herrando Posted June 4, 2013 Report post Posted June 4, 2013 Hi I have a problem I am hoping someone can assist with. A new secondary site was deployed recently and has been working absolutely fine. We could image PCs at the secondary site and they would appear in configmgr console with the correct primary site code within 25-35 minutes of them being deployed. This seems to have stopped happing recently. Any new PCS being built instead either don't appear in the configmgr console, or appear only when Active Directory system discovery is ran, when this happens the client are assigned to the secondary sitecode instead of the primary, and can receive no advertisements as a result of this. Here is the generic Info: All sites running SCCM 2007 SP2 R3 SQL database on separate server cluster AD schema extended Mixed mode operation. I have been though quite a few things to try and resolve this issue to no avail. Checked boundaries were correct - Initially I was using AD sites and services, I noticed that there were some issues with this so I have replaced this with the IP address ranges of the site instead. Verified no overlapping boundaries are present. Manually went though the boundaries and could find no overlaps, I also found a script that checked this and nothing was found. Check the health of the management point - Checked using MPTroubleshooter from ConfigMgr 2007 Toolkit V2, on the primary site all checks came up clear, but the secondary came up with a warning that the secondary site wasn't administrator on the SQL server - I run SQL in a separate cluster, I have now added the secondary site to the administrator group here and re-ran the test to find that it all comes back green now. running http://sbpssccm01/sms_mp/.sms_aut?mplist shows the primary and all secondary sites fine. running http://sbpssccm01/sms_mp/.sms_aut?mpcert shows the certificate Ensured that the secondary site is set to publish the site in active directory domain services, and also tried setting to publish in DNS. Checked System Management container in AD and verified that the secondary site is present and is being updated, the secondary site computer account has permissions to "this object and all descendant objects" On the Task sequence for my OSD I have specified additional installation properties MP=SBPSSCCM01.mydomain.net FSP=SBPSSCCM01.mydomain.net SLP=SBPSSCCM01.mydomain.net - my SBPSSCCM01 server runs in mixed mode and I have the FSP and SLP role installed on the same server. Using these settings my client is being installed, assigned to the correct site and appearing in SCCM correctly at all other secondary sites. On the client side I have verified that the client installs successfully, I have verified using Smsclient centre that the site is allocated to the correct MP and PMP. Sometimes the clients don't seem to automatically download a policy, I have to go onto the machine and trigger the machine policy retrieval and evaluation cycle. After this the machine gets it's policy fine. I run Discovery Data collection cycle, as I understand it, heartbeat discovery is the only thing that causes the configmanager console to update from client=no to client=yes and also renew the sitecode. This seems to run fine. Heartbeat discovery is set to run every hour on my secondary and primary site, I set the "clear client install flag" to only run once every 7 days. If I manually install the client over the top of the existing one pushed down via the TS with the MP hardcoded, or MP set to auto, the client will appear and work fine pointed at the correct site code (the primary). It seems that after a day or two the client reverts back to the secondary site, or is marked as client=no, yet you can still remote tools onto these machines and see the configmr client is still installed, and still seems be assigned correctly. Using Client Push results in the same issues. I have attached some logs when seem to have a few blips in them, but upon looking up the errors it seems to be normal operation. Open to any and all ideas... including crazy ones! - edit - uploaded some log files from the wrong client, reuploaded them. CAS.log CcmExec.log ClientIDManagerStartup.log ClientLocation.log DataTransferService.log execmgr.log LocationServices.log StatusAgent.log Quote Share this post Link to post Share on other sites More sharing options...
herrando Posted June 12, 2013 Report post Posted June 12, 2013 Well just a quick update, I had a specialist contractor come to look at this and he was also unable to resolve/identify the issue. It just looked like everything should have been working fine. Clients were getting the correct MP information and retrieving their policies fine. I've now taken the somewhat drastic step of building another secondary site at the same location and all is working fine there. I will be decomissioning this secondary site, the only assumption I can make is that there was something fundamentally broken on the site, despite all tests passing using MPTtroubleshooter and the CMRAP coming back with no issues the site did not function as expected. Quote Share this post Link to post Share on other sites More sharing options...
