locked
Secondary Sites in SCCM 2012 R2 - Doubts RRS feed

  • Question

  • I have a lot of doubts that i wish to clear w.r.t Secondary Sites in 2012 R2 

    What is the communication between Secondary and Primary Site in 2012

    If the primary Site goes down , what can a secondary serve the clients with. 

    Does Secondary Site Cache data received from clients and until how long.

    How is the data transferred between secondary in 2012 to its Primary.

    Will the clients always talk to the MP that Secondary site has , or will they keep jumping around 

    How to make sure that the clients in the secondary site location only talk to the Secondary site MP / SUP

    What happens if Secondary site goes down , will the clients move to other MP/SUP

    Will Clients stop sending data when the secondary is down ? do they cache data somewhere on the themselves.

    How many clients can we handle with Secondary , considering that SQL Express has some size limitations.

    Does Secondary site accumulate client data and send it together ? what is the time frame between communication between Secondary and Primary

    Does Secondary site write directly to the SQL DB of Primary or it transfers data to primary and then that writes the data.

    How soon does the Secondary get the Policies from Primary . if someone creates a new Package , how soon a client in secondary get notified of that. 

    Does it always mean if consider to have a  Secondary , we should consider installing SUP as well (for the same reason of local Access / Bandwidth etc)

    Why should i not install Secondary for a site that has 2500 clients in it.

    How much data does a secondary MP have w.r.t a Primary MP (What is it exactly)

    Saturday, July 19, 2014 2:35 PM

Answers

  • Wow, that a lot of questions. I think you should read through TechNet first as this will address most of your questions: http://technet.microsoft.com/en-us/library/gg712681.aspx#BKMK_ChooseSecondary Also, based on the number of questions you've posed, you would benefit by brining in some outside help from an experienced consultant or contractor. This is not something you want to learn on the fly and mess up. You also should read through TechNet and or pick up a book (or two) as much of this information is contained there along with things you may not even have thought about.

    Here goes though:

    - What is the communication between Secondary and Primary Site in 2012

    Both file replication and SQL replication happen. File replication is both ways. There's a lot of different things that go back and forth so detailing them would take pages.

    - If the primary Site goes down , what can a secondary serve the clients with. 

    You need to be more specific about what you want though. Clients don't communicate with Secondary Sites really, they communicate with the roles implemented within that secondary site. Thus, anything that is cached or available within those roles will still be available to the clients. Roles at the secondary site do not require involvement from anything at the primary site to continue to operate.

    - Does Secondary Site Cache data received from clients and until how long.

    Yes. Until it can deliver what's needed.

    - How is the data transferred between secondary in 2012 to its Primary.

    As mentioned above, both file and SQL replication are used.

    - Will the clients always talk to the MP that Secondary site has , or will they keep jumping around

    They will use the MP within a secondary site if they are also within a boundary contained within a boundary group that has that secondary site server specified as a site system. By the book, clients also require the ability to communicate with an MP in the primary site but will mainly use the MP in the secondary site.

    - How to make sure that the clients in the secondary site location only talk to the Secondary site MP / SUP

    For the MP, as described above with the caveat as stated. For the SUP, this is also based upon the boundary and boundary group but no communication is required with the SUP in the primary site (as stated in the other thread, it's not actually the SUP though that clients talk to, they talk with the underlying WSUS instance).

    - What happens if Secondary site goes down , will the clients move to other MP/SUP

    No. Secondary site recovery is usually fairly painless though and secondary sites can often be rebuilt in relatively short-order (8 hours) if you have the procedures documented and tested. This is typically more than sufficient for most scenarios and organizations.

    - Will Clients stop sending data when the secondary is down ? do they cache data somewhere on the themselves.

    Yes, but what "data" are you talking about here? Software and hardware inventory is not going to magically disappear if it's not sent.

    - How many clients can we handle with Secondary , considering that SQL Express has some size limitations.

    No client data is stored in the DB on a secondary site server, that's not it's purpose in life at all so the client scalability for a secondary site is not in any way based upon this. By the book, a secondary site can support 5,000 client although this is based on recommended hardware and more can be supported with better hardware.

    - Does Secondary site accumulate client data and send it together ? what is the time frame between communication between Secondary and Primary

    Yes. Communication is constant by default but it can be scheduled and throttled. It is also compressed. As mentioned above, this has nothing to do with the SQL DB though.

    Does Secondary site write directly to the SQL DB of Primary or it transfers data to primary and then that writes the data.

    - No, the DB at the secondary, from the perspective of the secondary site, is read-only. It is also only a very small subset of the data at the primary. Thus, all data that eventually winds up in the DB is submitted via the roles at the primary site.

    How soon does the Secondary get the Policies from Primary . if someone creates a new Package , how soon a client in secondary get notified of that.

    - As the clients request them. Package creation does not result in a policy so a secondary site would never know of that. Deployments (of anything included client settings, packages/programs, applications, update groups, compliance settings, etc.)  are the main policy item synced with and thus contained with the DB at the secondary site.

    - Does it always mean if consider to have a  Secondary , we should consider installing SUP as well (for the same reason of local Access / Bandwidth etc)

    Not necessarily. Client to WSUS traffic is very sporadic -- really only once a month in most cases and then its only deltas and done via BITS. The initial catalog transfer can be large (depending upon what you've included in it) but thereafter the deltas are usually relatively small.

    - Why should i not install Secondary for a site that has 2500 clients in it.

    It's ultimately all about bandwidth utilization. If the connection to that location (don't use the word site as that is ambiguous in this context) is sufficient to handle the upward flow of traffic handled by the secondary site, then there's no reason to. Only you and your network team can truly determine how much the connection can and should handle though and when there is too much traffic.

    How much data does a secondary MP have w.r.t a Primary MP (What is it exactly)

    - I have no idea what this means.


    Jason | http://blog.configmgrftw.com

    • Proposed as answer by Joyce L Monday, July 21, 2014 7:47 AM
    • Marked as answer by Joyce L Wednesday, July 30, 2014 9:50 AM
    Saturday, July 19, 2014 5:05 PM

All replies

  • Wow, that a lot of questions. I think you should read through TechNet first as this will address most of your questions: http://technet.microsoft.com/en-us/library/gg712681.aspx#BKMK_ChooseSecondary Also, based on the number of questions you've posed, you would benefit by brining in some outside help from an experienced consultant or contractor. This is not something you want to learn on the fly and mess up. You also should read through TechNet and or pick up a book (or two) as much of this information is contained there along with things you may not even have thought about.

    Here goes though:

    - What is the communication between Secondary and Primary Site in 2012

    Both file replication and SQL replication happen. File replication is both ways. There's a lot of different things that go back and forth so detailing them would take pages.

    - If the primary Site goes down , what can a secondary serve the clients with. 

    You need to be more specific about what you want though. Clients don't communicate with Secondary Sites really, they communicate with the roles implemented within that secondary site. Thus, anything that is cached or available within those roles will still be available to the clients. Roles at the secondary site do not require involvement from anything at the primary site to continue to operate.

    - Does Secondary Site Cache data received from clients and until how long.

    Yes. Until it can deliver what's needed.

    - How is the data transferred between secondary in 2012 to its Primary.

    As mentioned above, both file and SQL replication are used.

    - Will the clients always talk to the MP that Secondary site has , or will they keep jumping around

    They will use the MP within a secondary site if they are also within a boundary contained within a boundary group that has that secondary site server specified as a site system. By the book, clients also require the ability to communicate with an MP in the primary site but will mainly use the MP in the secondary site.

    - How to make sure that the clients in the secondary site location only talk to the Secondary site MP / SUP

    For the MP, as described above with the caveat as stated. For the SUP, this is also based upon the boundary and boundary group but no communication is required with the SUP in the primary site (as stated in the other thread, it's not actually the SUP though that clients talk to, they talk with the underlying WSUS instance).

    - What happens if Secondary site goes down , will the clients move to other MP/SUP

    No. Secondary site recovery is usually fairly painless though and secondary sites can often be rebuilt in relatively short-order (8 hours) if you have the procedures documented and tested. This is typically more than sufficient for most scenarios and organizations.

    - Will Clients stop sending data when the secondary is down ? do they cache data somewhere on the themselves.

    Yes, but what "data" are you talking about here? Software and hardware inventory is not going to magically disappear if it's not sent.

    - How many clients can we handle with Secondary , considering that SQL Express has some size limitations.

    No client data is stored in the DB on a secondary site server, that's not it's purpose in life at all so the client scalability for a secondary site is not in any way based upon this. By the book, a secondary site can support 5,000 client although this is based on recommended hardware and more can be supported with better hardware.

    - Does Secondary site accumulate client data and send it together ? what is the time frame between communication between Secondary and Primary

    Yes. Communication is constant by default but it can be scheduled and throttled. It is also compressed. As mentioned above, this has nothing to do with the SQL DB though.

    Does Secondary site write directly to the SQL DB of Primary or it transfers data to primary and then that writes the data.

    - No, the DB at the secondary, from the perspective of the secondary site, is read-only. It is also only a very small subset of the data at the primary. Thus, all data that eventually winds up in the DB is submitted via the roles at the primary site.

    How soon does the Secondary get the Policies from Primary . if someone creates a new Package , how soon a client in secondary get notified of that.

    - As the clients request them. Package creation does not result in a policy so a secondary site would never know of that. Deployments (of anything included client settings, packages/programs, applications, update groups, compliance settings, etc.)  are the main policy item synced with and thus contained with the DB at the secondary site.

    - Does it always mean if consider to have a  Secondary , we should consider installing SUP as well (for the same reason of local Access / Bandwidth etc)

    Not necessarily. Client to WSUS traffic is very sporadic -- really only once a month in most cases and then its only deltas and done via BITS. The initial catalog transfer can be large (depending upon what you've included in it) but thereafter the deltas are usually relatively small.

    - Why should i not install Secondary for a site that has 2500 clients in it.

    It's ultimately all about bandwidth utilization. If the connection to that location (don't use the word site as that is ambiguous in this context) is sufficient to handle the upward flow of traffic handled by the secondary site, then there's no reason to. Only you and your network team can truly determine how much the connection can and should handle though and when there is too much traffic.

    How much data does a secondary MP have w.r.t a Primary MP (What is it exactly)

    - I have no idea what this means.


    Jason | http://blog.configmgrftw.com

    • Proposed as answer by Joyce L Monday, July 21, 2014 7:47 AM
    • Marked as answer by Joyce L Wednesday, July 30, 2014 9:50 AM
    Saturday, July 19, 2014 5:05 PM
  • Thank you Jason. this Helped. 

    Starting to read your book now.

    Monday, July 21, 2014 4:22 PM