Can multiple `oraasm` resources be managed on a pacemaker cluster?

Environment

• Red Hat Enterprise Linux Server 7, 8, or 9 (with the High Availability Add On)

Issue

• Can multiple oraasm resources be managed on a pacemaker cluster?

Resolution

The oraasm resource is currently unable to have multiple instances (in other words, multiple configured orasm resources managed by pacemaker) running on a cluster. The oraasm can be ran as a cloned resource or a single instance resource.

The oraasm resource does not manage which disk groups are started and stopped. The resource oraasm basically starts and stops the init script /etc/init.d/ohasd. This means that oraasm does not control which disk groups are activated and there is no way to manage a subset of disk groups between multiple oraasm instances.

Current Limitations

• The oraasm resource can only have 1 instance that is configured in a cluster. That one instance of oraasm can be cloned or non-cloned resource. If the oraasm resource-agent is cloned then make sure that your Oracle ASM can support multiple cluster nodes accessing the same set of disk groups at the same time.
• If the oraasm resource is cloned then all the disk groups for Oracle ASM should be shareable.
• Only one disk group value should be configured for the oraasm attribute diskgroup when configured as a cloned or non-cloned resource. If multiple disk groups are configured then an error could occur if thehome attribute is not configured. The oraasm attributediskgroup is only used to find the value for the home attribute (if the home attribute is not configured).
• The diskgroup attribute is not an attribute designed to specify which disk groups are enabled when the oraasm resource starts. Currently, If the home attribute is configured then the diskgroup attribute is ignored. The diskgroup is only needed in order to figure out what the value of the home attribute if the home attribute is not configured by parsing the /etc/oratab file.

There is a bugzilla opened to improve the oraasm resource agent so that multiple instances of oraasm can run and manage different disk groups with the attribute diskgroup.

Red Hat Enterprise Linux 9

• The issue is being tracked with bugzilla 1947246: This content is not included.Bug 1947246 - [RFE] oraasm: The oraasm resource agent should manage the specified Oracle ASM disk group (RHEL 8). As of Thu, April 08 2021, the status of 1947246 is ASSIGNED. An engineer has been assigned to the bug but no patch has been posted that fixes the bug.

Red Hat Enterprise Linux 8

• The issue is being tracked with bugzilla 2088962: This content is not included.Bug 2088962 - [RFE] oraasm: The oraasm resource agent should manage the specified Oracle ASM disk group (RHEL 9). As of Sat, May 21, 2022, the status of 2088962 is NEW.

Related Articles

• Support Policies for RHEL High Availability Clusters - Management of Oracle ASM Services in a Cluster
• The monitor or start operation of an oraasm resource returns a false success in a Pacemaker cluster

Root Cause

When multiple oraasm resources are managed by pacemaker it cause pacemaker to detect oraasm resources as being active on multiple cluster nodes. Which will cause the oraasm instance to be stopped. For more information on this then see:

• pengine: error: native_create_actions: Resource <name> (<class>:<provider>:<type>) is active on 2 nodes attempting recoveryseen when a node rejoins a RHEL 6 or 7 High Availability cluster

This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.