이 콘텐츠는 선택한 언어로 제공되지 않습니다.
Chapter 8. Clustering
Dynamic Token Timeout for Corosync
The
token_coefficient option has been added to the Corosync Cluster Engine. The value of token_coefficient is used only when the nodelist section is specified and contains at least three nodes. In such a situation, the token timeout is computed as follows:
[token + (amount of nodes - 2)] * token_coefficient
This allows the cluster to scale without manually changing the token timeout every time a new node is added. The default value is 650 milliseconds, but it can be set to 0, resulting in effective removal of this feature.
This feature allows
Corosync to handle dynamic addition and removal of nodes.
Corosync Tie Breaker Enhancement
The
auto_tie_breaker quorum feature of Corosync has been enhanced to provide options for more flexible configuration and modification of tie breaker nodes. Users can now select a list of nodes that will retain a quorum in case of an even cluster split, or choose that a quorum will be retained by the node with the lowest node ID or the highest node ID.
Enhancements for Red Hat High Availability
For the Red Hat Enterprise Linux 7.1 release, the
Red Hat High Availability Add-On supports the following features. For information on these features, see the High Availability Add-On Reference manual.
- The
pcs resource cleanupcommand can now reset the resource status andfailcountfor all resources. - You can specify a
lifetimeparameter for thepcs resource movecommand to indicate a period of time that the resource constraint this command creates will remain in effect. - You can use the
pcs aclcommand to set permissions for local users to allow read-only or read-write access to the cluster configuration by using access control lists (ACLs). - The
pcs constraintcommand now supports the configuration of specific constraint options in addition to general resource options. - The
pcs resource createcommand supports thedisabledparameter to indicate that the resource being created is not started automatically. - The
pcs cluster quorum unblockcommand prevents the cluster from waiting for all nodes when establishing a quorum. - You can configure resource group order with the
beforeandafterparameters of thepcs resource createcommand. - You can back up the cluster configuration in a tarball and restore the cluster configuration files on all nodes from backup with the
backupandrestoreoptions of thepcs configcommand.
