Pacemaker 2.0 upgrade in Red Hat Enterprise Linux 8
In Red Hat Enterprise Linux 8, the Pacemaker packages have been upgraded to the upstream version of Pacemaker 2.0.0. Pacemaker 2.0 is a major release whose main goal is to remove support for deprecated legacy syntax, in order to improve code maintainability.
Pacemaker 2.0 configuration changes
Pacemaker 2.0 changed cluster configuration in the following ways.
Changes in behavior
- The
record-pendingcluster property now defaults totrue. This change was initially implemented in RHEL 7.5. - If the
stonith-watchdog-timeoutcluster property is set to a negative number, and theSBD_WATCHDOG_TIMEOUTenvironment variable is set, Pacemaker will calculate an appropriate value. To get the previous behavior of disabling Pacemaker watchdog fencing, use 0 (which is still the default). - The
reconnect_intervaloption ofocf:pacemaker:remoteresources now accepts ISO8601 period specifications, as well as the previously accepted bare number of seconds or number with SI units.
Promotable clone resources
The resource type referred to as "master/slave", "stateful", or "multi-state" is no longer a separate resource type, but a variation of clone now referred to as a "promotable clone". "Master scores" are now referred to as "promotion scores" (though the names of the crm_master tool and master-* node attributes are unchanged).
| Deprecated | Use instead |
|---|---|
| pcs resource create ... master | pcs resource create ... clone meta promotable=true |
master-max master meta-attribute | promoted-max clone meta-attribute |
master-node-max master meta-attribute | promoted-node-max clone meta-attribute |
| masters property of docker and rkt tags inside bundles | promoted-max property in the same place |
The deprecated syntax will continue to be accepted for at least the lifetime of the 2.0.x series.
Further changes in related terminology (such as the master and slave roles) are expected in future versions. It is not being done all at once because some changes will require updates to the OCF standard and existing resource agents, which will take longer.
Deprecated options that have been removed
Before upgrading to RHEL 8, you must manually replace the functionality of the following removed cluster properties and resource meta-attributes with the indicated replacement feature.
| Removed cluster property | Use instead |
|---|---|
| notification-agent, notification-recipient | alerts |
| Removed resource meta-attributes | Use instead |
|---|---|
| isolation, isolation-host, isolation-instance, isolation-wrapper | bundle resources |
NOTE: In addition to manually replacing the functionality of these cluster properties, it is recommended that you run the pcs cluster cib-upgrade command before and after upgrading to RHEL 8 to automatically convert other legacy syntax to the current syntax.
Pacemaker 2.0 daemon changes
The Pacemaker daemons have been renamed more intuitively, to help make the logs easier to follow.
| Previous name | Current name | Purpose |
|---|---|---|
| attrd | pacemaker-attrd | node attribute manager |
| cib | pacemaker-based | Cluster Information Base manager |
| crmd | pacemaker-controld | cluster controller |
| lrmd | pacemaker-execd | local resource agent executor |
| stonithd | pacemaker-fenced | node fencer |
| pacemaker_remoted | pacemaker-remoted | remote resource agent executor |
| pengine | pacemaker-schedulerd | action scheduler |
- The master process that spawns all the others,
pacemakerd, keeps the same name. - The
systemdunit files andinitscripts that startpacemakerandpacemaker_remotekeep the same names.