SMS, SCCM, MECM, MEM, MCM & other fun acronyms

I have recommended the appdeploy.com web site to countless customers over the year as THE software packaging reference site.  Now, that url takes you to itninja.com and it looks like Dell has done their best to turn a great site into something not nearly as intuitive and useful.

During the Configuration Manager Risk Assessment Program (RAP) offering that MS Premier provides to customers one issue we look for, and usually find, is the fact that SQL memory has not been limited.  Now, on a production SQL server that may not be an issue.  But, if SQL is running locally on the primary or CAS, that might cause some memory allocation issues for other functions running on the server (like Configuration Manager). 

It’s good to see that we now check for this condition during the pre-req check for the install of ConfigMgr 2012. 

 

 

Been using the beta for a while, now the 2.5 RTM is available here http://technet.microsoft.com/en-us/solutionaccelerators/cc835245.  Not sure what SCM is?  Come see our Instructor Led Labs and Breakout Session on Security Compliance Manager at MMS! #mms2012

 

…and found anything here that is useful please consider making a tax deductible donation to the bike ride (the MS150) I am doing in a few weeks to support Multiple Sclerosis (MS).  The ride is May 4-5 in the North Texas area and we will ride over 150 miles over the course of two days.  Any amount you want to donate would be very helpful in helping me reach my goal.  Here is the link http://main.nationalmssociety.org/site/TR/Bike/TXHBikeEvents?px=9029062&pg=personal&fr_id=18029

thanks!!!!

 

If you have worked with SMS/ConfigMgr over the years you understand that the product is continually evolving and adding more functionality.  In many companies, ConfigMgr is, or is close to becoming, a core critical IT service.  With the said, what are the high availability options for ConfigMgr 2012? 

We can utilize a SQL cluster for the central administration site (the CAS) and for the primary site servers.  We cannot use a SQL cluster for the secondary site.

Configuration Manager 2012 also provides the ability to install multiple instances of several site system roles to increase availability, including the management point, distribution point, state migration point, the application catalog roles and the reporting services point.  You can also use an NLB cluster for the software update point role.  In ConfigMgr 2007 we provided the ability to use an NLB cluster for the management point role, but in ConfigMgr 2012 we replace that with the ability to add multiple management point servers to the hierarchy.

But what about high availability for the CAS, primary or secondary site server roles themselves?  There is no clustering or NLB support for those roles.  If the CAS or primary site servers go down, our recovery model is to use the backup/recovery process in ConfigMgr.  If the secondary site server goes down, the only recovery model is to reinstall the site.  You could possibly make the case to backup the package files to avoid having to repush those across the WAN, but backing up the secondary site server itself (and the database) is unnecessary. Also we do not support restoring any ConfigMgr components/servers using the snapshot feature that virtual server products provide.  You might use it, it might work in a recovery, but completely unsupported.

But what if I decide to have two primary site servers (and a CAS, since you would need it to ‘bind’ the two primary’s together) and have one primary there (say it’s site code PR2) in case the other primary (PR1) goes down (or vice versa).  In that scenario, your plan might be to temporarily assign the clients from PR1 to PR2.  Then, rebuild PR1 and then move the clients back from PR2 to PR1.  Temporarily assigning clients to another primary site is possible, but may introduce issues. After assigning the clients to the other site the clients will begin submitting data (inventory, compliance settings data, Endpoint-related data, software update compliance, etc) to the newly assigned site.  Once the original site is recovered and the clients are assigned back to it, the clients would exist and be viewable in both primary sites and would remain there until that data was manually deleted or triggered to delete once it become aged. This may introduce various issues around targeting and software delivery and perhaps other unforeseen issues. The more effective recovery scenario would be to recover the original primary site and replicate information from the central administration site database (if a CAS has been implemented). Otherwise, restoring the site using the ConfigMgr backup/recovery process would be the recommended option.

A customer I work with alot has an opening in their Oklahoma City location for a ConfigMgr administrator.  They are great folks, they are aggressive with new technology and this would be a great opportunity for the right person.  If you have an interest email me and I’ll put you in contact.  breben @ microsoft . com

 

Configuration Manager has (and has had for a very long time) the ‘All Systems’ collection.  It contains discovered computers as well as actual clients.  In ConfigMgr 2012, what you see in this collection (especially at ‘child’ primary sites) may be a bit unexpected. 

For example, in ConfigMgr 2007, if you had a central ConfigMgr site (say site code CEN) and two primary sites (say site codes PR1 and PR2) and you viewed the ‘All Systems’ collection on PR1, all you see are the discovered objects and clients of that site.  Same holds true for PR2, which may have been configured to discover a different AD domain, OU, whatever.  Look at the ‘All Systems’ collection on the CEN site and would see all of the objects that PR1 and PR2 know about.  But, PR1 would have no knowledge of the computers discovered on PR2, and vice versa. Makes sense.

Now, in Configuration Manager 2012, EVERY primary knows about ALL objects.  PR1 knows about devices discovered on PR2, PR2 knows about devices discovered on PR1, and the CAS knows everthing too.  Welcome to the world of DRS in ConfigMgr 2012.  If you don’t plan RBAC properly (or at all), this might pose some issues, especially around targeting baselines, packages, client agent settings, etc.

 

You may encounter the error “SQL Server x is not running version SQL Server 2008 or SQL Server 2008 R2” even though the server IS running the correct version of SQL server.  Try the steps provided in this article and it should resolve the issue.
http://support.microsoft.com/default.aspx?scid=kb%3bEN-US%3b956013.  Note that your path may be a bit different than the one shown in the KB.

 

This information came out in RC2 and we'll see if it is changed once ConfigMgr 2012 goes RTM.

·        SQL Server 2008 SP2 Standard and Enterprise  CU7

·        SQL Server 2008 R2 SP1 and CU4

·        SQL Server Express 2008 R2 and CU4 (secondary sites)

 

http://technet.microsoft.com/en-us/library/gg682077.aspx#BKMK_SiteSysReqFunction