Saturday, February 3, 2018

[Domino 9.0.1 FP9] Occasional Domino Task Hang on non-transaction logged server

We have upgraded our mail server (non-transaction logged) few month ago from Domino 9.0.1 FP8 to FP9 IF1 and since then we experienced occasional Domino task hang few times in a month.

Domino task such as Indexer and Replicator will hang on a certain mailbox and the mailbox is no longer accessible. Everything looks fine again after we restarted Domino. And when this appears again the next time, it will be a different mailbox.

We managed to run manual NSD when this happened and requested IBM Support to check on this.

It appears that this is a known issue in  Domino 9.0.1 FP9 which happens on non transaction logged server and with warning quota set to DBs (mostly on mail server, as we set quota to mailboxes).

SPR #KBRNASPR6L - Domino server hangs in UBMDelayThreadDebug on non Transaction logging enabled servers.

Solution (Either one the following will prevent this):

1. Submit a PMR to IBM Support to obtain a custom hotfix. (this fix is not available in any of the public released interim fix for Domino 9.0.1 FP9.)

2. Enable transaction logging.

3. Upgrade to Domino 9.0.1 FP10 which includes this fix. (Not recommended for the moment, suggest to wait for Domino 9.0.1 FP10 IF1 which fixes problems in FP10)

4 comments:

  1. FP10 has some issues better wait FP10 IF1
    https://www.ibm.com/developerworks/community/blogs/LotusSupport/entry/Listening_to_your_feedback_on_Notes_Domino_9_0_1_FP10

    ReplyDelete
    Replies
    1. Thanks for the warning, added it to my solution above.

      Delete
  2. Hi, it seems, we had the same troubles and opened a PMR and received a HotFix 260 for FP9. Try to open a PMR and reference PMR 32845,999,618.

    Best regards, Rainer

    ReplyDelete
    Replies
    1. Is comforting to know that I'm not alone facing this problem. This had been troubling me many months as I'm not able to obtain a manual NSD and there's no technote nor interim fix that provides this fix.

      Delete

Beware not to abort compact -REPLICA -RESTART

Due to some of our DBs encountered "Unable to extend an ID table" error, I have scheduled compact -REPLICA -RESTART to run on thes...