Monday, January 16, 2012

Impact of SQL Outage on System Center Configuration Manager 2012

Application Description
Microsoft SCCM is a set of tools and technologies that allows us to manage EUC systems and applications across the enterprise. SCCM include technologies that are used for application delivery, OS delivery, software, system and security auditing, security patching, and antivirus definition deployments etc.

It is important to note that this version of SCCM will be tightly integrated with App-V. Stand-alone App-V streaming infrastructure can be deprecated in favor of full integration with SCCM management tools and distribution points. That being the case SCCM 2012 high availability becomes an even higher priority than if it was only being used for traditional push deployments and management.

Database Storage and Performance Characteristics
<Database storage would typicall suffice an organization with 10,000 end user computing devices, 10,000 users, and 2000 applications>

Role

SQL Requirements 

Features Enabled

Purpose

Size

Tier 1 - SCCM CAS Site Server

o Database Engine Services
o SQL Server Replication
o Full Text Search
o Analysis Services
o Reporting Services
o Integration Services
o Management Tools – Basic
o Management Tools – Complete

SQL Server 2008 R2 Database

100 GB

SQL Server Log Files

50 GB

Tier 2 - SCCM Primary Site Server

Database Engine Services
o SQL Server Replication
o Full Text Search
o Analysis Services
o Reporting Services
o Integration Services
o Management Tools – Basic
o Management Tools – Complete

SQL Server 2008 R2 Database

100 GB

SQL Server Log Files

50 GB

Tier 3 - SCCM Secondary Site Server

Default

SQL 2008 Express

50 GB


 

Impact of Database Outage:
SQL databases are associated with the Central Administration Site Server (CAS) and the Primary Site Server.  Database outages at each of these tiers will produce different impact. In order to understand the impact of database outage it is important to understand the SQL replication topology of SCCM:

  • Global Data Replication: Global data refers to collection rules, package metadata, program metadata, deployments, task sequence metadata, and configuration item metadata and site control file. The global data is replicated via SQL across CAS, Primary, and Secondary Site Servers.
  • Site Data Replication: Site data refers to collection membership results, alert messages, hardware inventory, software inventory and metering, status messages and client health data. The site replication data is replicated between the CAS and Primary Site Server only.
  • Content Replication: Content replication refers to replication of the package media. Content replication is file-based and SQL has no impact on the physical content.

Impact of Database Outage at CAS

If the database is down at the CAS the following impact will be seen:

  • Impact for Administrators:
    • Administrators will not be able to connect to the SCCM console that is connecting to the CAS
    • Administrators already accessing the SCCM Console on the CAS will have to manually reset the connection to point to the Primary Site Server
    • Reporting services on the CAS will be stopped and users pointing to the CAS for accessing SCCM reports will have to point to the primary site server for reporting requirements
  • Impact for End Users
    • No new application packages can be created on the CAS and distributed, though the applications can be created on the Primary Site Server
    • New patches can not be synced up with the Microsoft website, though existing patches will be available for all clients
    • SCCM based antivirus definitions will not be applied, though these can be applied through a UNC path on another server

Impact – Medium

Summary – No new patches and antivirus definitions can be deployed via CAS


 

Impact of Database Outage at Primary Site Server

If the database is down at the Primary Site Server level the following impact will be seen:

  • Impact for Administrators:
    • SCCM will not be able to store inventory of any new clients
    • No new updates from the Secondary Site Server will move up the hierarchy in SCCM
    • Discovery Data of clients will not be updated to the database
      • Heartbeat discovery will not be updated on the database


     

  • Impact for End Users:
    • New deployments of any of the below payloads will not be advertised to the clients:
      • Application Updates
      • Software
      • Antivirus
      • Image
    • No new application packages can be created on the primary site server, though these applications will not be created on the Primary Site Server by default
    • If the Primary Site Server goes down just after installing SCCM on a new client, then
      • It will not show up in SCCM
      • Any SCCM deployments on that new client will not happen
      • New clients will not recive antivirus policies
      • New clients will not get patched
      • New clients will not get any applications or baseline policies


 

Impact – High

Summary – No new patches, antivirus definitions, and packages can be deployed to any client in the environment and no new SCCM clients can be provisioned


 

Impact of Database Outage at Secondary Site Server

If the database is down at the secondary site server level the following impact will be seen:

  • Any client updates will not be replicated between the primary and the secondary site server i.e.
    • Inventory updates will not flow to the Primary Site Server
    • Meta data of packages also called as the site data will not replicate between the primary and secondary site servers
  • New deployments of any of the below payloads will not be advertised to the clients that are assigned to the secondary site:
    • Application Updates
    • Software
    • Antivirus
    • Image


     

Impact – Medium

Summary – No new patches, antivirus definitions, and packages can be deployed to any client that are local to the secondary site


 

HA/DR Requirements
SCCM leverage SQL replication heavily behind the scenes. Replication is not directly configured in SQL but is configured as a result of changes that are made via the SCCM installer and configuration toolset. Replication will exist between the Central Administration Site Server, the Primary Site Server and a number of secondary site servers located in satellite offices internationally. The replication topology is as follows:

  • Central Administration Site Server will replicate with Primary Site Server and vice versa
  • Secondary Site Server will replicate with Primary Site Server and vice versa


 


 

3 comments:

  1. Hi Dude,

    The SCCM is the largest multiprofessional organization dedicated to ensuring excellence and consistency in the practice of critical care. Thanks a lot!

    Exchange Public Folder

    ReplyDelete
  2. This is a very good article!
    Sandeep, i was wondering if you could expand on the aboce scenarios even further?
    If a Primary Site Server were to fail (Stand-Alone site), what would be the imapact be to Site, Devices and Clients?

    many thanks...

    ReplyDelete
    Replies
    1. Thanks .. I will try to elaborate on the article.. I will add this to my to do list.. I already have a back log of around 8 articles that I am working on :(

      Delete