none
Store.exe and MS Information Store Service

    Question

  • Hello, I am having to end the store.exe process and restart the MS Information Store service at least once or twice a day on my exchange server. It is exchange 2007, Win 2003. The reason I do it and know that it needs to be done is because outlook will disconnect and can't reconnect back until the service has been restarted, I have to end the store.exe process first or else the information store service will hang at stopping or starting resulting in a server reboot. Any ideas as to why I have to keep doing this? There are no entries in the event viewer when this happens.
    Friday, August 20, 2010 5:22 PM

Answers


  • Hello,

    This is one of those types of issues that could be just about anything, and at least initially should be looked at from a very high level.
    I'm going to attempt to provide a list of things i would do/try by combining previous suggestions, as well as some additional items.


    1. Update to Exchange 2007 Service Pack 3:
    http://www.microsoft.com/downloads/details.aspx?FamilyID=1687160b-634a-43cb-a65a-f355cff0afa6&displaylang=en

    This ensures we are not troubleshooting any known issues with the product. I would install, reboot, and monitor. (Please be sure to install on all applicable servers and as always, make backups first!)

    2. Run EXPBA to check for any known issues:
    Open Exchange Management Console\Toolbox\Best Practices Analyzer : Select the affected server \ then health check \ Start Scanning

    Upon completion of the scan, click to "View a report of this Best Practices scan"

    Any issues detected in here?

    3. Do you have other levels of antivirus protection?:
    (Transport server, gateway, etc.) If so, can you set antivirus services to disabled, reboot, then monitor to see if the issue returns? (This obviously would not be a solution, but if it helped, would give us a better idea of where to focus)

    4. Microsoft Exchange Server User Monitor
    Download and run EXMON while experiencing the outage:
    http://www.microsoft.com/downloads/details.aspx?familyid=9A49C22E-E0C7-4B7C-ACEF-729D48AF7BC9&displaylang=en

    Launch ExMon.exe from the install directory to run.
    This tool will capture at 1 minute intervals. You can see user statistics on the server.

    Do we have 1 or maybe a couple of users utilizing alot of CPU? Sessions? Etc.
    If so, what are they doing? (Searches, moves, deletions, etc)

    5. Gather some performance data:
    The first 4 items are sort of a "slam dunk" approach, but are not necessarily bad. They ensure you are not troubleshooting a known issue with the product, a user task/session, or potentially the antivirus. If these do not help, below is a link to a powershell script written by a colleague of mine. Please review the link and consider downloading the script to gather some statistics. The link contains the download, usage information, etc.

    Please review the page in it's entirety before using.

    ExPerfwiz 
    http://code.msdn.microsoft.com/ExPerfwiz

    Ideally you would start this while things are running good, and continue to let it run into the outage.
    Once we have some data, you can then review the blg log in Performance Monitor.

    So now that we have the data, what are we looking for?
    Check this out to give you an idea of what counters to look for and what the expected values (roughly) should be:

    Monitoring Without System Center Operations Manager
    http://technet.microsoft.com/en-us/library/bb201720(EXCHG.80).aspx

    Note: The values are in the respective links beneath the article based on server role.

    Did you see any counters that are not inline with recommended values?
    If so, do these lead us to anything going on with the server?

    6. Hang Dump of the Store Process during "hang" or "outage"
    This is something that would require the assistance of Microsoft Support. You could initially obtain the dump information, but to have it analyzed you would need a support case. A hang dump is a memory dump of the store.exe process taken during the outage. We typically obtain 3 hang dumps of the process around 2 minutes apart while experiencing the issue. By having multiple dumps we are sometimes able to identify a trend, or confirm something in 1 dump that was identified in another dump.

    To read more about this, see:
    How to use ADPlus to troubleshoot "hangs" and "crashes"
    http://support.microsoft.com/kb/286350

     

    So to start, try to knock out items 1 through 5 to see if that leads us anywhere. If it comes to Step # 6 and you end up calling for asssistance, you have a great head start on the troubleshooting and data collection!

    Please let me know if you have any questions.

    Thanks,
    Kevin Ca - MSFT

    • Marked as answer by Anleifr Tuesday, August 24, 2010 8:49 PM
    Tuesday, August 24, 2010 4:28 PM

All replies

  • Can you give a little more info?  What SP level and update rollup level are you on?  What is the interval between needs to restart the service?  What AV client is running on the server?  Do you just have one exchange server?  If not, explain your Exchange environment a little more.
    Tim Harrington - Catapult Systems - http://HowDoUC.blogspot.com
    Friday, August 20, 2010 5:49 PM
  • Exchange 2007 SP2. Two exchange servers, one that has user mailboxes on it and this one...it has 3 shared mailboxes. There isn't a certain time when it happens, it is different every day.
    Friday, August 20, 2010 5:58 PM
  • only 3 mailboxes on one server?  Do you have AV client installed?  If so, have you set the proper exclusions on it?  See link:

    http://technet.microsoft.com/en-us/library/bb332342(EXCHG.80).aspx

    You also might want to apply the latest update rollup (4) for SP2:

    http://www.microsoft.com/downloads/details.aspx?FamilyID=b8f7f872-16d5-49d6-9867-adc01351c06f&displaylang=en


    Tim Harrington - Catapult Systems - http://HowDoUC.blogspot.com
    Friday, August 20, 2010 6:37 PM
  • Yeah, the shared mailboxes are 100gb, 47gb and 71gb in size. There are about 100 users that access these mailboxes, move messages into them and out of them daily.
    Friday, August 20, 2010 7:04 PM
  • Hi

    Check the disk performance

     

    Is this a VM or physical server?

    If it's a VM, check so it get's the necessary I/O's from the underlying "OS"


    Jonas Andersson MCTS: Microsoft Exchange Server 2007/2010 | MCITP: EMA 2007/2010 | MCSE/MCSA Blog: http://www.testlabs.se/blog
    Monday, August 23, 2010 6:22 AM
  • Hi, this is a physical server. Specs are over the recommended. There is a lot of disk I/O on this box because of all the activity these 3 mailboxes get. Do you think that it just gets hosed up and locks up store.exe so it has to be restarted?
    Monday, August 23, 2010 12:45 PM

  • Hello,

    This is one of those types of issues that could be just about anything, and at least initially should be looked at from a very high level.
    I'm going to attempt to provide a list of things i would do/try by combining previous suggestions, as well as some additional items.


    1. Update to Exchange 2007 Service Pack 3:
    http://www.microsoft.com/downloads/details.aspx?FamilyID=1687160b-634a-43cb-a65a-f355cff0afa6&displaylang=en

    This ensures we are not troubleshooting any known issues with the product. I would install, reboot, and monitor. (Please be sure to install on all applicable servers and as always, make backups first!)

    2. Run EXPBA to check for any known issues:
    Open Exchange Management Console\Toolbox\Best Practices Analyzer : Select the affected server \ then health check \ Start Scanning

    Upon completion of the scan, click to "View a report of this Best Practices scan"

    Any issues detected in here?

    3. Do you have other levels of antivirus protection?:
    (Transport server, gateway, etc.) If so, can you set antivirus services to disabled, reboot, then monitor to see if the issue returns? (This obviously would not be a solution, but if it helped, would give us a better idea of where to focus)

    4. Microsoft Exchange Server User Monitor
    Download and run EXMON while experiencing the outage:
    http://www.microsoft.com/downloads/details.aspx?familyid=9A49C22E-E0C7-4B7C-ACEF-729D48AF7BC9&displaylang=en

    Launch ExMon.exe from the install directory to run.
    This tool will capture at 1 minute intervals. You can see user statistics on the server.

    Do we have 1 or maybe a couple of users utilizing alot of CPU? Sessions? Etc.
    If so, what are they doing? (Searches, moves, deletions, etc)

    5. Gather some performance data:
    The first 4 items are sort of a "slam dunk" approach, but are not necessarily bad. They ensure you are not troubleshooting a known issue with the product, a user task/session, or potentially the antivirus. If these do not help, below is a link to a powershell script written by a colleague of mine. Please review the link and consider downloading the script to gather some statistics. The link contains the download, usage information, etc.

    Please review the page in it's entirety before using.

    ExPerfwiz 
    http://code.msdn.microsoft.com/ExPerfwiz

    Ideally you would start this while things are running good, and continue to let it run into the outage.
    Once we have some data, you can then review the blg log in Performance Monitor.

    So now that we have the data, what are we looking for?
    Check this out to give you an idea of what counters to look for and what the expected values (roughly) should be:

    Monitoring Without System Center Operations Manager
    http://technet.microsoft.com/en-us/library/bb201720(EXCHG.80).aspx

    Note: The values are in the respective links beneath the article based on server role.

    Did you see any counters that are not inline with recommended values?
    If so, do these lead us to anything going on with the server?

    6. Hang Dump of the Store Process during "hang" or "outage"
    This is something that would require the assistance of Microsoft Support. You could initially obtain the dump information, but to have it analyzed you would need a support case. A hang dump is a memory dump of the store.exe process taken during the outage. We typically obtain 3 hang dumps of the process around 2 minutes apart while experiencing the issue. By having multiple dumps we are sometimes able to identify a trend, or confirm something in 1 dump that was identified in another dump.

    To read more about this, see:
    How to use ADPlus to troubleshoot "hangs" and "crashes"
    http://support.microsoft.com/kb/286350

     

    So to start, try to knock out items 1 through 5 to see if that leads us anywhere. If it comes to Step # 6 and you end up calling for asssistance, you have a great head start on the troubleshooting and data collection!

    Please let me know if you have any questions.

    Thanks,
    Kevin Ca - MSFT

    • Marked as answer by Anleifr Tuesday, August 24, 2010 8:49 PM
    Tuesday, August 24, 2010 4:28 PM
  • Thank you for the info. I have actually already done alot of the things you mentioned but can't find a clear cut problem/issue. I will keep looking, again thank you very much!
    Tuesday, August 24, 2010 8:49 PM