NetWorker: Cloned Saveset Incorrectly Marked as Valid Despite Source Corruption

• A saveset that was valid during backup becomes corrupted later.
• Cloning the corrupted saveset completes with errors but still marks the cloned saveset as valid in the media database.
• Affected NetWorker versions are 19.6.x, 19.7.x, and 9.8.x, and versions prior to 19.9.0.4.
• nsrclone reports failure, but mminfo shows the cloned saveset with valid flags.

Example:

root@server:#  mminfo -q ssid=2557547202 -r "clonetime(22),sscreate(22),volume,pool,totalsize(2),sumflags,clflags,name"
      clone time            ss created      volume         pool        total fl clflg name
11/13/2022 04:52:34 PM 11/13/2022 04:52:34 PM VM040089     DUMPS_5WKS 31 GB cb  RMAN:HOT_F_DB_name_1.rman  <=== Backup valid but corrupted data

• Starting a nsrclone of the corrupted saveset

root@server:# nsrclone -D9 -v -s nsr_server -b CLONE_5WKS -S 2557547202 >> clone_log.txt 2>&1 ; echo $?
1   <=== An error is correctly reported from the nsrclone command

• The NetWorker clone output log is reporting the saveset as invalid

root@server:/$ tail -50 clone_log.txt
      forced volume location: SN_name;
          inactivity timeout: 180;
                      job id: 1104237;
                      manual: Yes;
               NSR operation: cloning;
                   save sets: \
...
11/22/22 17:29:08.948529 start_regular_clone for Regular type failed: Unable to start clone session: other clones exist for failed save set RMAN:HOT_F_DB_name_1.rman ssid=2557547202
 11/22/22 17:29:08.948586 Exiting process_this_snode for Regular clone
11/22/22 17:29:08.948603 Function for Regular clone returned: Unable to start clone session: other clones exist for failed save set RMAN:HOT_F_DB_name_1.rman ssid=2557547202
 162560:nsrclone: Unable to start clone session: other clones exist for failed save set RMAN:HOT_F_DB_name_1.rman ssid=2557547202
...
198620:nsrclone: Some of the clones failed
198620:nsrclone: Some of the clones failed
11/22/22 17:29:08.953903 alldone(): ENTER
...
         completion severity: 50;
           completion status: failed;
...
11/22/22 17:29:08.954582 nsrclone, alldone(): EXIT

• Daemon.log from the NetWorker server reports an error:

181147 11/22/2022 05:29:08 PM  ansrd NSR critical Unable to clone save sets for nsr_prd: other clones exist for failed save set RMAN:HOT_F_DB_name_1.rman ssid=2557547202
71193 11/22/2022 05:29:08 PM  nsrd NSR info Error Alert: Clone operation failed with error: 'Some of the clones failed'.

• The Daemon.raw from the Storage Node reports an error:

198625 22/11/2022 5:29:04 PM  2 0 11 2180355872 29752 0 SN_name nsrmmd#43 NSR warning Reading volume 'VM040089' failed: 'Bad file descriptor' ^M
198625 22/11/2022 5:29:08 PM  2 0 11 2180355872 29752 0 SN_name nsrmmd#43 NSR warning Reading volume 'VM040089' failed: 'Bad file descriptor' ^M

• Once the clone is completed, you can see the saveset as follows:

root@server:/# mminfo -q ssid=2557547202 -r "clonetime(22),sscreate(22),volume,pool,totalsize(2),sumflags,clflags,name"
      clone time            ss created      volume         pool        total fl clflg name
11/22/2022 05:13:30 PM 11/13/2022 04:52:34 PM VC00500U     CLONE_DUMPS_5WKS 31 GB hb  RMAN:HOT_F_DB_name_1.rman  <= Clone is considered as VALID
11/22/2022 05:13:30 PM 11/13/2022 04:52:34 PM VC0250BO     CLONE_DUMPS_5WKS 31 GB tb  RMAN:HOT_F_DB_name_1.rman     <= Clone is considered as VALID
11/13/2022 04:52:34 PM 11/13/2022 04:52:34 PM VM040089     DUMPS_CIBOS_5WKS 31 GB cb s RMAN:HOT_F_DB_name_1.rman     <= saveset source has been newly set as suspect

NetWorker incorrectly marks cloned saveset as valid even when the source saveset is corrupted and the clone operation fails. This behavior is due to a defect in the clone validation logic.

A fix for this issue is in the following NetWorker versions:

• 19.9.0.5
• 19.10.0.2
• All versions above 19.11.x

Upgrade to one of the above versions to ensure that corrupted savesets are correctly marked as invalid during cloning.