches need to be rolled back

    No release update patches need to be installed

    No interim patches need to be applied




SQL Patching tool complete on Thu Nov 14 09:21:48 2019

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$ sqlplus / as sysdba




SQL*Plus: Release 18.0.0.0.0 - Production on Thu Nov 14 09:23:35 2019

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle.  All rights reserved.







Connected to:

Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0




SQL> set linesize 250

SQL> col comments format a50

SQL> select ACTION_TIME,VERSION,COMMENTS from dba_registry_history;




ACTION_TIME                                                                 VERSION                        COMMENTS

--------------------------------------------------------------------------- ------------------------------ --------------------------------------------------

                                                                            18                             RDBMS_18.6.0.0.0DBRU_LINUX.X64_190313.1

19-JUN-18 01.51.48.687187 PM                                                18.0.0.0.0                     Patch applied from 18.1.0.0.0 to 18.2.0.0.0

13-NOV-19 05.16.14.040821 PM                                                18.0.0.0.0                     Patch applied from 18.2.0.0.0 to 18.6.0.0.0




SQL> exit

Disconnected from Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0

[grid@exa01vm02 -MGMTDB]$

 

If you look above even the datapatch and database itself report that patch was applied correctly:

 

[grid@exa01vm02 -MGMTDB]$ sqlplus / as sysdba




SQL*Plus: Release 18.0.0.0.0 - Production on Thu Nov 14 09:19:05 2019

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle.  All rights reserved.







Connected to:

Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0




SQL> show pdbs




    CON_ID CON_NAME                       OPEN MODE  RESTRICTED

---------- ------------------------------ ---------- ----------

         2 PDB$SEED                       READ ONLY  NO

SQL> select instance_name from v$instance;




INSTANCE_NAME

----------------

-MGMTDB




SQL> exit

Disconnected from Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$ cd $ORACLE_HOME/OPatch

[grid@exa01vm02 -MGMTDB]$ ./datapatch -verbose

SQL Patching tool version 18.0.0.0.0 Production on Thu Nov 14 09:21:33 2019

Copyright (c) 2012, 2019, Oracle.  All rights reserved.




Log file for this invocation: /u01/app/grid/cfgtoollogs/sqlpatch/sqlpatch_242627_2019_11_14_09_21_33/sqlpatch_invocation.log




Connecting to database...OK

Gathering database info...done




Note:  Datapatch will only apply or rollback SQL fixes for PDBs

       that are in an open state, no patches will be applied to closed PDBs.

       Please refer to Note: Datapatch: Database 12c Post Patch SQL Automation

       (Doc ID 1585822.1)




Bootstrapping registry and package to current versions...done

Determining current state...done




Current state of interim SQL patches:

  No interim patches found




Current state of release update SQL patches:

  Binary registry:

    18.6.0.0.0 Release_Update 1903190102: Installed

  PDB CDB$ROOT:

    Applied 18.6.0.0.0 Release_Update 1903190102 successfully on 13-NOV-19 05.17.48.183101 PM

  PDB PDB$SEED:

    Applied 18.6.0.0.0 Release_Update 1903190102 successfully on 13-NOV-19 05.17.50.152110 PM




Adding patches to installation queue and performing prereq checks...done

Installation queue:

  For the following PDBs: CDB$ROOT PDB$SEED

    No interim patches need to be rolled back

    No release update patches need to be installed

    No interim patches need to be applied




SQL Patching tool complete on Thu Nov 14 09:21:48 2019

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$

[grid@exa01vm02 -MGMTDB]$ sqlplus / as sysdba




SQL*Plus: Release 18.0.0.0.0 - Production on Thu Nov 14 09:23:35 2019

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle.  All rights reserved.







Connected to:

Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0




SQL> set linesize 250

SQL> col comments format a50

SQL> select ACTION_TIME,VERSION,COMMENTS from dba_registry_history;




ACTION_TIME                                                                 VERSION                        COMMENTS

--------------------------------------------------------------------------- ------------------------------ --------------------------------------------------

                                                                            18                             RDBMS_18.6.0.0.0DBRU_LINUX.X64_190313.1

19-JUN-18 01.51.48.687187 PM                                                18.0.0.0.0                     Patch applied from 18.1.0.0.0 to 18.2.0.0.0

13-NOV-19 05.16.14.040821 PM                                                18.0.0.0.0                     Patch applied from 18.2.0.0.0 to 18.6.0.0.0




SQL> exit

Disconnected from Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0

[grid@exa01vm02 -MGMTDB]$

 

The Workaround

The workaround tries to bypass some checks executed to opatch when checking MGMTDB. If you look at the errors above, they were caused because CRS can’t communicate correctly with the database. Maybe because of some credential failure inside of CRS or listener registration.
The first step is to remove completely the MGMTDB. To do that you can use the dbca (from GI home and only one node):

 

[grid@exa01vm02 +ASM2]$ $ORACLE_HOME/bin/dbca -silent -deleteDatabase -sourceDB -MGMTDB

[WARNING] [DBT-19202] The Database Configuration Assistant will delete the Oracle instances and datafiles for your database. All information in the database will be destroyed.

Prepare for db operation

32% complete

Connecting to database

35% complete

39% complete

42% complete

45% complete

48% complete

52% complete

65% complete

Updating network configuration files

68% complete

Deleting instance and datafiles

84% complete

100% complete

Database deletion completed.

Look at the log file "/u01/app/grid/cfgtoollogs/dbca/_mgmtdb/_mgmtdb.log" for further details.

[grid@exa01vm02 +ASM2]$

 

After remove completely, you just need to recreate the MGMTDB. This can be done manually or using the script from MOS. One good source for information is the note How to Move/Recreate GI Management Repository (GIMR / MGMTDB) to Different Shared Storage (Diskgroup, CFS or NFS etc) (Doc ID 1589394.1) where you can find how to create manually or the script (the script just call the dbca).
Here I used the script, you can download it from the note MDBUtil: GI Management Repository configuration tool (Doc ID 2065175.1). The script call DBCA and you just need to specify the diskgroup that you want to use to store the database files.
In this case, I made:

 

[grid@exa01vm02 +ASM2]$ /tmp/mdbutil.pl --addmdb --target=+DATAC2

mdbutil.pl version : 1.98

2019-11-14 09:54:58: I Starting To Configure MGMTDB at +DATAC2...

2019-11-14 09:55:19: I Container database creation in progress... for GI 18.0.0.0.0

2019-11-14 10:12:50: I Plugable database creation in progress...

2019-11-14 10:16:59: I Executing "/tmp/mdbutil.pl --addchm" on exa01vm02 as root to configure CHM.

root@exa01vm02's password:

2019-11-14 10:17:26: I Executing "/tmp/mdbutil.pl --addchm" on exa01vm02 as root to configure CHM.

root@exa01vm02's password:

2019-11-14 10:17:34: I MGMTDB & CHM configuration done!

[grid@exa01vm02 +ASM2]$

 

You can even call with debug mode:

 

[grid@exa02vm03 -MGMTDB]$ /tmp/mdbutil.pl --addmdb --target=+DATAC3 --debug

mdbutil.pl version : 1.98

2020-01-15 15:30:58: D Executing: /u01/app/18.0.0/grid/bin/srvctl status diskgroup -g DATAC3

2020-01-15 15:30:59: D Exit code: 0

2020-01-15 15:30:59: D Output of last command execution:

Disk Group DATAC3 is running on exa02vm03,exa01vm03

2020-01-15 15:30:59: I Starting To Configure MGMTDB at +DATAC3...

2020-01-15 15:30:59: D Executing: /u01/app/18.0.0/grid/bin/srvctl status mgmtlsnr

2020-01-15 15:30:59: D Exit code: 0

2020-01-15 15:30:59: D Output of last command execution:

Listener MGMTLSNR is enabled

2020-01-15 15:30:59: D Executing: /u01/app/18.0.0/grid/bin/srvctl status mgmtdb

2020-01-15 15:31:00: D Exit code: 1

2020-01-15 15:31:00: D Output of last command execution:

PRCD-1120 : The resource for database _mgmtdb could not be found.

2020-01-15 15:31:00: D Executing: /u01/app/18.0.0/grid/bin/srvctl status mgmtdb

2020-01-15 15:31:01: D Exit code: 1

2020-01-15 15:31:01: D Output of last command execution:

PRCD-1120 : The resource for database _mgmtdb could not be found.

2020-01-15 15:31:01: D Executing: /u01/app/18.0.0/grid/bin/srvctl stop mgmtlsnr

2020-01-15 15:31:05: D Exit code: 0

2020-01-15 15:31:05: D Output of last command execution:

2020-01-15 15:31:05: D Executing: /u01/app/18.0.0/grid/bin/crsctl query crs activeversion

2020-01-15 15:31:05: D Exit code: 0

2020-01-15 15:31:05: D Output of last command execution:

Oracle Clusterware active version on the cluster is [18.0.0.0.0]

2020-01-15 15:31:05: D Executing: /u01/app/18.0.0/grid/bin/srvctl enable qosmserver

2020-01-15 15:31:06: D Exit code: 2

2020-01-15 15:31:06: D Output of last command execution:

PRKF-1321 : QoS Management Server is already enabled.

2020-01-15 15:31:06: D Executing: /u01/app/18.0.0/grid/bin/srvctl start qosmserver

2020-01-15 15:31:06: D Exit code: 2

2020-01-15 15:31:06: D Output of last command execution:

PRCC-1014 : qosmserver was already running

2020-01-15 15:31:06: I Container database creation in progress... for GI 18.0.0.0.0

2020-01-15 15:31:06: D Executing: /u01/app/18.0.0/grid/bin/dbca  -silent -createDatabase -createAsContainerDatabase true -templateName MGMTSeed_Database.dbc -sid -MGMTDB -gdbName _mgmtdb -storageType ASM -diskGroupName DATAC3 -datafileJarLocation /u01/app/18.0.0/grid/assistants/dbca/templates -characterset AL32UTF8 -autoGeneratePasswords -skipUserTemplateCheck

2020-01-15 15:47:19: D Exit code: 0

2020-01-15 15:47:19: D Output of last command execution:

Prepare for db operation

2020-01-15 15:47:19: I Plugable database creation in progress...

2020-01-15 15:47:19: D Executing: /u01/app/18.0.0/grid/bin/mgmtca -local

2020-01-15 15:49:28: D Exit code: 0

2020-01-15 15:49:28: D Output of last command execution:

2020-01-15 15:49:28: D Executing: scp /tmp/mdbutil.pl exa01vm03:/tmp/

2020-01-15 15:49:28: D Exit code: 0

2020-01-15 15:49:28: D Output of last command execution:

2020-01-15 15:49:28: I Executing "/tmp/mdbutil.pl --addchm" on exa01vm03 as root to configure CHM.

2020-01-15 15:49:28: D Executing: ssh root@exa01vm03 "/tmp/mdbutil.pl --addchm"

root@exa01vm03's password:

2020-01-15 15:49:39: D Exit code: 0

2020-01-15 15:49:39: D Output of last command execution:

mdbutil.pl version : 1.98

2020-01-15 15:49:39: D Executing: scp /tmp/mdbutil.pl exa02vm03:/tmp/

2020-01-15 15:49:39: D Exit code: 0

2020-01-15 15:49:39: D Output of last command execution:

2020-01-15 15:49:39: I Executing "/tmp/mdbutil.pl --addchm" on exa02vm03 as root to configure CHM.

2020-01-15 15:49:39: D Executing: ssh root@exa02vm03 "/tmp/mdbutil.pl --addchm"

root@exa02vm03's password:

2020-01-15 15:49:46: D Exit code: 0

2020-01-15 15:49:46: D Output of last command execution:

mdbutil.pl version : 1.98

2020-01-15 15:49:46: I MGMTDB & CHM configuration done!

[grid@exa02vm03 -MGMTDB]$

 

Resume and finish

So, after recreating the MGMTDB we can resume the opatch:

 

[root@exa02vm03 ~]# opatchauto resume




OPatchauto session is initiated at Wed Jan 15 15:50:15 2020

Session log file is /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/opatchauto2020-01-15_03-50-16PM.log

Resuming existing session with id NRZ1




Starting CRS service on home /u01/app/18.0.0/grid

CRS service started successfully on home /u01/app/18.0.0/grid




OPatchAuto successful.




--------------------------------Summary--------------------------------




Patching is completed successfully. Please find the summary as follows:




Host:exa02vm03

CRS Home:/u01/app/18.0.0/grid

Version:18.0.0.0.0

Summary:




==Following patches were SUCCESSFULLY applied:




Patch: /u01/patches/grid/29301682/28435192

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-15_13-40-42PM_1.log




Patch: /u01/patches/grid/29301682/28547619

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-15_13-40-42PM_1.log




Patch: /u01/patches/grid/29301682/29301631

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-15_13-40-42PM_1.log




Patch: /u01/patches/grid/29301682/29301643

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-15_13-40-42PM_1.log




Patch: /u01/patches/grid/29301682/29302264

Log: /u01/app/18.0.0/grid/cfgtoollogs/opatchauto/core/opatch/opatch2020-01-15_13-40-42PM_1.log










OPatchauto session completed at Wed Jan 15 15:57:24 2020

Time taken to complete the session 7 minutes, 9 seconds

[root@exa02vm03 ~]#

 

But why this work? So, the idea is that when the opatch tries to check if it is needed to apply the patch over the MGMTDB, it will reach it and verify that everything was already there and the call for mgmtca applysql will be successful. And since all the credentials between GI and MGMTDB are fine, the check can be done correctly. Another point is that you are recreating the database with one binary that already has the RU running.
Again, this solution can be applied for these errors when the communication between MGMTDB and GI fails because of credential errors or listener errors. And I include the complete check for opatch error to identify exactly the error (it is just because mgmtca applysql), and if the SQL’s was applied by datpatch and are registered inside of the database. To finish, always open SR.

 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”


19c Grid Infrastructure Upgrade
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

19c Grid Infrastructure Upgrade

 
Upgrade GRID infrastructure is one activity that usually is postponed because it involves a sensible area that, when not works, causes big downtime until be fixed. But, in the last versions, it is not a complicated task and if you follow the basic rules, it works without problems.
Here I will show a little example of how to upgrade the GI from 18.6.0 to 19.5. The steps below were executed at Exadata running version 19.2.7.0.0.191012 and GI 18.6.0.0, but can be done in every environment that supports Oracle GI.

 

Requirements and Plan

 

The basis for every upgrade is, first, check the requirements to do that. This means not just for operational system version, but the current GI version (that one that will be upgraded). For Exadata and GI 19 you can follow the note 19c Grid Infrastructure and Database Upgrade steps for Exadata Database Machine running on Oracle Linux (Doc ID 2542082.1) but can use as a guide even in the non-Exadata environment.
Every upgrade or patch needs to be planned to be successful. Check previously the current versions, backup, and others are the basis. There is no rule of thumb, but a base plan has:
  • Check operation system requirements (like version).
  • Check current GI requirements (maybe install additional patches, like 28553832 for Exadata).
  • Check system requirements (space and others).
  • Check Oracle Home Requirements (maybe you need to apply additional patches for Oracle Homes).
  • Download and copy to one server that will be patched (this includes the patches to be applied over the new GI).
  • Prepare the X (you can use xauth and redirection through ssh).
You can follow the note hinted previously to see all the requirements. Another REQUIRED reference is the blog of Mike Dietrich that is dedicated to Oracle Upgrade Practices. Since I am doing this at Oracle Exadata, I downloaded the compatible versions from note Exadata Database Machine and Exadata Storage Server Supported Versions (Doc ID 888828.1).

 

Pre-Upgrade

 

When you start the upgrade phase you already have all the requirements OK (operational system, previous GI patched). The steps here are based in the note 19c Grid Infrastructure and Database Upgrade steps for Exadata Database Machine running on Oracle Linux (Doc ID 2542082.1) but I will include some extra details and tricks that will help you.

 

Memory parameters

It is necessary to have at least 3GB for SGA to correctly upgrade (and runs) the GI 19c version. Check this in the current GI to guarantee that this is OK (executed as grid user):

 

[grid@exacl04n1 +ASM1]$ sqlplus / as sysdba




SQL*Plus: Release 18.0.0.0.0 - Production on Fri Jan 24 14:57:24 2020

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle.  All rights reserved.







Connected to:

Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0




SQL> show parameter memory_target




NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

memory_target                        big integer 0

SQL> show parameter memory_max_target




NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

memory_max_target                    big integer 0

SQL> show parameter use_large_pages




NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

use_large_pages                      string      TRUE

SQL> show parameter sga_max_size




NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

sga_max_size                         big integer 3G

SQL> show parameter sga_target




NAME                                 TYPE        VALUE

------------------------------------ ----------- ------------------------------

sga_target                           big integer 3G

SQL> exit

Disconnected from Oracle Database 18c Enterprise Edition Release 18.0.0.0.0 - Production

Version 18.6.0.0.0

[grid@exacl04n1 +ASM1]$

 

Folders and Unzip

Create all the folders for oracle GI installation. This means (usually) the /u01/app/19.0.0.0/grid for GI at Exadata.

 

[root@exacl04n1 ~]# dcli -g /zfs/EXADATA/exacl04.txt -l root mkdir -p /u01/app/19.0.0.0/grid

[root@exacl04n1 ~]# dcli -g /zfs/EXADATA/exacl04.txt -l root chown grid /u01/app/19.0.0.0/grid

[root@exacl04n1 ~]# dcli -g /zfs/EXADATA/exacl04.txt -l root chgrp -R oinstall /u01/app/19.0.0.0/grid

[root@exacl04n1 ~]#

 

In Exadata, I used dcli to do that automatically in every node of the cluster. But remember to check the ownership of the folder in every node.
After you create the folders, in the first node (where +ASM1 runs), you can unzip the GI Base version for 19c (downloaded from edelivery.oracle.com) in the folder for GI destination. Remember to execute this as the user that will run GI, in this case, I made with grid user and unzipped directly from an NFS folder (but you can put in a local folder and unzip):

 

[grid@exacl04n1 +ASM1]$ unzip -q /nfs/19c/Exadata-Patch/19c-Grid/V982068-01.zip -d /u01/app/19.0.0.0/grid

[grid@exacl04n1 +ASM1]$

 

runcluvfy

Cluster verify utility is amazing because automates a lot of tests and checks. It is needed to execute before you continue because it will report to you the success (if everything is OK), or error and points to be fixed.
In this case, I made (the output was cropped but a full output can be checked here):

 

[grid@exacl04n1 +ASM1]$ cd /u01/app/19.0.0.0/grid/

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ ./runcluvfy.sh stage -pre crsinst -upgrade -rolling -src_crshome /u01/app/18.0.0/grid -dest_crshome /u01/app/19.0.0.0/grid -dest_version 19.0.0.0.0 -fixup -verbose




Verifying Physical Memory ...

  Node Name     Available                 Required                  Status

  ------------  ------------------------  ------------------------  ----------

  exacl04n2     84.1372GB (8.8224204E7KB)  8GB (8388608.0KB)         passed

  exacl04n1     84.1372GB (8.8224204E7KB)  8GB (8388608.0KB)         passed



Verifying ASM Filter Driver configuration ...PASSED

Verifying Systemd login manager IPC parameter ...PASSED




Pre-check for cluster services setup was successful.

Verifying RPM Package Manager database ...INFORMATION

PRVG-11250 : The check "RPM Package Manager database" was not performed because

it needs 'root' user privileges.







CVU operation performed:      stage -pre crsinst

Date:                         Jan 24, 2020 3:17:24 PM

CVU home:                     /u01/app/19.0.0.0/grid/

User:                         grid

[grid@exacl04n1 +ASM1]$

 

As you can see, was a success. But one example of a failed report is:

 

Checks did not pass for the following nodes:

        zeroing02,zeroing01







Failures were encountered during execution of CVU verification request "stage -pre crsinst".




Verifying Node Connectivity ...FAILED

zeroing02: PRVG-11068 : Highly Available IP (HAIP) is enabled on the nodes

            "zeroing01,zeroing02".




zeroing01: PRVG-11068 : Highly Available IP (HAIP) is enabled on the nodes

            "zeroing01,zeroing02".




Verifying RPM Package Manager database ...INFORMATION

PRVG-11250 : The check "RPM Package Manager database" was not performed because

it needs 'root' user privileges.







CVU operation performed:      stage -pre crsinst

Date:                         Dec 9, 2019 1:38:46 PM

CVU home:                     /u01/app/19.0.0.0/grid/

User:                         oracle

[root@zeroing01 ~]#

 

Prepare Patch for the New GI

It is possible to patch the new GI even before install the 19c GI (the same was possible to do at 12 and 18c too). It is recommended by the way. Fortunately, you don’t need to patch manually, you just call the gridSetup.sh with -applyRU if it is Release Update (or -applyRUR for RUR patch).
Since some files will be overwritten, I create in node 1 one new folder to backup them (executed as the same owner from GI):

 

[grid@exacl04n1 +ASM1]$ mkdir /u01/patches/grid-19c

[grid@exacl04n1 +ASM1]$

 

To apply these patches correctly it is needed to update the opatch from the GI base release (that was unzipped previously). Execute this as the same user that will run GI:

 

[grid@exacl04n1 +ASM1]$ #backup current opatch to the folder created before

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ cp /u01/app/19.0.0.0/grid/OPatch /u01/patches/grid-19c/OPatch-ORG -R

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ # go to GI home folder

[grid@exacl04n1 +ASM1]$ cd /u01/app/19.0.0.0/grid

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ # unzip the new version of Opatch for 19c

[grid@exacl04n1 +ASM1]$ unzip -qa /zfs/EXADATA_PATCHING/19c/Exadata-Patch/p6880880_190000_Linux-x86-64.zip

replace OPatch/emdpatch.pl? [y]es, [n]o, [A]ll, [N]one, [r]ename: A

[grid@exacl04n1 +ASM1]$

 

As you saw, I create the backup of Opatch and overwrote with the new version. Remember to execute this with the same user that will runs GI.
Another point is unzipping the patch that you want to apply. Here I used the same folder create before to store the patch. Everything executed in the first node and as the user for GI:

 

[grid@exacl04n1 +ASM1]$ cd /u01/patches/grid-19c/

[grid@exacl04n1 +ASM1]$ unzip -qa /zfs/EXADATA_PATCHING/19c/Exadata-Patch/19c-Grid/p30116789_190000_Linux-x86-64.zip

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ ls -l

total 232

drwxr-x---  7 grid oinstall   4096 Oct  9 17:11 30116789

drwxr-x--- 14 grid oinstall   4096 Jan 24 15:23 OPatch-ORG

-rw-rw-r--  1 grid oinstall 225499 Oct 15 13:24 PatchSearch.xml

[grid@exacl04n1 +ASM1]$

 

The version here was downloaded from node 888828.1 for Exadata and it is the GI Release Update 19.5.0.0 – Patch 30116789.

 

Fix know issues

Little know issues need to be fixed before you upgrade the GI. These errors were based on my experience during passed upgraded (These errors were not reported by cluster verify utility).

 

Lock for oraInventory

The first is related to inventory lock files. It is common to have an environment with role separation (where GI runs with a different user than OH) and can occur that lock file can be set as other ownership. To fix execute in both nodes (or dcli for Exadata):

 

[root@exacl04n1 ~]# dcli -g /zfs/EXADATA/exacl04.txt -l root "chown grid:oinstall /u01/app/oraInventory/locks -R"

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# dcli -g /zfs/EXADATA/exacl04.txt -l root "ls -l /u01/app/oraInventory/locks"

exacl04n1: total 0

exacl04n1: -rw-r--r-- 1 grid oinstall 0 Jan 20 13:07 inventory.lock

exacl04n2: total 0

exacl04n2: -rw-r--r-- 1 grid oinstall 0 Jan 21 13:36 inventory.lock

[root@exacl04n1 ~]#

 

ACFS for ghchkpt

Can occur that GHCHKPT filesystem was removed (intentionally or no) from the current GI home and during the rootupgrade.sh you can hit the error belowing it does not exists:

 

2019/12/04 17:09:28 CLSRSC-692: Checking whether CRS entities are ready for upgrade. This operation may take a few minutes.

PRCR-1129 : Failed to upgrade resource type ora.rhpserver.type

PRCR-1071 : Failed to register or update resource ora.rhpserver

CRS-2510: Resource 'ora.datac3.ghchkpt.acfs' used in dependency 'hard' does not exist or is not registered.

CRS-2514: Dependency attribute specification 'hard' is invalid in resource 'ora.rhpserver'

2019/12/04 17:11:14 CLSRSC-180: An error occurred while executing the command '/u01/app/19.0.0.0/grid/bin/srvctl upgrade model  -s 18.0.0.0.0 -d 19.0.0.0.0 -p first'

2019/12/04 17:11:17 CLSRSC-694: failed to validate CRS entities for upgrade, aborting the upgrade

2019/12/04 17:11:17 CLSRSC-362: The pre-upgrade checks failed, aborting the upgrade

Died at /u01/app/19.0.0.0/grid/crs/install/crsupgrade.pm line 3772.

[root@exacl03n1 ~]#

 

To avoid this you need to recreate it manually doing this:

 

[root@exacl04n1 ~]# su - grid

Last login: Fri Jan 24 14:51:16 CET 2020 from 35.213.248.138 on ssh

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ #CHECK if ghchkpt exists. In this case NO

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ crsctl stat res -w "TYPE = ora.acfs.type" -p | grep VOLUME

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/acfsoh-256

VOLUME_DEVICE=/dev/asm/acfsoh-256

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/acfsoh-256

VOLUME_DEVICE=/dev/asm/acfsoh-256

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ #Create the volume at ASM. The size it is important

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ asmcmd

ASMCMD>

ASMCMD> volcreate -G DATAC4 -s 1536M ghchkpt

ASMCMD>

ASMCMD> volinfo -G DATAC4 ghchkpt

Diskgroup Name: DATAC4




         Volume Name: GHCHKPT

         Volume Device: /dev/asm/ghchkpt-256

         State: ENABLED

         Size (MB): 1536

         Resize Unit (MB): 512

         Redundancy: MIRROR

         Stripe Columns: 8

         Stripe Width (K): 1024

         Usage:

         Mountpath:




ASMCMD>

ASMCMD> exit

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ #Format the volume as ACFS

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ /sbin/mkfs -t acfs /dev/asm/ghchkpt-256

mkfs.acfs: version                   = 18.0.0.0.0

mkfs.acfs: on-disk version           = 46.0

mkfs.acfs: volume                    = /dev/asm/ghchkpt-256

mkfs.acfs: volume size               = 1610612736  (   1.50 GB )

mkfs.acfs: Format complete.

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$




[root@exacl04n1 ~]#

[root@exacl04n1 ~]# #As root, add the filesystem at current GI

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# export ORACLE_HOME=/u01/app/18.0.0/grid

[root@exacl04n1 ~]# $ORACLE_HOME/bin/srvctl add filesystem -d /dev/asm/ghchkpt-256 -g DATAC4 -v GHCHKPT -m /mnt/oracle/rhpimages/chkbase/ -user oracle,grid

[root@exacl04n1 ~]#

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# #Check that everything is OK

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# $ORACLE_HOME/bin/crsctl stat res -t |grep acfs

ora.datac4.acfsoh.acfs

               ONLINE  ONLINE       exacl04n1                mounted on /u01/acfs

               ONLINE  ONLINE       exacl04n2                mounted on /u01/acfs

ora.datac4.ghchkpt.acfs

[root@exacl04n1 ~]#

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# $ORACLE_HOME/bin/crsctl stat res -w "TYPE = ora.acfs.type" -p | grep VOLUME

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/acfsoh-256

VOLUME_DEVICE=/dev/asm/acfsoh-256

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/acfsoh-256

VOLUME_DEVICE=/dev/asm/acfsoh-256

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/ghchkpt-256

VOLUME_DEVICE=/dev/asm/ghchkpt-256

AUX_VOLUMES=

CANONICAL_VOLUME_DEVICE=/dev/asm/ghchkpt-256

VOLUME_DEVICE=/dev/asm/ghchkpt-256

[root@exacl04n1 ~]#

[root@exacl04n1 ~]#

 

Important here is the size for GHCHKPT volume that needs to be 1536M and registered at current GI. When registered the filesystem remember to set correctly the volume name.

 

ACFS filesystem and CRS resources

If you have some mountpoint over ACFS, it is recommended to stop it “safely” before executing the gridSetup. This occurs because during the upgrade phase the script will try to shutdown the entire clusters in the node,  and if the unmount of ACFS goes wrong you will receive an error (and this can be boring and stressful to handle). The same is valid for resources that you registered at CRS.
So, just verify ACFS and unmount it (including stop the ACFS filesystem) and stop additional CRS resources:

 

[root@exacl04n1 ~]# /u01/app/18.0.0/grid/bin/crsctl stop res flk.mountbind -n exacl04n1

CRS-2673: Attempting to stop 'flk.mountbind' on 'exacl04n1'

CRS-2675: Stop of 'flk.mountbind' on 'exacl04n1' failed

CRS-2679: Attempting to clean 'flk.mountbind' on 'exacl04n1'

CRS-2681: Clean of 'flk.mountbind' on 'exacl04n1' succeeded

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# #Stopped two times (the error above was expected)

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# /u01/app/18.0.0/grid/bin/crsctl stop res flk.mountbind -n exacl04n1

CRS-2500: Cannot stop resource 'flk.mountbind' as it is not running

CRS-4000: Command Stop failed, or completed with errors.

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# #Stop ACFS

[root@exacl04n1 ~]#

[root@exacl04n1 ~]# /u01/app/18.0.0/grid/bin/srvctl stop filesystem -d /dev/asm/acfsoh-256 -n exacl04n1

[root@exacl04n1 ~]#

 

Run gridSetup.sh

After preparing everything, patches, folders, and fix some issues we can call the gridSetup.sh to install the Oracle GI 19c.
The first is clear all the current session variables that point to the current GI, and set some variables (as DISPLAY for X and move to the new GI home to call the grid setup):

 

[root@exacl04n1 ~]# su - grid

.Last login: Fri Jan 24 16:06:12 CET 2020 from 35.213.248.138 on ssh

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ unset ORACLE_HOME

[grid@exacl04n1 +ASM1]$ unset ORACLE_BASE

[grid@exacl04n1 +ASM1]$ unset ORACLE_SID

[grid@exacl04n1 ]$

[grid@exacl04n1 ]$ cd /u01/app/19.0.0.0/grid/

[grid@exacl04n1 ]$

[grid@exacl04n1 ]$ export DISPLAY=35.213.248.116:1.0

[grid@exacl04n1 ]$

 

Now the gridSetup.sh can be called, and I do this with some special parameters:
  • -applyRU: This will apply the Release Update passed as parameter BEFORE start to install the 19c GI.
  • mgmtDB parameters: With 19c the MGMTDB it is not needed anymore, these parameters disable the installation and even the options are not shown during the graphical installation.
So, the call it is:

 

[grid@exacl04n1 ]$ ./gridSetup.sh -applyRU /u01/patches/grid-19c/30116789 -J-Doracle.install.mgmtDB=false -J-Doracle.install.mgmtDB.CDB=false -J Doracle.install.crs.enableRemoteGIMR=false

Preparing the home to patch...

Applying the patch /u01/patches/grid-19c/30116789...

Successfully applied the patch.

The log can be found at: /u01/app/oraInventory/logs/GridSetupActions2020-01-17_04-49-57PM/installerPatchActions_2020-01-17_04-49-57PM.log

Launching Oracle Grid Infrastructure Setup Wizard...




The response file for this session can be found at:

 /u01/app/19.0.0.0/grid/install/response/grid_2020-01-17_04-49-57PM.rsp

 

As you saw, the report says that the patch 30116789 was installed over the GI home (this took around 10 minutes). Just after that the graphical installer appears and the installation is basically NNF (next next finish) until the request for root upgrade scripts appears:

 

 

rootupgrade.sh – Node 1

At this point it is needed to execute, node by node, the rootupgrade.sh to “finish” the installation. The rootupgrade.sh execute important steps like:
  • TFA upgrade
  • Upgrade OCR
  • Upgrade ACFS driver
  • Upgrade GI scripts
If something wrong occurs here (mainly after step 3) the recommendation is open SR to verify the errors and if the issue.
Another point it is stop all Oracle Databases running in the node before executing the script.
So, in the first node was called the rootupgrade.sh:

 

[root@exacl04n1 ~]# /u01/app/19.0.0.0/grid/rootupgrade.sh

Performing root user operation.




The following environment variables are set as:

    ORACLE_OWNER= grid

    ORACLE_HOME=  /u01/app/19.0.0.0/grid




Enter the full pathname of the local bin directory: [/usr/local/bin]:

The contents of "dbhome" have not changed. No need to overwrite.

The file "oraenv" already exists in /usr/local/bin.  Overwrite it? (y/n)

[n]: y

   Copying oraenv to /usr/local/bin ...

The file "coraenv" already exists in /usr/local/bin.  Overwrite it? (y/n)

[n]: y

   Copying coraenv to /usr/local/bin ...




Entries will be added to the /etc/oratab file as needed by

Database Configuration Assistant when a database is created

Finished running generic part of root script.

Now product-specific root actions will be performed.

Relinking oracle with rac_on option

Using configuration parameter file: /u01/app/19.0.0.0/grid/crs/install/crsconfig_params

The log of current session can be found at:

  /u01/app/grid/crsdata/exacl04n1/crsconfig/rootcrs_exacl04n1_2020-01-24_04-42-49PM.log

2020/01/24 16:43:14 CLSRSC-595: Executing upgrade step 1 of 18: 'UpgradeTFA'.

2020/01/24 16:43:14 CLSRSC-4015: Performing install or upgrade action for Oracle Trace File Analyzer (TFA) Collector.

2020/01/24 16:43:14 CLSRSC-595: Executing upgrade step 2 of 18: 'ValidateEnv'.

2020/01/24 16:43:19 CLSRSC-595: Executing upgrade step 3 of 18: 'GetOldConfig'.

2020/01/24 16:43:19 CLSRSC-464: Starting retrieval of the cluster configuration data

2020/01/24 16:47:50 CLSRSC-692: Checking whether CRS entities are ready for upgrade. This operation may take a few minutes.

2020/01/24 16:48:25 CLSRSC-4003: Successfully patched Oracle Trace File Analyzer (TFA) Collector.

2020/01/24 16:50:06 CLSRSC-693: CRS entities validation completed successfully.

2020/01/24 16:50:12 CLSRSC-515: Starting OCR manual backup.

2020/01/24 16:50:48 CLSRSC-516: OCR manual backup successful.

2020/01/24 16:50:57 CLSRSC-486:

 At this stage of upgrade, the OCR has changed.

 Any attempt to downgrade the cluster after this point will require a complete cluster outage to restore the OCR.

2020/01/24 16:50:57 CLSRSC-541:

 To downgrade the cluster:
  1. All nodes that have been upgraded must be downgraded.
2020/01/24 16:50:57 CLSRSC-542:
  1. Before downgrading the last node, the Grid Infrastructure stack on all other cluster nodes must be down.
2020/01/24 16:51:07 CLSRSC-465: Retrieval of the cluster configuration data has successfully completed.

2020/01/24 16:51:07 CLSRSC-595: Executing upgrade step 4 of 18: 'GenSiteGUIDs'.

2020/01/24 16:51:08 CLSRSC-595: Executing upgrade step 5 of 18: 'UpgPrechecks'.

2020/01/24 16:51:13 CLSRSC-363: User ignored prerequisites during installation

2020/01/24 16:51:25 CLSRSC-595: Executing upgrade step 6 of 18: 'SetupOSD'.

2020/01/24 16:51:25 CLSRSC-595: Executing upgrade step 7 of 18: 'PreUpgrade'.

2020/01/24 16:55:02 CLSRSC-468: Setting Oracle Clusterware and ASM to rolling migration mode

2020/01/24 16:55:02 CLSRSC-482: Running command: '/u01/app/18.0.0/grid/bin/crsctl start rollingupgrade 19.0.0.0.0'

CRS-1131: The cluster was successfully set to rolling upgrade mode.

2020/01/24 16:55:09 CLSRSC-482: Running command: '/u01/app/19.0.0.0/grid/bin/asmca -silent -upgradeNodeASM -nonRolling false -oldCRSHome /u01/app/18.0.0/grid -oldCRSVersion 18.0.0.0.0 -firstNode true -startRolling false '




ASM configuration upgraded in local node successfully.




2020/01/24 16:55:12 CLSRSC-469: Successfully set Oracle Clusterware and ASM to rolling migration mode

2020/01/24 16:55:18 CLSRSC-466: Starting shutdown of the current Oracle Grid Infrastructure stack

2020/01/24 16:56:08 CLSRSC-467: Shutdown of the current Oracle Grid Infrastructure stack has successfully completed.

2020/01/24 16:56:11 CLSRSC-595: Executing upgrade step 8 of 18: 'CheckCRSConfig'.

2020/01/24 16:56:12 CLSRSC-595: Executing upgrade step 9 of 18: 'UpgradeOLR'.

2020/01/24 16:56:23 CLSRSC-595: Executing upgrade step 10 of 18: 'ConfigCHMOS'.

2020/01/24 16:56:23 CLSRSC-595: Executing upgrade step 11 of 18: 'UpgradeAFD'.

2020/01/24 16:56:30 CLSRSC-595: Executing upgrade step 12 of 18: 'createOHASD'.

2020/01/24 16:56:37 CLSRSC-595: Executing upgrade step 13 of 18: 'ConfigOHASD'.

2020/01/24 16:56:37 CLSRSC-329: Replacing Clusterware entries in file 'oracle-ohasd.service'

2020/01/24 16:57:29 CLSRSC-595: Executing upgrade step 14 of 18: 'InstallACFS'.

2020/01/24 16:58:01 CLSRSC-595: Executing upgrade step 15 of 18: 'InstallKA'.

2020/01/24 16:58:21 CLSRSC-595: Executing upgrade step 16 of 18: 'UpgradeCluster'.

2020/01/24 17:00:27 CLSRSC-343: Successfully started Oracle Clusterware stack

clscfg: EXISTING configuration version 5 detected.

Successfully taken the backup of node specific configuration in OCR.

Successfully accumulated necessary OCR keys.

Creating OCR keys for user 'root', privgrp 'root'..

Operation successful.

2020/01/24 17:00:52 CLSRSC-595: Executing upgrade step 17 of 18: 'UpgradeNode'.

2020/01/24 17:00:57 CLSRSC-474: Initiating upgrade of resource types

2020/01/24 17:02:09 CLSRSC-475: Upgrade of resource types successfully initiated.

2020/01/24 17:02:24 CLSRSC-595: Executing upgrade step 18 of 18: 'PostUpgrade'.

2020/01/24 17:02:34 CLSRSC-325: Configure Oracle Grid Infrastructure for a Cluster ... succeeded

[root@exacl04n1 ~]#

 

After finish with success, you can start the Oracle Databases in this node, or even relocate RACOneNode databases to this node.

 

rootupgrade.sh – Node 2

Here it is called the script in the second node. But some details before execute it:
  • Stop and unmount ACFS filesystem running in the node
  • Stop any additional resources at CRS running in this node
  • Stop all Databases running in the node (you relocate to others nodes already upgraded
After checking the points above, you can call the script:

 

[root@exacl04n2 ~]# /u01/app/18.0.0/grid/bin/crsctl stop res flk.mountbind -n exacl04n2

CRS-2673: Attempting to stop 'flk.mountbind' on 'exacl04n2'

CRS-2675: Stop of 'flk.mountbind' on 'exacl04n2' failed

CRS-2679: Attempting to clean 'flk.mountbind' on 'exacl04n2'

CRS-2681: Clean of 'flk.mountbind' on 'exacl04n2' succeeded

[root@exacl04n2 ~]#

[root@exacl04n2 ~]#

[root@exacl04n2 ~]# /u01/app/18.0.0/grid/bin/crsctl stop res flk.mountbind -n exacl04n2

CRS-2500: Cannot stop resource 'flk.mountbind' as it is not running

CRS-4000: Command Stop failed, or completed with errors.

[root@exacl04n2 ~]#

[root@exacl04n2 ~]#

[root@exacl04n2 ~]# /u01/app/18.0.0/grid/bin/srvctl stop filesystem -d /dev/asm/acfsoh-256 -n exacl04n2

[root@exacl04n2 ~]#

[root@exacl04n2 ~]#

[root@exacl04n2 ~]# /u01/app/19.0.0.0/grid/rootupgrade.sh

Performing root user operation.




The following environment variables are set as:

    ORACLE_OWNER= grid

    ORACLE_HOME=  /u01/app/19.0.0.0/grid




Enter the full pathname of the local bin directory: [/usr/local/bin]:

The contents of "dbhome" have not changed. No need to overwrite.

The file "oraenv" already exists in /usr/local/bin.  Overwrite it? (y/n)

[n]: y

   Copying oraenv to /usr/local/bin ...

The file "coraenv" already exists in /usr/local/bin.  Overwrite it? (y/n)

[n]: y

   Copying coraenv to /usr/local/bin ...




Entries will be added to the /etc/oratab file as needed by

Database Configuration Assistant when a database is created

Finished running generic part of root script.

Now product-specific root actions will be performed.

Relinking oracle with rac_on option

Using configuration parameter file: /u01/app/19.0.0.0/grid/crs/install/crsconfig_params

The log of current session can be found at:

  /u01/app/grid/crsdata/exacl04n2/crsconfig/rootcrs_exacl04n2_2020-01-24_06-01-55PM.log

2020/01/24 18:02:12 CLSRSC-595: Executing upgrade step 1 of 18: 'UpgradeTFA'.

2020/01/24 18:02:12 CLSRSC-4015: Performing install or upgrade action for Oracle Trace File Analyzer (TFA) Collector.

2020/01/24 18:02:12 CLSRSC-595: Executing upgrade step 2 of 18: 'ValidateEnv'.

2020/01/24 18:02:13 CLSRSC-595: Executing upgrade step 3 of 18: 'GetOldConfig'.

2020/01/24 18:02:13 CLSRSC-464: Starting retrieval of the cluster configuration data

2020/01/24 18:02:42 CLSRSC-465: Retrieval of the cluster configuration data has successfully completed.

2020/01/24 18:02:42 CLSRSC-595: Executing upgrade step 4 of 18: 'GenSiteGUIDs'.

2020/01/24 18:02:43 CLSRSC-595: Executing upgrade step 5 of 18: 'UpgPrechecks'.

2020/01/24 18:02:43 CLSRSC-363: User ignored prerequisites during installation

2020/01/24 18:02:44 CLSRSC-595: Executing upgrade step 6 of 18: 'SetupOSD'.

2020/01/24 18:02:44 CLSRSC-595: Executing upgrade step 7 of 18: 'PreUpgrade'.




ASM configuration upgraded in local node successfully.




2020/01/24 18:03:01 CLSRSC-466: Starting shutdown of the current Oracle Grid Infrastructure stack

2020/01/24 18:04:45 CLSRSC-467: Shutdown of the current Oracle Grid Infrastructure stack has successfully completed.

2020/01/24 18:04:57 CLSRSC-595: Executing upgrade step 8 of 18: 'CheckCRSConfig'.

2020/01/24 18:04:57 CLSRSC-595: Executing upgrade step 9 of 18: 'UpgradeOLR'.

2020/01/24 18:05:05 CLSRSC-595: Executing upgrade step 10 of 18: 'ConfigCHMOS'.

2020/01/24 18:05:06 CLSRSC-595: Executing upgrade step 11 of 18: 'UpgradeAFD'.

2020/01/24 18:05:07 CLSRSC-595: Executing upgrade step 12 of 18: 'createOHASD'.

2020/01/24 18:05:09 CLSRSC-595: Executing upgrade step 13 of 18: 'ConfigOHASD'.

2020/01/24 18:05:09 CLSRSC-329: Replacing Clusterware entries in file 'oracle-ohasd.service'

2020/01/24 18:05:54 CLSRSC-595: Executing upgrade step 14 of 18: 'InstallACFS'.

2020/01/24 18:06:18 CLSRSC-4003: Successfully patched Oracle Trace File Analyzer (TFA) Collector.

2020/01/24 18:06:20 CLSRSC-595: Executing upgrade step 15 of 18: 'InstallKA'.

2020/01/24 18:06:34 CLSRSC-595: Executing upgrade step 16 of 18: 'UpgradeCluster'.

2020/01/24 18:08:31 CLSRSC-343: Successfully started Oracle Clusterware stack

clscfg: EXISTING configuration version 19 detected.

Successfully taken the backup of node specific configuration in OCR.

Successfully accumulated necessary OCR keys.

Creating OCR keys for user 'root', privgrp 'root'..

Operation successful.

2020/01/24 18:08:46 CLSRSC-595: Executing upgrade step 17 of 18: 'UpgradeNode'.

Start upgrade invoked..

2020/01/24 18:08:50 CLSRSC-478: Setting Oracle Clusterware active version on the last node to be upgraded

2020/01/24 18:08:50 CLSRSC-482: Running command: '/u01/app/19.0.0.0/grid/bin/crsctl set crs activeversion'

Started to upgrade the active version of Oracle Clusterware. This operation may take a few minutes.

Started to upgrade CSS.

CSS was successfully upgraded.

Started to upgrade Oracle ASM.

Started to upgrade CRS.

CRS was successfully upgraded.

Started to upgrade Oracle ACFS.

Oracle ACFS was successfully upgraded.

Successfully upgraded the active version of Oracle Clusterware.

Oracle Clusterware active version was successfully set to 19.0.0.0.0.

2020/01/24 18:10:02 CLSRSC-479: Successfully set Oracle Clusterware active version

2020/01/24 18:10:04 CLSRSC-476: Finishing upgrade of resource types

2020/01/24 18:10:18 CLSRSC-477: Successfully completed upgrade of resource types

2020/01/24 18:10:45 CLSRSC-595: Executing upgrade step 18 of 18: 'PostUpgrade'.

Successfully updated XAG resources.

2020/01/24 18:11:21 CLSRSC-325: Configure Oracle Grid Infrastructure for a Cluster ... succeeded

[root@exacl04n2 ~]#

[root@exacl04n2 ~]#

 

As you can see, everything was fine and the cluster was upgraded in both nodes. If you have more nodes, continue with the others.

Continue with Graphical Installation – AND INTENTIONAL ERROR

After that is possible to continue the installation at the graphical side. But we will hit on intentional error.

 

 

During the phase Upgrade RHP Repository will occur one error. But this is expected. Remember that during the gridSetup.sh the mgmtDB parameters were disabled? So, since it was not created, the RHP will fail because the mgmtDB was not found (this occurred in all of 10 clusters that upgraded recently):
If you check in the log you will see details informing that mgmtDB was not found. To solve, just click OK and Skip. Look below that upgrade RHP was Ignored.

Post Upgrade

After finish the upgrade above, everything will be working correctly. But some details need to be checked to deliver everything at 100%.

 

ASM Compatibility

Remember that now the ASM.COMPATIBILITY parameter can be upgraded to allow you to use some new features. Remember, just ASM.COMPATIBILITY for all diskgroups:

 

[grid@exacl04n1 +ASM1]$ echo "ALTER DISKGROUP DATAC4 SET ATTRIBUTE 'compatible.asm' = '19.3.0.0.0';" | sqlplus -s / as sysasm




Diskgroup altered.




[grid@exacl04n1 +ASM1]$ echo "ALTER DISKGROUP RECOC4 SET ATTRIBUTE 'compatible.asm' = '19.3.0.0.0';" | sqlplus -s / as sysasm




Diskgroup altered.




[grid@exacl04n1 +ASM1]$

 

Inventory Update

Since Oracle 19C, GI did not register the nodes in the oraInventory. So, it is needed to add to allow older versions (11, 12 and 18) to see the GI nodes. Simple call resolves this (take attention the correct path and node names):

 

[grid@exacl04n1 +ASM1]$ /u01/app/19.0.0.0/grid/oui/bin/runInstaller -nowait -waitforcompletion -ignoreSysPrereqs -updateNodeList ORACLE_HOME=/u01/app/19.0.0.0/grid "CLUSTER_NODES={exacl04n1,exacl04n2}" CRS=true LOCAL_NODE=exacl04n1

Starting Oracle Universal Installer...




Checking swap space: must be greater than 500 MB.   Actual 16174 MB    Passed

The inventory pointer is located at /etc/oraInst.loc

[grid@exacl04n1 +ASM1]$

 

Remove previous GI

Remember to delete (or at least mark) in the Oracle inventory that previous GI was removed. To do that, you need to call runInstaller (from previous OH) with option detachHome. Be careful to set the ORACLE_HOME correctly (and in just one node):

 

[grid@exacl04n1 +ASM1]$ cat /u01/app/oraInventory/ContentsXML/inventory.xml |grep grid

<HOME NAME="OraGI18Home1" LOC="/u01/app/18.0.0/grid" TYPE="O" IDX="6">

<HOME NAME="OraGI19Home1" LOC="/u01/app/19.0.0.0/grid" TYPE="O" IDX="11" CRS="true">

<HOME NAME="OraGridHome" LOC="/u01/app/12.1.0.2/grid" TYPE="O" IDX="1" REMOVED="T"/>

<HOME NAME="OraGI12Home1" LOC="/u01/app/12.2.0.1/grid" TYPE="O" IDX="5" REMOVED="T"/>

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ unset ORACLE_HOME

[grid@exacl04n1 +ASM1]$ export ORACLE_HOME=/u01/app/18.0.0/grid/

[grid@exacl04n1 +ASM1]$ $ORACLE_HOME/oui/bin/runInstaller -detachHome -silent ORACLE_HOME=/u01/app/18.0.0/grid

Starting Oracle Universal Installer...




Checking swap space: must be greater than 500 MB.   Actual 16174 MB    Passed

The inventory pointer is located at /etc/oraInst.loc

'DetachHome' was successful.

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ cat /u01/app/oraInventory/ContentsXML/inventory.xml |grep grid

<HOME NAME="OraGI19Home1" LOC="/u01/app/19.0.0.0/grid" TYPE="O" IDX="11" CRS="true">

<HOME NAME="OraGridHome" LOC="/u01/app/12.1.0.2/grid" TYPE="O" IDX="1" REMOVED="T"/>

<HOME NAME="OraGI12Home1" LOC="/u01/app/12.2.0.1/grid" TYPE="O" IDX="5" REMOVED="T"/>

<HOME NAME="OraGI18Home1" LOC="/u01/app/18.0.0/grid" TYPE="O" IDX="6" REMOVED="T"/>

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$

 

After that, you can delete the previous GI Home.

 

Others details

To finish, remember to fix the bash_profile and fix oratab (if you have something particularly set there). Remember to execute in both nodes:

 

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ cat .bash_profile |grep HOME

export ORACLE_HOME=/u01/app/18.0.0/grid

export PATH=$ORACLE_HOME/bin:/usr/local/bin:/bin:/usr/bin

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ vi .bash_profile

[grid@exacl04n1 +ASM1]$

[grid@exacl04n1 +ASM1]$ cat .bash_profile |grep HOME

export ORACLE_HOME=/u01/app/19.0.0.0/grid

export PATH=$ORACLE_HOME/bin:/usr/local/bin:/bin:/usr/bin

[grid@exacl04n1 +ASM1]$




#################

#Second node

#################




[root@exacl04n2 ~]# su - grid

Last login: Fri Jan 24 18:17:43 CET 2020 from exacl04n1.flisk.rulz.org on ssh

[grid@exacl04n2 +ASM2]$ cat .bash_profile |grep HOME

export ORACLE_HOME=/u01/app/18.0.0/grid

export PATH=$ORACLE_HOME/bin:/usr/local/bin:/bin:/usr/bin

[grid@exacl04n2 +ASM2]$

[grid@exacl04n2 +ASM2]$ vi .bash_profile

[grid@exacl04n2 +ASM2]$

[grid@exacl04n2 +ASM2]$ cat .bash_profile |grep HOME

export ORACLE_HOME=/u01/app/19.0.0.0/grid

export PATH=$ORACLE_HOME/bin:/usr/local/bin:/bin:/usr/bin

[grid@exacl04n2 +ASM2]$

 

Conclusion

Upgrade GI to the 19c version is easier than the previous versions. If you follow correctly the requirements probably you will not see errors. At the beginning of 2018, I made the same for 18c upgrade (migrating from 12c) and the process was similar. You can check in my post Reaching Exadata 18c.  
But be careful with ACFS and the repost from cluster verify script. The output from it (even if still requires additional checks as I showed before) is a good indication of system health and “OK to continue”.
The process that I showed above was executed in Exadata, but with small adaptions (mainly for dcli), it can be used in other environments as well.
The post-upgrade has some additional tricks as I showed in the previous post about Exadata, workaround for oracka.ko error and TFA error after GI upgrade to 19c. Nothing special but will require additional attention too.

 

 

 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”


ZDLRA, Protection Policies
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

ZDLRA, Protection Policies

For ZDLRA the protection policies have a significant role in the appliance management, but not just that, for the architecture design too. And usually (and unfortunately) policies do not take a lot of attention as deserved.
To create a good ZDLRA design, and avoid future problems, it is important to understand all the requirements for the protection policies and all the impacts. You can check the official documentation for this, but I will explain deeply the details that can pass without you notice them in the documentation.

 

DBMS_RA.CREATE_PROTECTION_POLICY

To create the policy it is easy, just need to use the DBMS_RA.CREATE_PROTECTION_POLICY and set the parameters:

 

SQL> BEGIN

  2  DBMS_RA.CREATE_PROTECTION_POLICY(

  3      protection_policy_name => 'ZDLRA_BRONZE'

  4      , description => 'Policy ZDLRA MAA BRONZE'

  5      , storage_location_name => 'DELTA'

  6      , recovery_window_goal => INTERVAL '30' DAY

  7      , max_retention_window => INTERVAL '60' DAY

  8      , recovery_window_sbt => INTERVAL '120' DAY

  9      , guaranteed_copy => 'NO'

 10      , allow_backup_deletion => 'YES'

 11  );

 12  END;

 13  /




PL/SQL procedure successfully completed.




SQL>

 

As you can see, the parameters are self-explained, you just define the name, description, and recovery/retention goals. But these retention windows are important and need some attention. To check inside of ZDLRA database, you can check the table RASYS.RA_PROTECTION_POLICY.

 

Retention Window

As you saw above, when you create the policy you have three parameters related to the retention window:
  • RECOVERY_WINDOW_GOAL: This parameter defines the time that ZDLRA will keep the backups (inside the appliance disks) for the database that is covered by this policy. In the example above, all the databases will stay for 30 days. Oldest, it is not guaranteed and can be deleted.
  • MAX_RETENTION_WINDOW: If ZDLRA has free space the backups can be retained until this retention window days inside of ZDLRA. In the example above, will be 60 days. If you do not specify, it will be “until having space”. If you specify, ZDLRA will delete everything after that period.
  • RECOVERY_WINDOW_SBT: It is the window that ZDLRA will retain for backups cloned to tape. In the example above, it will be 120 days and after that, the backup is not valid and will be expired.
The import here is understanding the small details. By the ZDLRA rules, it always tries to support (for each database) the point-in-time recovery from today until the recovery window goal. As the documentation says: “For each protected database, the Recovery Appliance attempts to ensure that the oldest backup on disk can support a point-in-time recovery to any time within the specified interval, counting backward from the current time.”.

 

One collateral effect for RECOVERY_WINDOW parameter is that it is global for policy (and not per database), and if you remember, when you enroll the database at ZDLRA, you need to define the “reserved_space” for it. And the detail it is that this value (reserved_space) needs to cover the recovery_window_goal. So, if your database changes a lot (or it is a big database) you need to constantly check the “Recovery Window Goal” and adjust the reserved space for your database. You can read some best practices here (page 15).
The MAX_RETENTION_WINDOW means the maximum time that your databases will be inside of ZDLRA. Imagine that the period between RECOVERY_WINDOW_GOAL and the MAX_RETENTION_WINDOW as a bonus, the backups are not guaranteed that will remains or will be there. If ZDLRA needs to delete some backups (because of lack of space), it will delete these backups between these dates first. And since the management is based in backupset, it can occur that one backupset for your database is deleted and you can’t use this point in time to restore the database.
One detail here it is that backups that pass MAX_RETENTION_WINDOW are forcibly deleted by ZDLRA. So, if you have a close date/time/day between RETENTION_WINDOW and MAX_RETENTION_WINDOW (like 10 for the first and 11 for the second), you can put a high pressure over ZDLRA because it will never stop of doing delete tasks for backups. In the ZDLRA best practices (or the old version) there is some vague indication of how to set it, but the idea is not to be aggressive with this value. By experience, I recommend that, at least, the MAX_RETENTION_WINDOW be 20% higher than RETENTION_WINDOW (and the same for reservered_space – be at least 20% higher than database size). But if you have big databases, this value needs to be high because the delete task will demand more time to finish and you can lead for a non-ending delete queue/load inside ZDLRA.
RECOVERY_WINDOW_SBT means that the period that backups will be available (for recovery purpose) in the cloned destination (tape or cloud). Since these backups are not inside of ZDLRA, it will not struggle for a lack of space.
More than one Policy

 

For ZDLRA probably you will have more than one policy to protect your databases since you probably will have databases (PROD, UAT, TST, DEV) with different requirements for recovery window. And even inside of same type (like PROD) it is possible to have different requirements (because of law regimentation as an example) and these lead/force you to create more than one policy.

 

Whatever the case, all databases will “fight each other” for disk space, and if you badly design your policies, or left the database in the wrong protection policy, you can have a system with high pressure for disk usage. ZDLRA always will accept new backups, if needed will delete the oldest backup (if you think this is adequate because probably the newest data is more important). But it is true too that ZDLRA will try to support the point-in-time recovery for all databases to reach what was defined in the policy. If you want to control this behavior you can set parameter GUARANTEED_COPY to YES. Doing this ZDLRA will delete old backups just if they were already copied to tape or replicated.
Don’t be afraid to create more than one policy because to move one database from one policy to another it is a simple command and will be more adequate to manage space usage if needed. One drawback of the huge number of protection policies is that clone to tape backups are based/scheduled in protection policies. If you have a lot of them, you need to create one clone to tape job for each one.

 

 

Protection Policies and ZDLRA Replication

One important detail is ZDLRA replication and how it interacted with protection policies. This is important because the replication between ZDLRA’s is purely based on policies, this means that replicates everything/all databases for the protection policy that you defined as a parameter. So, as you can imagine, if you want to replicate just some part of your databases between ZDLRA’s you need to create a specific protection policy.
Another interesting point is that on both sides of replicated ZDLRA the protection policies can have different recovery window goal. As example, in the primary site, the upstream ZDLRA can have 30 days of recovery windows and guaranteed copy as YES (because this ZDLRA receive more backups), but in the downstream ZDLRA, the destination protection policy can have 120 days as recovery window goal (because this ZDLRA protect fewer databases and the pressure for space usage will be less).
Let’s imagine protection policy for SILVER databases (https://www.oracle.com/a/tech/docs/maa-overview-onpremise-2019.pdf), that you want to replicate just some of them. In this case both ZDLRA’s (upstream and downstream) will have the “normal” silver protection policy (named as policy_silver as an example), as well another policy just for replicate some silver databases (named as policy_replicated_silver).

 

Architecture Design

 

The correct definition for your protection policies it is important for ZDLRA maintenance and usability. Design correctly the polices are important to avoid high pressure over the storage location for ZDLRA, even if you start to used (or maintain) one already deployed ZDLRA.
Understand recovery window goals and max retention windows constraint will avoid reaching full space utilization. You don’t need to create just one or two protection policies for your ZDLRA, but be careful with your design if you have replicated ZDLRA or protection a mix of database types. Group them correctly.
As explained before, there is a direct link between retention_window and reserved_space for your databases. If you create a unique protection policy for all of your databases, you can lead to putting a high value for reserved space and this can cause problems (like ZDLRA deny to add databases because you already reserved all the available space – even existing free space).
There is no rule of thumb to follow, like create policy A or B with X or Y values for the recovery window. The correct way is checking the requirements (and rules) that you need to follow and design the architecture that meets your requirements. Don’t worry if you need to change it in the future, it is possible and easy to do.
So, the most important is to know and understand the links that exist between the ZDLRA functionalities. Protection policies, replicated backups, and reserved space are some examples. A good time understanding them will reduce rework in the future.
 
 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”


ZDLRA, Virtual Private Catalog User – VPC
Category: Engineer System Author: Fernando Simon (Board Member) Date: 5 years ago Comments: 0

ZDLRA, Virtual Private Catalog User - VPC

The Virtual Private Catalog (VPC) user is a key piece for a good ZDLRA architecture design. The detail is not how to create it, but how to correctly integrate it in your design, and this is more important if you have replicated ZDLRA or using Real-Time redo transport.
Here I will show and discuss VPC implications for your architecture design when deploying ZDLRA. Even for a complete and new implementation (together with database) or adding ZDLRA at your already running environment. All points here try to show some perspectives and key points that can help you to correct use and define VPC’s.

 

VPC

For the simple definition, VPC user is created directly inside ZDLRA database using the “racli” (or EM/CC) as you can see below and here in the doc:

 

[root@zdlras1n1 ~]# /opt/oracle.RecoveryAppliance/bin/racli add vpc_user --user_name=vpcsrc

[vpcsrc] New Password:

Mon Nov 25 23:41:45 2019: Start: Add vpc user vpcsrc.

Mon Nov 25 23:41:46 2019:        Add vpc user vpcsrc successfully.

Mon Nov 25 23:41:46 2019: End:   Add vpc user vpcsrc.

[root@zdlras1n1 ~]#

 

The VPC is used when you connect using rman from the client, and it is the catalog owner for your rman:

 

[oracle@orcloel7 ~]$ rman target=/ catalog=vpcsrc/vpcsrc@zdlras1




Recovery Manager: Release 18.0.0.0.0 - Production on Tue Nov 26 00:23:40 2019

Version 18.6.0.0.0




Copyright (c) 1982, 2018, Oracle and/or its affiliates.  All rights reserved.




connected to target database: ORCL18 (DBID=3914023082)

connected to recovery catalog database

recovery catalog schema version 19.03.00.00. is newer than RMAN version




RMAN>

 

Inside ZDLRA it is just a logical organization for your connections, it does not store tables or views. Just made the link for rman catalog views (RC_*) between the VPC user itself and the internal RASYS (ZDLRA user/schema that have all tables).
When you add the database inside of ZDLRA you grant the access for the database using VPC user as a parameter too:

 

SQL> BEGIN

  2  DBMS_RA.GRANT_DB_ACCESS (

  3        db_unique_name => 'ORCL18'

  4        , username => 'VPCSRC'

  5  );

  6  END;

  7  /




PL/SQL procedure successfully completed.




SQL>

 

But to be honest this is just a simple definition. VPC is more important in the architecture of your solution than “creation and connection commands”. I will explain.
For ZDLRA, the VPC is just a logical definition (don’t store data itself), you can create more than you single VPC. You can create multiple VPC’s and separate databases based on your own rules. Like one VPC for DEV database, one for PROD, or can be one per rack (like Exadata, AIX, Power). You choose.
Remember that the VPC is the rman catalog user, so every database with the grant, “share” it. If you do “list db_unique_name all” (and if do “set dbid”) everyone in the same catalog can be access the backups.
So, when you are planning your architecture for ZDLRA deployment, think about the VPC users that you will have. How you can divide/aggregate your databases. But don’t worry if you need to create more VPC’s in the future, remember that it is just a logical definition. If you want to move databases between VPC’s you just need to call DBMS_RA.GRANT_DB_ACCESS for your database to the new VPC. All the backups will appear automatically with/in the new catalog.
But it is not just a rman catalog that VPC defines. There are other points that can change your architecture, I will explain below.

 

VPC and Real-Time Redo transport

The Real-Time Redo is the feature for ZDLRA that guarantee the zero RPO, you can see more details in my previous posts (here and here). But the point is that the VPC needs to be used as REDO_TRANSPORT_USER parameter for your database to real-time redo work.
This is needed because the ZDLRA is one destination for archive logs (log_archive_dest parameter) and since ZDLRA “simulate” dataguard in this case, both sides need to have the same user (as a normal DG does). Usually, you don’t change this for DG since both sides talks using sys user. But for ZDLRA, you need to use VPC as redo user. And besides that, the automatic backups for archivelog need to appear inside rman catalog (list backup of archive as an example). As told in my post here, create user inside your database (with sysoper permission at least) and change the database parameter.
So, unfortunately, this can impact again at the architecture design and not just for ZDLRA, but for all databases. Sometimes the database username needs to follow some naming/security rules and VPC definition will need to follow this too.

 

VPC and Real-Time Redo transport and DataGuard

The point for VPC and Real-Time Redo when using Dataguard is that in this scenario you have at least two sites, two ZDLRA’s, two databases with dataguard, and real-time redo transport enabled. But, everything needs to have/use the same, and only one, VPC username.
Database in site 1 will talk with ZDLRA at site 1 using the VPC for backup and real-time redo. And since it is dataguard, the database in site 1 talks with the database in site 2 (for DG) using the redo_transport_user. So, if you are doing a backup of the database in site 2 at ZDLRA in site 2, and even the real-time redo, the VPC for ZDLRA in site 2 needs to be the same than redo_transport_user. As you can imagine, the same VPC username in both ZDLRA’s in this case. If you are not planning to use real-time redo for your database, you can use different VPC usernames.
So, if you are planning to use multiple ZDLRA’s to protect your sites the architecture design needs to be well planned. If you were planning to use different VPC usernames (to represent different rman catalogs for each site), you need to review. You can use this approach for databases that don’t use dataguard, but will not fit for DG databases.

 

VPC and Replicated ZDLRA

Another point that needs attention for VPC is when using replicated ZDLRA. Besides, if you are using the real-time redo, here, you need two additional VPC’s (three at total):
  1. ZDLRA site 1 VPC: The normal VPC used to database (from the same site) connect and do backups and use real-time redo.
  2. ZDLRA site 2 VPC: VPC at second that will be used as is normal rman catalog. This is needed if you want to connect the database from site 1 in ZDLRA site 2 (think that your ZDLRA from site 1 is out).
  3. ZDLRA side 2 VPC for replication: VPC user to receive the replicated backups from ZDLRA in the other site. This is used internally by ZDLRA to intercommunication but still is a VPC.
When using replicated ZDLRA you have one database doing backup at a local ZDLRA in the same site, and this ZDLRA sends this backup to other replicated ZDLRA. The connection between the database and local ZDLRA uses the VPC from point 1 listed above. The replication between ZDLRA’s will use the VPC from point number 3. And if you want, you can have the VPC of point #2 to do backups when the “source” ZDLRA is down.
This scenario can complicate a little more if you are planning to use real-time redo because the VPC from points 1 and 2 needs to be the same.
As you can see, other points that you need to verify and take care of ZDLRA architecture design.

 

Architecture

 

In this level of architecture design probably you will have more than one ZDLRA, protecting more than one site, or doing replication to achieve ideal protection. But even if you use only one ZDLRA, the concerns are the same.
Independently the number of ZDLRA’s is possible that you will mix all the available features, real-time redo, and replication. It is common that you have different requirements for your databases, you may need to handle (and protect) single/isolated databases (like DEV), databases with just real-time redo in use, databases with ZDLRA replication (for multi-site protection), and probably databases with DG that needed to be protected in both sides.
Usually, when deploying ZDLRA you are dealing with huge and mixed environments that need different levels of protection. But before “just deploy” ZDLRA there you need to start with good architecture design, identify the needs and how you care VPC to handle correctly everything.
Maybe you will need to create more than one VPC, this totally depends on your requirements. It does not exist a magic receipt to do that, but if you know all the details that can impact VPC definition, the success will be achieved. As you saw, VPC it is not just the user for rman connection.
The idea of this post was pointing most of the details that are important to know about VPC when designing the architecture to deploy and integrate ZDLRA.
 

 

Disclaimer: “The postings on this site are my own and don’t necessarily represent my actual employer positions, strategies or opinions. The information here was edited to be useful for general purpose, specific data and identifications were removed to allow reach the generic audience and to be useful for the community.”


1 2 3 4 5 7