Critical issue in SharePlex versions 9.1.x might cause data loss after activation.
After activating a configuration on source, reader on source and/or import on target might go into a recovery and indicate that some messages are skipped, causing data loss.
Import Recovery message example :
Notice 2018-10-10 19:45:27.941353 16122188 1 Import: Queue write recovery started, qrw_srcseq=10098143231339 msg.mpseq=10098143223896 (importing from source MainQueue) [module que]
Notice 2018-10-10 19:45:27.960541 16122188 1 Import: Queue write recovery complete, 8 duplicate messages skipped , 7446 bytes total (importing from source queue MainQueue) [module que]
From the above example, 8 messages were lost because of the issue.
There are no workarounds available for this issue, please upgrade to 9.1.4, before you activate your configuration again.
You can download the SharePlex 9.1.4 version from our Product Downloads page.
Who does this effect :
If you are replicating from Oracle, using SharePlex versions 9.1.0, 9.1.1, 9.1.2, or 9.1.3 and if you activating a config, this critical issue might cause data loss.
If you are replicating from Oracle, using SharePlex versions 9.1.0, 9.1.1, 9.1.2, or 9.1.3, but did not activate or plan to activate, this critical issue will not affect you.
If you are using SharePlex versions 9.0.x or 8.6.x, this issue will not affect you.
This issue is database version and platform agnostic.
Additional Conditions:
If there is no user activity during activation, there wont be any data loss.
and this issue only effect setups with object cache bigger than 1MB in size(Check $SP_SYS_VARDIR\state\*objcache*).
Refer to the SharePlex 9.1.4 Upgrade Guide for Oracle Source for detailed instructions on upgrade.
© 2025 Quest Software Inc. ALL RIGHTS RESERVED. Terms of Use Privacy Cookie Preference Center