Archive task is failing
Created: 15 Aug 2012 | Updated: 24 Sep 2012 | 22 comments
This issue has been solved. See solution.
EV 10.0.1.1171. Frequently I find that the archive task associated with one of my Exchange 2010 mailbox servers is failed. I see nothing in the event logs that mentions the failure. Restarting the task is always successful. Is there a powershell script that I can schedule that will alert me that the task is failed?
Discussion Filed Under:
Comments 22 Comments • Jump to latest comment
well you're best off finding out why it failed
Typically with Exchange 2010 its because the CAS Server is forcefully disconnecting the EVAdmin account from its session, usually remedied by making sure the throttling script was correctly applied.
You could see on the CAS Servers event logs where it mentions that the EVAdmin has breached one of its policies (for instance TimeSpentInCAS or whatever its called)
Another can be things like CAS Loadbalancers that distribute calls between CAS Servers so that no one exchange server is being overloaded, sometimes that causes issues for EV and other mapi services such as Blackberry Enterprise Server.
You could specify a specific CAS server using the DS Server registry key or by using the ExchangeGCOverride column within the ExchangeServerEntry table in your EnterpriseVaultDirectory database.
I would have thought that there would be an event in the event log indicating why it was failed...
Many Thanks,
Rob
www.quadrotech-it.com - All your EV Tools
PS I hope that the post proves helpful.
I had a similar problem following a migration to Exchange 2010 from Exchange 2003. My issue was caused by a failing GC. I removed the option to force EV to use the "Closest GC" and also removed the registry setting pointing it at the closest GC and all of a sudden the Archive Task stopped failing.
This may not resolve your issue, but it is worth checking.
Make sure the you have all the required hotfixes isntalled for oulook on the Ev server.There are some hotfixes which are mandatory to be isntalled on the Ev server for oulook 2007 sp2 .
If this response answers your concern, please mark it as a "solution"
Is there a list of hotfixes?
You can find the hotfixes details on the Compatiblity list/chart - For Outlook 2007 SP2 on EV server section. There should be two hotfixes that you need.
I'm using SP3. Doesn't appear to be any hotfixes.
EV is pointing to a single member of the CAS using a HOSTS file. I see these error in the EV logs today:
Source: Enterprise Vault
Date: 8/16/2012 5:48:30 AM
Event ID: 3432
Task Category: Archive Task
Level: Error
Keywords: Classic
User: N/A
Computer: EVAULT.domain.com
Description:
One or more errors occurred during the creation of a profile to connect to an Exchange Server.
Targeted Exchange Server: MBX02
Mailbox: SMTP:EV_MBX02@domain.com
ConfigureMsgService failed with the following errors:
Out of memory error has occurred for the following connection points:
DC01.domain.com
MBX02
This indicates that the maximum number of connections has been consumed on this Enterprise Vault Server.
Source: Enterprise Vault
Date: 8/16/2012 5:48:30 AM
Event ID: 2216
Task Category: Archive Task
Level: Information
Keywords: Classic
User: N/A
Computer: EVAULT.domain.com
Description:
Message dispenser will suspend processing for 10 minutes due to a recoverable error
Task: Exchange Mailbox Archiving Task for MBX02
Source: Enterprise Vault
Date: 8/16/2012 5:48:46 AM
Event ID: 3305
Task Category: Archive Task
Level: Error
Keywords: Classic
User: N/A
Computer: EVAULT.domain.com
Description:
The Task 'Exchange Mailbox Archiving Task for MBX02' failed to log on to Exchange server 'MBX02' using mailbox 'SMTP:EV_MBX02@domain.com'. Please ensure the mailbox has not been hidden, that the server is running and that the Vault account has sufficient permissions on the server.
Source: Enterprise Vault
Date: 8/16/2012 5:48:46 AM
Event ID: 2256
Task Category: Archive Task
Level: Error
Keywords: Classic
User: N/A
Computer: EVAULT.domain.com
Description:
The Archiving Task 'Exchange Mailbox Archiving Task for MBX02' could not be started due to startup errors.
I don't see any errors in the CAS server logs.
1. You need to determine if these errors are occuring at these particular times or do your archiving task randomly stop.
2. At these particular times, anything happening on the Exchange servers (CAS and MBX servers) and/or your GC servers? i.e. maintenance or/and server restart etc.
3. As for the EV server, any reason why you are using HOSTS file?
Out of memory error has occurred for the following connection points:
DC01.domain.com
MBX02
If you go to Control Panel and open Mail 32, click on Profiles, do you see a lot of MAPI profiles created by EV? Try clear these out.
The other suggestions above would help your issue too.
oh...have you run the SetEVThrottling policy for the EV service account yet? I would have thought you'd have done this but just checking.
1. That's why I was looking for a way to alert on the task failure. I don't know when it's failing, I can only look for logs after I see it has failed.
2. no maintenance
3. We had intermittent communication problems to our CAS array address so we force it to use one. Same issue with BES and GOOD. It seems that some things can't deal with Windows NLB.
Throttling policy is set. Is there any down side to deleting all of the extra profiles?
1. Do you have SCOM or some sort of monitoring for EV. EV does not have these features that you are looking for...This will be the key point to your problem.
2. That's fine.
3. ah...If I remember correctly, Exchange 2010 Windows load balancing may not even be supported or it doesn't work well and you have to use hardware load balancing. Check with Exchange 2010 expert/consultant/architech.
Clearing out those windows profile will do you good, they will get recreated when EV task starts to run as they are needed to access the user's mailbox. Hence they need to be cleaned out every now and then.
As for the Outlook 20037 SP3, did you install this from the start of EV settup of was it upgraded from Outlook 2007 SP2? Not that this make a lot of difference to your issue but I am interested as I have seen people upgraded and it caused MAPI issues.
Take a look at this regarding Exchange 2010 Windows NBL.
http://www.msexchange.org/articles_tutorials/exchange-server-2010/high-availability-recovery/load-balancing-exchange-2010-client-access-servers-using-hardware-load-balancer-solution-part1.html
Not sure how your environment is setup and configured, basically it is not recommended or at least not suitable. HNBL is the Microsoft recommendation.
1. no
3. It's supported. And since we're pointing to just one server it would seem to me that NLB isn't part of the equation.
I cleared the profiles. Outlook was installed from the SP1 media. Upgraded to SP3 via Windows Updates. Don't remember if it stopped at SP2 first.
I would suggest setting the CAS in the ExchangeGCOverride as opposed to the hosts file
Aren't the ExchangeGCOverride and DS Server entries used to specify the Global Catalog, not the CAS Server?
So what's the need for Windows NLB if you are only using one CAS server? Just curious...What happens if you remove the Windows NLB to take it out of the issue equation completely as you seem to think that this may be causing this issue and relating to Good and BES as well?
On your exchange environment I believe you can setup monitoring relating client MAPI connections and connectivities etc. You may have already done so....see from this monitoring results if there is a high usage at a particular time of the day and if they matches the time when the EV tasks are failing.
Also reducing the EV archiving task settings to just one connection per concurrent connection for each task...but this will slow down the round robin of archiving if you have lots of mailboxes per exchange server.
I understand the Windows NLB is normally setup to scale/spanning the client connection via TCP/IP, just wondering if it might be worth you add another CAS server and use HNLB due to lots of client connection with BES and Good Technology and god knows what else? of course tis will add cost into this equation.
I am not saying that this will definitely resolve your issue, just my thoughts and opinions from what I can see from a consultancy point of view. Also this will help planning your company growth, more users more client's connections.
NLB works fine for general client access via Outlook 2007 and 2010. No problems reported. We only see issues when other servers are trying to access.
I'm going to focus on the archive schedule to see if I can find anything.
Let us know what you can find as there has been a few queries on this forum recently regarding EV archiving tasks failed and it's a bit of a mystery after all the normal troubleshooting for EV is done for EV9 and EV10. So it may be an outlook on the EV server issue or/and CAS server issue etc. and we need to report back to Symantec or/and Microsoft to get it resolved permanently.
Symantec support provided this attempted fix:
So far it is working, but I'll have to monitor it for a few days to see if it is the solution.
Would you like to reply?
Login or Register to post your comment.