Jump to content


Established Members
  • Posts

  • Joined

  • Last visited

Contact Methods

  • Website URL

Profile Information

  • Gender
  • Location
    Manchester, UK

Gatt's Achievements


Newbie (1/14)



  1. Thanks Peter CCMEval rings a bell with causing problems - I'll have a look and see what I can see in those logs
  2. Bear with me as this could get a bit confusing Setup: All Servers = Windows 2012 R2 (Update 1) Database Server = SQL Server 2012 R2 Problem: Site Server (SCCM-01) - the Client Agent reverts from CU3 to RTM versions I've deployed the CU3 package to all our servers (There is a separate Desktop team who have not yet deployed CU3 to Workstations). The site server's SMS Agent Host successfully updates to CU3 (5.00.7958.1401) and appears in a "Servers with CU3 Installed" Collection. After a day or so, I find that the SMS Agent host has reverted back to either RTM (5.00.7958.1000) or CU1 (5.00.7958.1203) I have made sure that CU1 is NOT deployed to any of our servers - only CU3 I've manually installed the MSP file to the Site server, but this also reverts back after a day or so. Not clear if this is affecting any other non-SCCM Servers - for now it only appears to be the SCCM Site Server Any ideas what could be causing such an issue? Happy to try and attach any appropriate logs / screenshots if needed.
  3. Got a bit of an odd problem here Setup: SCCM Server: Windows 2008 R2 SP1, SCCM 2012 R2 CU2, WSUS 3 DB Server: Windows 2008 R2 SP1, SQL Server 2008 R2 Clients - Mix of Windows 2008 R2, Sever 2012 R2, Windows 7 & Windows 8 Problem: I recently had to rebuild SCCM due to database corruption (In turn due to a faulty VM host) All went well with the rebuild and all the clients that were on at the time (all servers and 1 Windows 8.1 client) all reported in fine, updates to CU2, got the test App deployment (VLC) and their SCEP clients all updated via ADRs However.. Not the same story for any client that was NOT ON (Windows 7 / Windows 8.1) The get VLC, they get CU2, the get the Custom Client Policies, the SCEP definition policies What they dont get is any updates (the SCEP ADR to be specific) I have check the logs and whilst WUHandler reports the correct SUP server URL, the WindowsUpdate.log shows the client going direct to Microsoft to get the SCEP definitions I have tried repairing the clients, uninstalling the clients and removing the SMSCFG.ini and then reinstalling - to no avail. What I have also noticed that if the Registry Key (HKLM\SOFTWARE\Policies\Microsoft\Windows\WindowsUpdate, etc) are removed (either manually or via uninstalling the client, the DONT get recreated in full (ie the key is created along with the AU subkey, but the WUServer values are not created with the SUP URL). If I manually create these then the logs show it is trying to contact the WSUS server, and the UpdateStore log lists a string of available updates. The WUHandler logs also appear normal with no errors anymore I've attached my windowsUpdate.log - the main error is near the end: 2014-09-08 10:31:23:689 1000 118c Report WARNING: CSerializationHelper:: InitSerialize failed : 0x80070002 But googling just tells me to stop wuauserv and to delete softwaredistribution and a couple of reg keys, and then restart it and reauthorise it - done that to no avail I'm loathed to remove the SUP role as it is working for about half my clients.. Any and all help/advise greatly appreciated....
  4. Hi folks Sorry not had a chance to email you yet Rob Chris - I've watched through your webcast and its made things a lot clearer now regarding RBA and can see how that can work for us. There is only one more are that I forgot to mention - and its causing us some grief at the moment with CM07 - but I suspect this may be easily remedied with the correct setup And that is this: We have multiple VLANS - two of important note are: Front-End: which is basically all our Desktops and various servers that need to be seen by the Desktops, but not all servers are on this one Management: which ALL our servers are connected to. CM07, when it was installed [not by me I hasten to add!] currently uses the Front-End VLan for communicating to clients and this is causing us great issues with those servers that do not have visibility of of it (IE only have a connection to the Management side) We suspect that it's due to the MP being on the Front End side and not the Management, so would having 2 MP roles installed (one on the Front End, and one on the Management side) or via boundary groups (as they have different IPs) Or does CM12 handle this better anyway?
  5. Thanks Rob - will write one up shortly (UK Teatime here!)
  6. Thanks, I'm now suitable unsure as to CAS or not to CAS to the point where I'm swaying to the latter... I'll take a look at the webcast and take all the comments under advisement with regards to our design. I'll setup a singe test Primary Site and look at the RBA configurations, etc Damn you MS and the CAS!!
  7. I understand that a CAS is not technically needed, but in our organisation it is - for a few reasons 1) Reduce the load from the clients - at the moment CM07 gets a right battering from all our clients - desktops and servers - to the point where the log files fill up fast, the inboxes regularly get 50,000 retry entries which then grinds the system to a crawl (first indication is that we cannot refresh collection membership) until we stop CM07, delete the Retries and reboot the server(s) 2) Internal Politics - The "Desktop" team and the "Server" Team are both battling over who has the most control over the system. Mis-management of packages on both sides have caused serious problems across the university These 2 points are both listed here as possible scenarios for a CAS - http://blogs.technet.com/b/server-cloud/archive/2012/02/29/hierarchy-design-in-system-center-2012-configuration-manager.aspx So, for these reasons we are looking at a CAS so that both teams have their own site to deal with, and a select few have access to the CAS itself should there be a need to apply settigns to the entire estate. Nor am I saying it will solve our problems - but it may limit them a bit. So, back to my original query - can there only be ONE server in the "CAS" - ie the CAS site [or role] server itself? From what I've found out, when using a CAS you MUST install SCEP and the Asset Intelligence Sync Point on the CAS, we would also look to install the SUP role here too with more granular control of SUP given the the Server and Desktop sites. So, we NEED two primary sites in order to end the tug-of-war between teams, and for that we NEED a CAS. All I need to know is the query above.
  8. Hi folks looking for some help in understanding how the CAS works with regards to Server Roles I work at a University in the UK and we are in the middle of migrating our ailing SCCM 2007 environment to a new SCCM 2012 R2 environment. At present our CM07 environment is a single primary site that covers ALL computers and servers across the campus (Staff, Students, Servers) and we have something like 5000 clients in total and the whole environment is a mess and its common to see software deployed incorrectly (packages for desktops have been assigned to servers as well causing much disruption) So, with the CM12 setup we are looking at implementing a CAS with 2 Primary Sites - one for the Desktop side and One for the Server side - to not only prevent accidental deployments of desktop packages to servers but to also lessen the load being placed on it (CM07 regularly falls over due to the amount of clients talking to it, etc). I know that we technically do not need a CAS, but our environment would benefit more from splitting the Desktop/Server sides into separate sites. Anyway - My query regarding this CAS... Is the CAS a single "All-In-One" site? That is - Is it literally only the CAS Server itself with all appropriate roles installed on it? (Specifically SCEP, SUP and the Asset Intell Sync Point) or can we have a 2nd server with these roles installed on them? Hope someone can help
  • Create New...