High availability groups
Scope
A high availability group cannot extend beyond the boundaries of a core group. Therefore, a highly available component cannot fail over from a server process defined in one core group to a server process defined in a different core group.
Life cycle
High availability groups are dynamically created; administrators have no direct control over when they are created or destroyed. A group is created when component code running in a process calls the high availability manager framework to join a group. The calling component must provide the name of the high availability group for the high availability manager framework to join.
If a group with this name does not currently exist, the high availability manager creates one, and makes this member the first member of the newly created group. If the high availability group already exists, this member is added to the set of high availability group members.
To view information about the high availability groups that are part of a core group...
- In the administrative console, click...
- Servers > Core Groups > Core group settings > core_group > Runtime tab
- Specify a value in the Group name field.
- Specify an asterisk in this field if we want a list of all the high availability groups that are part of this core group.
- Specify a set of name-value pairs, separated by a comma, to get a list of only those high availability groups that contain the specified name-value pairs in their group name. For example, we might specify the following value to obtain a list of all of the high availability groups that contain...
-
IBM_hc=MyCluster
...and...
-
type=WAS_TRANSACTIONS
...in their names.
- IBM_hc=MyCluster,type=WAS_TRANSACTIONS
- Click Show groups.
A list of high availability groups contained in this core group that meet the specified criteria is displayed, with pertinent information about these groups.
Because several different components might use the high availability manager framework, it is possible to have several different high availability groups across the same set of processes. However, each high availability group always has a unique group name.
A high availability group ceases to exist when all of the group members leave the group, which typically occurs when all of the processes that host members of a given high availability group stop.
Group name
Every high availability group has a unique name. Because any component can create a high availability group for that component to use, it is the high availability group name that ties a given component to a particular high availability group. This name is a set of name-value pairs that the creating component specifies. For example:
- Company=IBM,ComponentName=TM,policy=DefaultNoQuorumOneOfNPolicy
A component can specify any number of name-value pairs to create a unique name for their high availability group.
Member state
Each member of a high availability group is either idle, active or disabled. Typically, a high availability group member will be either idle or active. A member that is idle is not assigned any work, but is available as a backup if a member that is active fails. A member that is active is designated as the member to handle the component workload.
If a member is disabled, it cannot participate in the high availability group. A disabled member is not assigned any work, and is not available as a backup if an active member fails. An administrator might disable a member if they plan to remove, delete, or cycle power on the associated server. However, this action is not required.
Policy
Every high availability group has an associated policy. The policy is used to determine which members of a high availability group are active at a given point in time. The policies available for high availability groups to use are stored as part of the core group configuration.
Related: