System Center 2012 Configuration Manager sites,
hierarchy of sites, and Configuration Manager clients can each take
advantage of options that maintain a high level of available
service. These include the following:
- Sites support multiple instances of site
system servers that provide important services to clients.
- Central administration sites and primary
sites support the backup of the site database. The site database
contains all the configurations for sites and clients, and it is
shared between sites in a hierarchy that contain a central
administration site.
- Built-in site recovery options can reduce
server downtime and include advanced options that simplify recovery
when you have a hierarchy with a central administration site.
- Clients can automatically remediate typical
issues without administrative intervention.
- Sites generate alerts about clients that fail
to submit recent data, which alerts administrators to potential
problems.
- Configuration Manager provides several
built-in reports that enable administrators to identify problems
and trends before they become problems for server or client
operations.
Configuration Manager does not provide a real-time service and
you must expect it to operate with some data latency. Therefore, it
is unusual for most scenarios that involve a temporary interruption
of service to become a critical problem. When you have configured
your sites and hierarchies with high availability in mind, downtime
can be minimized, autonomy of operations maintained, and a high
level of service provided.
For example, Configuration Manager clients typically operate
autonomously by using known schedules and configurations for
operations, and schedules to submit data to the site for
processing. When clients cannot contact the site, they cache data
to be submitted until they can contact the site. Additionally,
clients that cannot contact the site continue to operate by using
the last known schedules and cached information, such as a
previously downloaded application that they must run or install,
until they can contact the site and receive new policies. The site
monitors its site systems and clients for periodic status updates,
and can generate alerts when these fail to register. Built-in
reports provide insight to ongoing operations as well as historical
operations and trends. Finally, Configuration Manager supports
state-based messages that provide near real-time information for
ongoing operations.
Use the information in the following sections to help you
understand the options to deploy Configuration Manager in a highly
available configuration.
High Availability for Configuration
Manager Clients
The following table provides information about the
operations of Configuration Manager clients that promote high
availability.
Feature |
More information |
Client operations are autonomous
|
Configuration Manager client autonomy includes the
following:
- Clients do not require continuous contact
with any specific site system servers. They use known
configurations to perform preconfigured actions on a schedule.
- Clients can use any available instance of a
site system role that provides services to clients, and they will
attempt to contact known servers until an available server is
located.
- Clients can run inventory, software
deployments, and similar scheduled actions independent of direct
contact with site system servers.
- Clients that are configured to use a fallback
status point can submit details to the fallback status point when
they cannot communicate with a management point.
|
Clients can repair themselves
|
Clients automatically remediate most typical issues without
direct administrative intervention:
- Periodically, clients self-evaluate their
status and take action to remediate typical problems by using a
local cache of remediation steps and source files for repairs.
- When a client fails to submit status
information to its site, the site can generate an alert.
Administrative users that receive these alerts can take immediate
action to restore the normal operation of the client.
|
Clients cache information to use in the future
|
When a client communicates with a management point, the client
can obtain and cache the following information:
- Client settings.
- Client schedules.
- Information about software deployments and a
download of the software the client is scheduled to install, when
the deployment is configured for this action.
When a client cannot contact a management point the following
actions are taken:
- Clients locally cache the status, state, and
client information they report to the site, and transfer this data
after they establish contact with a management point.
|
Client can submit status to a fallback status point
|
When you configure a client to use a fallback status point, you
provide an additional point of contact for the client to submit
important details about its operation:
- Clients that are configured to use a fallback
status point continue to send status about their operations to that
site system role even when the client cannot communicate with a
management point.
|
Central management of client data and client identity
|
The site database rather than the individual client retains
important information about each client’s identity, and associates
that data to a specific computer, or user. This has the following
results:
- The client source files on a computer can be
uninstalled and reinstalled without affecting the historical
records that are associated with the computer where the client is
installed.
- Failure of a client computer does not affect
the integrity of the information that is stored in the database.
This information can remain available for reporting.
|
High Availability for Configuration
Manager Sites
At each site, you deploy site system roles to provide
the services that you want clients to use at that site. The site
database contains the configuration information for the site and
for all clients. Use one or more of the available options to
provide for high availability of the site database, and the
recovery of the site and site database if needed.
The following table provides information about the
available options for Configuration Manager sites that support high
availability.
Option |
More information |
Use a SQL Server cluster to host the site database
|
When you use a SQL Server cluster for the database at a
central administration site or primary site, you use the fail-over
support built into SQL Server.
Secondary sites cannot use a SQL Server cluster, and do not
support backup or restoration of their site database. You recover a
secondary site by reinstalling the secondary site from its parent
primary site.
|
Deploy a hierarchy of sites with a central administration site,
and one or more child primary sites
|
This configuration can provide fault tolerance when your sites
manage overlapping segments of your network. In addition, this
configuration offers an additional recovery option to use the
information in the shared database available at another site, to
rebuild the site database at the recovered site. You can use this
option to replace a failed or unavailable backup of the failed
sites database.
|
Create regular backups at central administration sites and
primary sites
|
When you create and test a regular site backup, you can ensure
that you have the data necessary to recover a site, and the
experience to recover a site in the minimal amount of time.
|
Install multiple instances of site system roles
|
When you install multiple instances of critical site system
roles such as the management point and distribution point, you
provide redundant points of contact for clients in the event that a
specific site system server is off-line.
|
Install multiple instances of the SMS Provider at a
site
|
The SMS Provider provides the point of administrative
contact for one or more Configuration Manager consoles. When you
install multiple SMS Providers, you can provide redundancy for
contact points to administer your site and hierarchy.
|
Details for Sites and Site System Roles
that are Highly Available
The following table provides information about features
available at sites, and the site system roles that are part of a
high availability configuration.
Feature |
More information |
Redundancy for important site system roles
|
You can install multiple instances of the following site system
roles to provide important services to clients:
- Management point
- Distribution point
- State migration point
- System Health Validator point
- Application Catalog web service point
- Application Catalog website point
- Software update point (Configuration
Manager SP1 only)
You can install multiple instance of the following site system
role to provide redundancy for reporting on sites and clients:
You can install the following site system role on a Windows
Network Load Balancing (NLB) cluster to provide failover
support:
- Software update point
Note |
For Configuration Manager SP1, you must use Windows
PowerShell if you want to configure an NLB software update point
instead of using the automatic redundancy that Configuration
Manager SP1 provides when you install multiple software update
points. |
|
Built-in site backup
|
Configuration Manager includes a built-in backup task to help
you back up your site and critical information on a regular
schedule. Additionally, the Configuration Manager Setup wizard
supports site restoration actions to help you restore a site to
operations.
|
Publishing to Active Directory Domain Services and DNS
|
You can configure each site to publish data about site system
servers and services to Active Directory Domain Services and to
DNS. This enables clients to identify the most accessible server on
the network, and to identify when new site system servers that can
provide important services, such as management points, are
available.
|
SMS Providers and Configuration Manager consoles
|
Configuration Manager supports installing multiple
SMS Providers, each on a separate computer, to ensure multiple
access points for Configuration Manager consoles. This ensures that
if one SMS Provider computer is offline, you maintain the ability
to view and reconfigure Configuration Manager sites and
clients.
When a Configuration Manager console connects to a site, it
connects to an instance of the SMS Provider at that site. The
instance of the SMS Provider is selected nondeterministically.
If the selected SMS Provider is not available, you have the
following options:
- Reconnect the console to the site. Each new
connection request is nondeterministically assigned an instance of
the SMS Provider and it is possible that the new connection
will be assigned an available instance.
- Connect the console to a different
Configuration Manager site and manage the configuration from that
connection. This introduces a slight delay of configuration changes
of no more than a few minutes. After the SMS Provider for the
site is on-line, you can reconnect your Configuration Manager
console directly to the site that you want to manage.
You can install the Configuration Manager console on multiple
computers for use by administrative users. Each SMS Provider
supports connections from multiple Configuration Manager
consoles.
|
Management point
|
Install multiple management points at each primary site, and
enable the sites to publish site data to your Active Directory
infrastructure, and to DNS.
Multiple management points help to load-balance the use of any
single management point by multiple clients. In addition, you can
install one or more database replicas for management points to
decrease the CPU-intensive operations of the management point, and
to increase the availability of this critical site system role.
Because you can install only one management point in a secondary
site, which must be located on the secondary site server,
management points at secondary sites are not considered to have a
highly available configuration.
Note |
Mobile devices that are enrolled by Configuration Manager can
connect to only one management point in a primary site. The
management point is assigned by Configuration Manager to the mobile
device during enrollment and then does not change. When you install
multiple management points and enable more than one for mobile
devices, the management point that is assigned to a mobile device
client is non-deterministic. If the management point that a mobile
device client uses becomes unavailable, you must resolve the
problem with this management point or wipe the mobile device and
re-enroll the mobile device so that it can assign to an operational
management point that is enabled for mobile devices. |
|
Distribution point
|
Install multiple distribution points, and deploy content to
multiple distribution points. You can configure overlapping
boundary groups for content location to ensure that clients on each
subnet can access a deployment from two or more distribution
points. Finally, consider configuring one or more distribution
points as fallback locations for content.
For more information about fallback locations for content, see
the
Planning for Preferred Distribution Points and Fallback section
in the Planning
for Content Management in Configuration Manager topic.
|
Application Catalog web service point and Application Catalog
website point
|
You can install multiple instances of each site system role, and
for best performance, deploy one of each on the same site system
computer.
Each Application Catalog site system role provides the same
information as other instances of that site system role regardless
of the location of this site server role in the hierarchy.
Therefore, when a client makes a request for the Application
Catalog and you have configured the Default Application Catalog
website point device client setting for Automatically
detect, the client can be directed to an available instance,
with preference given to local Application Catalog site system
servers, based on the current network location of the client.
For more information about this client setting and how automatic
detection works, see the
Computer Agent client setting section in the About Client Settings in
Configuration Manager topic.
|
Details for Sites and Site System Roles
that are not Highly Available
Several site systems do not support multiple instances
at a site or in the hierarchy.
Use the information in the following table to help you
plan if these site systems go off-line.
Site system server |
More information |
Site server (site)
|
Configuration Manager does not support the installation of the
site server for each site on a Windows Server cluster or NLB
cluster.
The following information can help you prepare for when a site
server fails or is not operational:
- Use the built-in backup task to regularly
create a backup of the site. In a test environment, regularly
practice restoring sites from a backup.
- Deploy multiple Configuration Manager primary
sites in a hierarchy with a central administration site to create
redundancy. If you experience a site failure, consider using
Windows group policy or logon scripts to reassign clients to a
functional site.
- If you have a hierarchy with a central
administration site, you can recover the central administration
site or a child primary site by using the option to recover a site
database from another site in your hierarchy.
- Secondary sites cannot be restored, and must
be reinstalled.
|
Asset Intelligence synchronization point (hierarchy)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. If this
site system goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server.
|
Endpoint Protection point (hierarchy)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. If this
site system goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server.
|
Enrollment point (site)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. If this
site system goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server.
|
Enrollment proxy point (site)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. However,
you can install multiple instances of this site system role at a
site, and at multiple sites in the hierarchy. If this site system
goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server.
When you have more than one enrollment proxy server in a site,
use a DNS alias for the server name. When you use this
configuration, DNS round robin provides some fault tolerance and
load balancing for when users enroll their mobile devices. For more
information, see How to Install Clients
on Mobile Devices and Enroll Them by Using Configuration
Manager.
|
Fallback status point (site or hierarchy)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. If this
site system goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server. Because clients are assigned
the fallback status point during client installation, you will need
to modify existing clients to use the new site system server.
|
Out of band service point (site)
|
This site system role is not considered mission critical and
provides optional functionality in Configuration Manager. If this
site system goes offline, use one of the following options:
- Resolve the reason for the site system to be
off-line.
- Uninstall the role from the current server,
and install the role on a new server.
|
See Also