Results 1 to 7 of 7

Ok, so here's the senerio. 5 Domain controllers get rebooted spread out over a weeks ... Server Admins forum

  1. #1
    TheLizardFish's Avatar
    TheLizardFish no está en línea Stack level 1
    Join Date
    Jul 2009
    Posts
    17

    BES mail services & DC reboot correlation

    Advertisement



    Ok, so here's the senerio. 5 Domain controllers get rebooted spread out over a weeks time as part of a monthly patch wave. The last one gets re-booted around 2 am Friday morning.

    The next morning its noticed email is no longer getting out to the handhelds, although pins and data services were working just fine. Mail is flowing through the mail client and OWA, but just not getting sent to the blackberry devices. A simple re-boot of the BES resolves the issue. Any correlation here between the DC getting re-booted and the BES mail service connection? Any suggestions on what to look for in the application event logs to help confirm?

    We're running BES 5.0 btw with the BAS setup on seperate servers from the BES.

  2. #2
    TheLizardFish's Avatar
    TheLizardFish no está en línea Stack level 1
    Join Date
    Jul 2009
    Posts
    17
    Just in case anyone does happen to read this the answer is if you reboot your last DC the BES will loose its connection to the global catelog and not be able to send mail to the handhelds until the BES is rebooted.

    KB13157 - http://www.blackberry.com/btsc/KB13157

  3. #3
    jrockett's Avatar
    jrockett no está en línea BES Pro
    Join Date
    May 2007
    Posts
    962
    Great find thanks.
    Data Outages
    Reinstall OS
    Desktop/Device Page

    Yes, I'm a BES admin with an iPhone 3G.

    Follow me on Twitter!


  4. #4
    MWPatterson's Avatar
    MWPatterson no está en línea BES Pro
    Join Date
    Dec 2006
    PIN/ID
    Ask Plz
    Posts
    1,031
    Thx, that explains what happened to me last week.
    BlackBerry9630/4.7.1.53 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/104

  5. #5
    BES'm All's Avatar
    BES'm All no está en línea Stack level 1
    Join Date
    Feb 2007
    Posts
    10
    Indeed, this started to happen in 4.0, still occuring in 4.1.6 (MR6).as a best practice, we've been asking the DC guys to let us know once the last DC should be patched, and we schedual a BES (all of them)to reboot for early morning.
    There is also a RegHack you can use, it basically gets a DC referal from Exch instead of going directly to the DC (works most of the time for 98% of the BES). We gave up on that and just reboot BES early am's.
    We've put in several enhancement request, nothing was ever done. We did get an SDR for it so RIM is aware of it.
    If you ck your magt logs, you should be able to find keyword "CDO Mutex" as a good hint that that issue is occuring for some, but not all threads. It does something to the OS MAPI subsystem hens the OS reboot. MAPI still used in BES 5 thus still an issue, but should go away with using Exch 2K10 which uses LDP.
    BlackBerry8330/4.5.0.138 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/126

  6. #6
    BES'm All's Avatar
    BES'm All no está en línea Stack level 1
    Join Date
    Feb 2007
    Posts
    10
    Indeed, this started to happen in 4.0, still occuring in 4.1.6 (MR6).as a best practice, we've been asking the DC guys to let us know once the last DC should be patched, and we schedual a BES (all of them)to reboot for early morning.
    There is also a RegHack you can use, it basically gets a DC referal from Exch instead of going directly to the DC (works most of the time for 98% of the BES). We gave up on that and just reboot BES early am's.
    We've put in several enhancement request, nothing was ever done. We did get an SDR for it so RIM is aware of it.
    If you ck your magt logs, you should be able to find keyword "CDO Mutex" as a good hint that that issue is occuring for some, but not all threads. It does something to the OS MAPI subsystem hens the OS reboot. MAPI still used in BES 5 thus still an issue, but should go away with using Exch 2K10 which uses LDP.
    BlackBerry8330/4.5.0.138 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/126

  7. #7
    BES'm All's Avatar
    BES'm All no está en línea Stack level 1
    Join Date
    Feb 2007
    Posts
    10
    Indeed, this started to happen in 4.0, still occuring in 4.1.6 (MR6).Not being able to re-establish a DC link seems to affect the BES MAPI. As a best practice we've been asking the DC guys to let us know once the last DC should be patched, and we schedual a BES (all of them)to reboot for early morning.
    There is also a Reg Hack that kind of works, it basically gets a DC referal from Exch instead of going directly to the DC (works most of the time for 98% of the BES). We gave up on that and just reboot BES early am's.
    We've put in several enhancement request, nothing was ever done. We did get an SDR for it so RIM is aware of it.
    If you ck your magt logs, you should be able to find keyword "CDO Mutex" as a good hint that that issue is occuring for some, but not all threads. We set our notification system to scan the BES logs for this keywords.
    It does something to the OS MAPI subsystem hens the OS reboot. MAPI still used in BES 5 thus still an issue, but should go away with using Exch 2K10 which uses LDP.
    BlackBerry8330/4.5.0.138 Profile/MIDP-2.0 Configuration/CLDC-1.1 VendorID/126

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •