Video Screencast Help

Journal mails not being archived

Created: 14 Jun 2011 • Updated: 04 Apr 2012 | 3 comments
mart g's picture
This issue has been solved. See solution.

I have 3 journal mailboxes. mailbox A, B & C

Journal maibox A and B is using the same EV system mailbox on same Exchange server. Exchange server A

Journal mailbox C is on another Exchange server. Exchange server B.

Journal mailbox A & C have no issue but mailbox B was not being archived. 

This happened after migration of the EV to another physical hardware of the same model (HP DL380 G5) but different CPU speed [3ghz(old) and 2.5ghz(new)]. The old EV server was running Windows 2003 SP2 and SQL 2000 SP4 and the new EV server is running Windows 2008 SP2 x64 and SQL 2008 SP2 x64. There was no error in the eventlog and I couldn't see any problem in dtrace. Everything seems to work fine.

When I tried to open the mailbox B using outlook 2003 on the EV server (EV 8 SP5), it takes a long time and most time will hang up.

Now there are about 70k mails not archived in the Journal mailbox B.

Could this be due to MAPI issue ? 

Or there is too many emails to archive but not enough memory? Mailbox B got alot more mails than the other 2 mailboxes.
The EV server got 4gb of ram, but normally 2 or 3gb or ram will be used up.

 

 

 

Comments 3 CommentsJump to latest comment

Nick White's picture

As there don't seem to be any errors I think a Dtrace is going to be the best way to work out what is going on, and it's probably best to stop any other Journal tasks when you are doing it to cut down on the noise in the file. If you are not familiar with Dtrace you can enable it in the Admin Console by selecting Tools > Advanced features, refreshing the tree and then you will see a Traces container under the EV server in the servers node.

AndrewB's picture

your question about "not enough memory" reminded me to ask if you've increased the setting for the MSMQ memory? the default is only 1gb which may not be enough to handle that huge backlog, however, you said you're not getting any errors in the event log. a dtrace should show us more.

Andy Becker | Authorized Symantec Consultant | Trace3 | Symantec National Partner | www.trace3.com

KarlW's picture

When you migrated did you upgrade EV version at all?

Additionally it could be due to Exchange server architecture.  Is Mailbox B in the same database as Mailbox A?  It might be worth checking out CPU usage of Store.exe on the Exchange server.

When EV accesses a mailbox Exchange may need to build a new index for that folder based on the columns EV is requesting.  If there are a large number of messages in the inbox this can take some time.

This happens the first time EV processes the folder but can also happen after an EV upgrade if we've requested more information from Exchange.  

If this is the case Dtrace won't be of much help as the Journal task is waiting for Exchange to complete and return the data.  Also it can just take Exchange a long time to return the data - which is what you are seeing with Outlook hanging.  This would also impact the journal task.

Regards

Karl

SOLUTION