iwat command trigger not working

Is there anything in the event log after you run the lock ?

Update...

We've discovered that sometimes when we use Import to pull files into TS, files are locked but the locking doesn't show up in the event log. But other times locking does show up in the event log. WTF?

is this specific to a branch or workarea ? Or is it one time it happens, one time it does not to the same file ?

Looks like there is a lag time between when an event occurs and when it is being logged in the event log - as in minutes. That appears to be the issue. I love Interwoven.

What DB are you using for the event subsystem ?

We think it shares the same instance with the OD reporting DB, which is Oracle.

You should be able to tell us which DB by looking in:

iw-home/eventsubsystem/conf/jmsconfignew.xml

[PHP]

[/PHP]

It's definitely Oracle. The DBA's found a table that wasn't using an index and was killing the performance of the DB, so after they put an index on the table, the performance is a lot better and we're getting quicker updates of the log. Thanks!

Can you let us know which column wasn't indexed. We've just turned on the Event Subsystem using Oracle and may be having performance issues too...

Thanks!

It's definitely Oracle. The DBA's found a table that wasn't using an index and was killing the performance of the DB, so after they put an index on the table, the performance is a lot better and we're getting quicker updates of the log. Thanks!

Hmm, but how can you blame this on TeamSite ? You might have to conclude that TeamSite is not the cause of all the evil in this world. Smiley Very Happy

ISCBorisB

Hmm, but how can you blame this on TeamSite ? You might have to conclude that TeamSite is not the cause of all the evil in this world.

Well... IWOV has not invented distributed processing based on the event subscription. CORBA, Oracle, Forte and whatnot all do that.

IWOV however is the only System I know of that insists on keeping Event's history persisting, be it a (buggy) File, RDB or else.
In the processes like that, one hour old event is an ancient history, one day old event happened when dinosauruses were alive.

If some subscribers (IWOV processes!) not listening, events stay in the corresponding DB and Tables grow indefinitely.
The end result is a monumental resource waste and Users/Administrators/DBAs headache. Sure, if you have millions of event
entries (I've worked with systems where depending on load a rate easily went over 200K events a day) you better index Tables.
I wonder what IWOV need this garbage for.

Do all of those events in the messages table get left there forever?!?

ISCBorisB

Do all of those events in the messages table get left there forever?!?

Normally, no. Processed Event is removed. Problem is, only IWOV knows who subscribers (IWOV processes) are and if some of them do not listen for whatever reason, Tables start to grow really (and I mean really!) fast