spelaben

Hi, OS Deployment usually works without any issues for us. Sometimes (like 1 out of 15 deployments) the SCCM client will not install correctly and the system will boot up without any applications. That's okay, we just restart it and it works on the second attempt. Lately however, I discovered an issue with a certain model (Lenovo R500): The SCCM client configuration fails every time (reproducible!). Here are some answers to questions some of you might probably consider asking. - We are using diskpart /clean at the start of every task sequence - The system is domain joined before the SCCM client install step - Jason Sandys ConfigMgrStartup Script (client health GPO) is active - Client push install is active - The BIOS is up to date - These are our setup parameters during OSD: SMSMP=*FQDN* CCMEVALHOUR=12 PATCH="\\*FQDN*\hotfixes\configmgr2012ac-r2-kb3026739-x64.msp" - I tried installing newer network card drivers during the OSD in case the old drivers suddenly can't handle it any more but that also didn't improve the situation I thought maybe the GPO or the client push are interfering with the setup while running the OSD, but that also happens if the GPO is disabled. Also, the client push install never interfered with the OSD before and I don't think that it suddenly would. I attached the ccmsetup.log and client.msi.log, I would really appreciate it if someone could take a look at the logfiles and identify any potential issue that I'm not seeing. Thank you! logs.zip

I configured Jasons startup script (http://blog.configmgrftw.com/configmgr-client-startup-script/) in our environment and that significantly improved our client health issues. Now there are only few machines left which haven't been online in a long time and those with WMI issues.

Don't do it. Here is a link to a German blog post which describes an issue we had in our company: http://www.mssccmfaq.de/2014/04/07/die-gefahr-lauert-in-der-task-sequence-supersedence/ Basically the problem is if you use supersedence to update applications, and the specific application is also used in a task sequence, the application deployment evaluation cycle will evaluate every single machine in your environment and the machines with the specific application will receive the update... That can cause a lot more pain than the described annoyances. We only deploy task sequences to unknown computers, and in case of a reinstall just search for the MAC adress / name (if it is known) and then delete the object. If anyone has a less time consuming way of initiating a reinstall I would be glad to hear suggestions.

Hi all, I noticed an issue concerning the client health in our SCCM 2012 R2 environment. There are about 600 SCCM clients in our hierarchy, today I realized that about 40 of them don't have the "installed flag" any more (installed = no). The client was installed before because I could still access the hardware inventory etc. with the ressource explorer. The heartbeat interval is set to 7 days and the clear install flag interval is set to the maximum of 40 days. Earlier this day I had the chance to take a look at two of those clients and both had the same symptoms: - The SMS Agent Host service was just gone (as in it wasn't visible in services.msc any longer). - The site code, all the client actions etc. were gone from the Configuration Manager console in the control panel. - The eventviewer was spammed with WMI (event id: 63) warnings which appeared to be related to the SCCM client. - ccmsetup.exe /uninstall and a manual install / client push install didn't work to remove or to reinstall the client. The uninstall just started a rollback after a while and the installation gave me a return value '3' in the ccmsetup.log What I finally did to resolve the issue was to delete the folder C:\Windows\CCM and to empty the folder C:\Windows\ccmsetup because I couldn't delete the whole folder and then start a client push installation again. My questions are: - How did this happen? - What can I do to fix this error on all clients which have this problem without the need to connect to every system manually and repeat the above steps? - What can I do to prevent this from happening in the future? Thanks in advance guys! Best regards