locked
Problems after 2012 R2 SP1 CU2 update RRS feed

  • Question

  • Hi all, 

    I have installed the CU2 update at the primary site and two of my secondary sites. At one of my secondary sites, the computer installs via OSD fine, but the client will not register in the console. It looks to be fully installed on the client though and shows all the Actions and I can run them and the logs come up with no errors. The problem seems to be that the SCCM database does not recognise the client as installed. 

    I have rebooted the secondary site sever, and had a look at rcmctrl.log and despool.log and there are no replication issues. Would the client report to the secondary site server first, or to the primary directly after OSD?

    If anybody has any advice, or any logs I should look at I would be very grateful. Otherwise I will need to spend my weekend doing a full restore of my SCCM infrastructure.

    Michael


    Saturday, February 13, 2016 12:02 PM

Answers

  • OKay - found the problem but not the solution.

    The DDR record hits the primary site server, and can be observed in the MP_ddr.log

    But it is not processed due to a timestamp mismatch

    I am going to increase the heartbeat discovery from 7 days to every 1 day. I am not sure why this has become a problem in the last week when it was working fine before this?

    Michael

    • Marked as answer by MichaelSpencer Monday, February 15, 2016 5:23 PM
    Monday, February 15, 2016 5:13 PM

All replies

  • I rolled back the changes. Did a full restore of the primary site server, restored the database, restored the secondary site servers from backup, and performed a site recovery on both secondary servers. As it  turns out this issue existed before the CU2 upgrade! 

    I can build at all sites from PXE, the task sequence installs about 15 applications during the build process and they all install fine. America and Europe register the client in the console as soon as it is completed, however Asia does still shows the 'Client as no' - even though the client appears to install ok. 

    No errors in ClientIDManager.log, no errors in LocationServices. 

    What is strange though is I noticed the client showing as fully installed in the console and then it changed to 'No client'.

    Any ideas what would cause this?

    Any help would be greatly appreciated - I can see other people have had a similar issue, but these all appear to be boundary related? And have been resolved by investigating the above two logs?

    Michael

    Sunday, February 14, 2016 5:32 PM
  • Have you checked the site maintenance tasks for that site to make sure Clear Install Flag is not set or anything like that?

    Jeff

    Sunday, February 14, 2016 6:42 PM
  • Hi Jeff - the Clear Install Flag is set to every 21 days on the primary site, but clients install fine on the primary site and the secondary site in the Americas, it is only the Asia secondary site which has problems so I don't think the Clear Install Flag is the problem.

    I found a duplicate MP entry in DNS for the site 003 - which is the problem site. I have deleted this, let DNS replicate and rebuilt a client but it still shows up in the console as 'no client', currently about 5 hours after the build. Both other sites show as installed within 5 minutes of completing the build.

    These are the logs I have checked on the client.

    ClientLocation.log

    LocationServices.log

    ClientIDManager.log

    ExecMgr.log

    PolicyAgent.log

    All of these logs indicate that the client is installed fine, it can see the default MP on the primary site server, and then it successfully downloads policy.

    If I manually run a Data Discovery Cycle I believe this should send a DDR to the secondary site server, which will forward it to the primary site server. I can't see any bad DDR forwarding on the secondary in the ddr.log, so the issue I believe is on the primary site server. It doesn't seem to receive a DDR in MP_DDR.log from my client? The MP_DDR.log shows the DDR within a few seconds from my primary site and the other functioning secondary site.

    Is this the correct flow for data discovery - could anyone confirm if I am looking in the right places? How would I troubleshoot the DDR flow, how would I know if my client even sent it correctly?

    I am going to try and uninstall and re-install the MP role on the secondary site server, I will report back.

    Regards,

    Michael



    Monday, February 15, 2016 3:21 PM
  • OKay - found the problem but not the solution.

    The DDR record hits the primary site server, and can be observed in the MP_ddr.log

    But it is not processed due to a timestamp mismatch

    I am going to increase the heartbeat discovery from 7 days to every 1 day. I am not sure why this has become a problem in the last week when it was working fine before this?

    Michael

    • Marked as answer by MichaelSpencer Monday, February 15, 2016 5:23 PM
    Monday, February 15, 2016 5:13 PM
  • I'm closing this thread and opening a new one as this had nothing to do with the CU2 upgrade.

    New thread - https://social.technet.microsoft.com/Forums/en-US/51cdf6ad-0d71-4892-bc48-a2cd76dd6f66/ddr-records-discarded-from-secondary-site-in-future-timezone?forum=configmanagerdeployment

    Regards,

    Michael

    Monday, February 15, 2016 5:23 PM