U2 Data Replication Issue

Shubham,

Have you tried reviewing the error log files to see why the all the replication writer (rw) process have died.

On UniData, check the bin directory of directory where UniData is installed and examine the contents of the rm.log, rm.errlog and rw.errlog files.

The version of the files in bin directory contain the current set of error log files, if you have restated UniData please also check the historical logs in the saved_logs directory under bin for the historical errors.

On UniVerse the files are called uvrm.log, uvrm.errlog,uvrw.errlog and are located in the UniVerse insall directoty the saved_logs directory is under that.

The detail in these logs, should allow you to work out why the repwritter died and correct the underlying problem.

Regards,

Hi Jonathan,

Thanks for reverting back.

Yes, we have reviewed uvrw.errlog and uvrm.errlog on publisher as well as deferred server but can't see anything which suggests the reason for the issue.

Publisher Error logs

Thu Sep 11 15:19:29 Replication Timeout(2119)!
Thu Sep 11 15:19:29 2025
 Group XXX Replication to uat-02->uat-04 is suspended. 
Thu Sep 11 15:19:56 2025,  report:
uvsh(1345246)--TP ERROR missing transaction log.

Deferred Error logs:

Thu Sep 11 15:19:28 2025 Subscribing Group XXX started suspending. Thu Sep 11 15:19:29 Replication Timeout(2119)!

Thu Sep 11 15:41:10 2025, Group 0 report:
Udsub[pid=1500597,RepGroupNo=0,Status=27,errorCode=4260,datetime=Thu Sep 11 15:41:10 2025
]:U_rep_dump_lirf(): a gap. numlogs=4499, startLSN=3350759, LRFhd->lastInfofd.nextLSN=3196187, ret=154572.

Thu Sep 11 15:41:10 2025 Subscribing Group ABC_GRP started suspending. Thu Sep 11 15:41:10 Replication Timeout(2119)!
Thu Sep 11 15:41:11 2025 Subscribing Group DEF_GRP started suspending. Thu Sep 11 15:41:11 Subscriber has no alive RWs(2358)!
Thu Sep 11 15:44:30 RM request waiting timeout(2551)!
Cross-Group Suspension Suspended GROUP DEF_GRP, DISTRIBUTION uat-02->uat-04 failed.

Shubham,

Have you logged a support call on this for Rocket to look at ?

Replication problems if not immediately obvious, require detailed investigation as this should be done through a support call and not through a public forum such as this.

From the very limited logs you've provided, then I've already noticed replication suspended because of a timeout and it appears you have no exception action script in place to recover the suspension.

You still do not appear to have shown any of the errors from the subscriber uvrw.errlog, which would detail why the repwritter processes have all disappeared.

You are also using a deferred replication, which adds to the complications.

Please raise a support call with us in order for the problem to be correctly investigated.

Thanks,