none
users suddenly started getting when they try to send out a email to internal or external users "This message could not be sent. Try sending the message again later, or contact your network administrator. Error is [0x80004005-00000000-00000000]"

    Question

  • users suddenly started getting following error " when they try to send out a email to internal or external users

    This message could not be sent. Try sending the message again later, or contact your network administrator.  Error is [0x80004005-00000000-00000000]. "


    checked the eventviewer on exchange ..found tons of following events..

    event ID 1022
    "Logon Failure on database "First Storage Group\Mailbox Store (XYZ-Server)" - Windows 2000 account XYZ\testuser; mailbox /O=XYZ/OU=ABC/cn=Recipients/cn=cstier3.
    .. 

    Event ID 1025
    "An error occurred on database "First Storage Group\Mailbox Store (XYZ-Server)".

    Function name or description of problem: EcGenerateNDR:

    Error: 0xfffff9bf

    Error: -1069"


    Event ID-623
    nformation Store (5876) First Storage Group: The version store for this instance (0) has reached its maximum size of 155Mb. It is likely that a long-running transaction is preventing cleanup of the version store and causing it to build up in size. Updates will be rejected until the long-running transaction has been completely committed or rolled back.

      EVENT ID-9791
    Possible long-running transaction:

    SessionId: 0x7FD38220

    Session-context: 0x00000000

    Session-context ThreadId: 0x00000734

    Cleanup: 1

    For more information, click

    Cleanup of the DeliveredTo table for database 'First Storage Group\Mailbox Store (XYZ-Server)' was pre-empted because the database engine's version store was growing too large. 0 entries were purged.

     

    I rebooted the box and it fixed the issue…and couldn’t find anyting substantial cause of the issue..  I have exchagne sp2 enterprise running on windows 2003 sp1

     

    Any help would be appreciated..

    -D

     


    • Edited by mod 13 Wednesday, August 12, 2009 12:19 AM
    Wednesday, August 12, 2009 12:16 AM

Answers

All replies

    • Marked as answer by Allen Song Friday, August 21, 2009 8:56 AM
    Wednesday, August 12, 2009 6:40 AM
  • Please scan the exchange server using Best Practises Analyser
    Vinod |CCNA|MCSE 2003 +Messaging|MCTS|ITIL V3|
    Wednesday, August 12, 2009 12:55 PM
  • Hi,

    Hope the below article can help you on this issue:
    http://msexchangeteam.com/archive/2006/04/19/425722.aspx

    Thanks

    Allen

    • Marked as answer by Allen Song Friday, August 21, 2009 8:56 AM
    Thursday, August 13, 2009 9:53 AM
  • We had the same issue on Ex2003, the initial ESE error was caused by a user with outlook 2010 (Non cached mode) searching through his 100,000 item 4Gb inbox for an email.

    Whenever he did this the 623 error was logged after about 5-10 mins then everyone ho sent in this period or tried a web based access got denied for another 2-10 mins. public folders were also affected as they are delivered over IIS.

    Stuart


    Stuart
    Friday, September 23, 2011 10:44 AM
  • Hi Stuart, I know this is an old post, but we are now starting to see the same error with our Exchange server. How did you determine the ESE error was caused by a user with Outlook 2010 searching through his mailbox? I suspect the same is happening to use and it would be nice to point out who the culprit is.

    Many thanks in advance

    • Proposed as answer by Fat Frog Monday, September 24, 2012 5:46 PM
    Wednesday, February 29, 2012 10:59 AM
  • Hi Stuart,  we have just rolled out 400+ 2010 clients using Outlook 2010 (non cached mode) and we are seeing an Exchange database fall over on a weekly basis with  623 error .  we have 16 production databases, and we are seeing this issue with most of our databases.  we were wondering if there is a corrupt message somewhere in the system, thats triggering these long running transactions
    Monday, April 23, 2012 10:02 AM
  • HP DL380 G5 P400 RAID controller. Exchange 2003 Ent, Outlook 2010 clients

    We only started haing this problem after upgrading the clients from Office 2003 to 2010. After contacting MS they suggested running esutil commands or creating fresh databases and moving mailboxed and also apply hotfixes. We opted for new databases as the others were old and defragmented. Applied hotfix also. THat was 4 months ago and it stopped the errors immediately. Now they are back. Im pretty sure it has s.thing to do with Exch 2003 and Outlook 2010 as the server was running for years before this. THe clients have always had caching disabled as we found the performance was poor.

    Not sure what do do at this stage, will probably contact MS again. Otherwise we are upgradnig to exchange 2010 in a few months.

    Thursday, July 05, 2012 4:26 AM
  • Hi

    We  have exchange 2003 and some clients on outlook 2010. I proved it last night that when a user with a 10GB mailbox tries to do a search we get the ese 623 error and exhange server hangs for approx 20 mins.

    Has anyone come across a fix for this?

    Thursday, July 19, 2012 8:04 AM
  • Donkey, if the client hasn't cached mode enabled I'd try that first. Let it sync all the data locally and then do the same search to see if the errors appear in the event viewer. We have a support account with MS and decided to try once more with the tech but didn't get too much further. We did everything from new databases, esutil to isinteg, installing the 3 main hotfixes 953952, 893599(SP1 only), 919417. Nothing has fixed it. There is another KB911836 that I have not tried yet. MS have suggested we upgrade to a newer version of exchange server.

    One thing I did find was that emailing all staff to ensure they have less than 5000 messages in one folder did seem to improve the situation temporarily. Apparently that is a recommended limit also but isn't sustainable as people will still go over.

    From the suggestions and others findings, I'm putting my money on enabling cache mode on the clients to fix this and am absolutely convinced that it is an incompatibility or bug in an exchange 2003 server and Outlook 2010 client environment. As to why it works with some of our other offices - I don’t know. Could be more powerful servers/better disk storage or disk IO but not sure how that works with the version store memory. The HP P400 Raid controller is known to be poor without the BBWC enabled.

    As far as I am aware outlook cached mode is the recommended setting but we have had some performance problems in the past. If I'm not wrong, caching will reduce the load on the server especially when it comes to searching or working with folders that contail thousands of emails, i.e. the root of the Inbox without folders/listing 1000s of emails. Based on the forums and feedback from people it sounds like enabling caching on the clients might be the answer in our case at least.

    The errors are happening at least once a day, I'll will test over the next couple of weeks and revert.

    Tuesday, August 07, 2012 3:52 AM