locked
New SCCM 2012 environment setup RRS feed

  • Question

  • We are planning a greenfield SCCM 2012 environment for our new domain.

    It will have 1 primary server and 3 secondary servers.

    We currently have 4 physical sites that we will need to cover in this new domain.

    to start out with it will have about 1000 users and then will grow upto a maximum of 2000 users.

    Is there a tool or something that I can use to estimate requirements (server disk space /WSUS disk space/ RAM / SQL requirements/ Mandatory roles etc)?

    Thanks in advance for any help with the above.

    Tuesday, December 10, 2013 4:45 PM

Answers

  • There's not a good tool for this but Johan published some stuff very recently. It's probably on his site: deploymentresearch.com

    That's a very small site though, you can get by with fairly minimal hardware. Personally I wouldn't use secondary sites, I'd do remote DP's at the 3 sites. Those servers can be any decent hardware with lots of disk space. I generally ask for 1Tb on my DP's.

    For the primary you can host all roles and SQL on a single server. Give it at least 1Tb of disk space and as much RAM as possible. I'd say 24Gb but more is always better. CPU's aren't that important if you get RAM and fast disks. Shared SAN storage that's improperly configured is your worst enemy.


    John Marcum | http://myitforum.com/myitforumwp/author/johnmarcum/

    Tuesday, December 10, 2013 4:54 PM
  • The vast majority of traffic to remote locations (please don't use the word "sites" because that is ambiguous in this context as it has an explicit meaning in ConfigMgr) is in the form of content which is handled by the DPs; client to MP and client to SUP traffic is generally negligible. It is a sliding scale however and depends upon the number of clients as well as the amount of available bandwidth. Without knowing the details (clients/bandwidth) know one will/should give you an exact answer of which way to go though.

    One reason I tend to avoid secondary sites though is that they are single points of failure. If a secondary site server (and thus the roles it is hosting) becomes unavailable, the clients are orphaned until that secondary sites back online -- there simply is no other path for them to take.


    Jason | http://blog.configmgrftw.com

    Wednesday, December 11, 2013 1:18 AM

All replies

  • Hi

    I've used the ConfigMgr Server Sizing Spreadsheet located on this page in the past - never an exact science but it does give you a good indication of what you need from a resource perspective

    http://www.rflsystems.co.uk/resources/#!


    RDunne76

    Tuesday, December 10, 2013 4:52 PM
  • There's not a good tool for this but Johan published some stuff very recently. It's probably on his site: deploymentresearch.com

    That's a very small site though, you can get by with fairly minimal hardware. Personally I wouldn't use secondary sites, I'd do remote DP's at the 3 sites. Those servers can be any decent hardware with lots of disk space. I generally ask for 1Tb on my DP's.

    For the primary you can host all roles and SQL on a single server. Give it at least 1Tb of disk space and as much RAM as possible. I'd say 24Gb but more is always better. CPU's aren't that important if you get RAM and fast disks. Shared SAN storage that's improperly configured is your worst enemy.


    John Marcum | http://myitforum.com/myitforumwp/author/johnmarcum/

    Tuesday, December 10, 2013 4:54 PM
  • Thank you very much for your response John however can you please tell me why you think secondary sites is a bad idea?

    We were thinking of secondary sites to minimize WAN traffic which is a major issue.

    Would remote distribution points in your opinion be effective if we have slow links in these sites?

    Tuesday, December 10, 2013 11:52 PM
  • The vast majority of traffic to remote locations (please don't use the word "sites" because that is ambiguous in this context as it has an explicit meaning in ConfigMgr) is in the form of content which is handled by the DPs; client to MP and client to SUP traffic is generally negligible. It is a sliding scale however and depends upon the number of clients as well as the amount of available bandwidth. Without knowing the details (clients/bandwidth) know one will/should give you an exact answer of which way to go though.

    One reason I tend to avoid secondary sites though is that they are single points of failure. If a secondary site server (and thus the roles it is hosting) becomes unavailable, the clients are orphaned until that secondary sites back online -- there simply is no other path for them to take.


    Jason | http://blog.configmgrftw.com

    Wednesday, December 11, 2013 1:18 AM