3310 error mails are pending - english version

[Mailmonster]

We are currently evaluating Enterprise Vault 7.5 (running against exchange 2003 sp2) 

The EV Server runs using windows 2003, sp1 with 4 gb of ram and has already been reinstalled (blank new virtual server) several times

Exchange  Server and EV server are virtual servers under vmware esx

Database is sql2005 express on the ev server itself

Problem:

The well known "3310 unable to access a network..." is appearing very often and blocking all archiving actions (manually as well as per schedule)

Mails are stuck in different private message queues (a3 run now archive and a6 sync requests)

Restart of either ev taks controller service or message queue oder server reboot helps as a short time workaround. 

When the mails are stuck in the queue outlook mails are stuck in "pending state"

We reinstalled the message queue, we updated the ev server, we changed the msmq hard drive path, we changed kernel memory registry settings, we installed .net 2.0 sp1.

what else. Where in the world can this problem has its root cause?

Thanks for any help

best regards

christian

[PhilCo_2]

these 3310 events are often caused by mailboxes that have been moved to a mail server that the Enterprise Vault does not have an Archiving Task for, perhaps the mail server could be in a different site. The user who owns the mailbox will still have the rights to manually archive his mail, but, when the EV server comes to process the pending items and replace them by shortcuts, it has no way to contact the server. If you see the MSMQ items stuck in the A2 queue, this is most often the case, as far as I have seen anyway.

You will most likely see the 3310 events every ten minutes by default, as the  dispenser will 'sleep for ten minutes', then it will start again and the event will occur again, for every email that has been archvied from the mailbox in question.

If you expand the A2 queue, you can identify which mailbox is causing the problem and disable the user from writing to the archive.

There are third party tools available which will allow you to delete only the troublesome messages, rather than purging the whole queue, for example 'Queue Explorer, by Cogin' is very simple and effective.

You can also change the amount of time that the EV server will 'sleep' by reducing the

'ProblemPauseInterval'   in the 'archiveretrievaltask' sql table, although its up to you if you want to try that. I have set ours to one minute, which as far as I know is the minimum time allowed.

[Mailmonster]

Hello

thanks for your feedback. Unfortunately we only have one Mailbox Exchange Server being the target for

Enterprise Vault and always having been the target.

Our environment has only one connector and one exchange mailbox server. Since writing my forum post this morning the situation is still the same. The mails are "pending archive" and The Queues are not emptied

However they are in the A3 Queue where one message is generated for each mail that is subject to manual archiving. The other queue not emptied is he a6 queue where the synchronization requests messages are put. However the SQL Entry is  a very good hint. I guess I will also put 1 minute There.

However currently I have not 3310 Error in the log since this night but the queues are not processed.

There is at least what I know no possibility to run a dtrace on stuck messages.

Is there a way to increase logging for message queue Issues on the microsoft side?

best regards

Christian

[Mailmonster]

Hi Philco - does changing this setting need a reboot ?

I changed it to one minute but the waiting time is still more like 10 minutes.

best regards

christian

[patrickkuah]

well...you need to at least restart the archiving task to take effect.

Is 3310 well known??? i have never seen such issue before. I would suggest you to log a call to the support and get them to take a look....

regards,

patrickkuah

[patrickkuah]

ha....yes, 3310 is well known :) my apologies...

Check the mapi connectivity, check the archival task and use smtp format for the system mailbox...

patrickkuah

[Mailmonster]

Thanks to everybody for all your answers so far.

We did:

- Apply all Service Packs to the EV machine (including  .net 2.0 sp1 and language pack)

- put in the smtp address of the system mailbox

- applied all kernel memory settings

- used fixmapi

- moved message queues away from system drive

- increased memory from 2gb to 4gb (with /3gb switch in boot ini)
- reinstalled message queues

- reinstalled EV server

- put in registy key to disable strict name checking

- checked that the patch level of the VIRTUAL Exchange 2003 server is actual (SP2)

However the status remains the same.

1. At some unpredictable point in time the messages in the message queues are no longer processed and the queues can only be "revived" by either restarting the enterprise vault task controller service or the message queue service (which again also restarts ev task controller and ev storage service)

2. I get 3310 Errors from time to time and archiving resumes only after the sleeping period which I could thanks to this forum reduce to 1 minute which helps a lot although it is "only" a workaround and not the cure of the root cause

Summary:

I think that there is no relationship between both issues but I am not sure.

However I could live with partially getting 3310 errors with the reduced waiting time of 1 minute (even if I one time got the error 3 times in a row)

However I need to fix the message queue issue because I do not want to all the time check the server during day and night and restart the services when the message queue stops being processed.

Again our settings.

- Virtual exchange 2003, sp2

- virtual Windows 2003 standard with sql express and EV 7.5 on the same machine.

- virtual Domain Controllers/Global Catalogs

 The whole virtualization is realized by vmware infrastructure 3 (ESX)

Thanks for any advice in advance

And one side question - is there anybody else out there using ev 7.5 successfully with sql express 2005 (on the same machine) or is everbody using a full blown MS SQL Server (which version)

best regards

Christian

[patrickkuah]

1) remove /3GB

2) If you have antivirus, exlude the MSMQ location

3) Check your Global Catalog. If possible, do a dtrace on the archival task check for possible global catalog issue.

regards,

patrickkuah

[Mailmonster]

Hello back

- removed /3gb

- currently no antivirus setup

- according to dtrace no gc issue but will try to set a fix one as a test anyway

best regards

christian

[MichelZ]

Are you still experiencing this?

Cheers

Michel