Red Hat Training

A Red Hat training course is available for RHEL 8

Chapter 63. Pacemaker cluster properties

Cluster properties control how the cluster behaves when confronted with situations that may occur during cluster operation.

63.1. Summary of cluster properties and options

The following table summaries the Pacemaker cluster properties, showing the default values of the properties and the possible values you can set for those properties.

There are additional cluster properties that determine fencing behavior. For information on these properties, see the table of cluster properties that determine fencing behavior in General properties of fencing devices.

Note

In addition to the properties described in this table, there are additional cluster properties that are exposed by the cluster software. For these properties, it is recommended that you not change their values from their defaults.

Table 63.1. Cluster Properties

OptionDefaultDescription

batch-limit

0

The number of resource actions that the cluster is allowed to execute in parallel. The "correct" value will depend on the speed and load of your network and cluster nodes. The default value of 0 means that the cluster will dynamically impose a limit when any node has a high CPU load.

migration-limit

-1 (unlimited)

The number of migration jobs that the cluster is allowed to execute in parallel on a node.

no-quorum-policy

stop

What to do when the cluster does not have quorum. Allowed values:

* ignore - continue all resource management

* freeze - continue resource management, but do not recover resources from nodes not in the affected partition

* stop - stop all resources in the affected cluster partition

* suicide - fence all nodes in the affected cluster partition

* demote - if a cluster partition loses quorum, demote any promoted resources and stop all other resources

symmetric-cluster

true

Indicates whether resources can run on any node by default.

cluster-delay

60s

Round trip delay over the network (excluding action execution). The "correct" value will depend on the speed and load of your network and cluster nodes.

dc-deadtime

20s

How long to wait for a response from other nodes during startup. The "correct" value will depend on the speed and load of your network and the type of switches used.

stop-orphan-resources

true

Indicates whether deleted resources should be stopped.

stop-orphan-actions

true

Indicates whether deleted actions should be canceled.

start-failure-is-fatal

true

Indicates whether a failure to start a resource on a particular node prevents further start attempts on that node. When set to false, the cluster will decide whether to try starting on the same node again based on the resource’s current failure count and migration threshold. For information on setting the migration-threshold option for a resource, see Configuring resource meta options.

Setting start-failure-is-fatal to false incurs the risk that this will allow one faulty node that is unable to start a resource to hold up all dependent actions. This is why start-failure-is-fatal defaults to true. The risk of setting start-failure-is-fatal=false can be mitigated by setting a low migration threshold so that other actions can proceed after that many failures.

pe-error-series-max

-1 (all)

The number of scheduler inputs resulting in ERRORs to save. Used when reporting problems.

pe-warn-series-max

-1 (all)

The number of scheduler inputs resulting in WARNINGs to save. Used when reporting problems.

pe-input-series-max

-1 (all)

The number of "normal" scheduler inputs to save. Used when reporting problems.

cluster-infrastructure

 

The messaging stack on which Pacemaker is currently running. Used for informational and diagnostic purposes; not user-configurable.

dc-version

 

Version of Pacemaker on the cluster’s Designated Controller (DC). Used for diagnostic purposes; not user-configurable.

cluster-recheck-interval

15 minutes

Pacemaker is primarily event-driven, and looks ahead to know when to recheck the cluster for failure timeouts and most time-based rules. Pacemaker will also recheck the cluster after the duration of inactivity specified by this property. This cluster recheck has two purposes: rules with date-spec are guaranteed to be checked this often, and it serves as a fail-safe for some kinds of scheduler bugs. A value of 0 disables this polling; positive values indicate a time interval.

maintenance-mode

false

Maintenance Mode tells the cluster to go to a "hands off" mode, and not start or stop any services until told otherwise. When maintenance mode is completed, the cluster does a sanity check of the current state of any services, and then stops or starts any that need it.

shutdown-escalation

20min

The time after which to give up trying to shut down gracefully and just exit. Advanced use only.

stop-all-resources

false

Should the cluster stop all resources.

enable-acl

false

Indicates whether the cluster can use access control lists, as set with the pcs acl command.

placement-strategy

default

Indicates whether and how the cluster will take utilization attributes into account when determining resource placement on cluster nodes.

priority-fencing-delay

0 (disabled)

(RHEL 8.3 and later) Allows you to configure a two-node cluster so that in a split-brain situation the node with the fewest resources running is the node that gets fenced.

The priority-fencing-delay property can be set to a time duration. The default value for this property is 0 (disabled). If this property is set to a non-zero value, and the priority meta-attribute is configured for at least one resource, then in a split-brain situation the node with the highest combined priority of all resources running on it will be more likely to survive.

For example, if you set pcs resource defaults priority=1 and pcs property set priority-fencing-delay=15s and no other priorities are set, then the node running the most resources will be more likely to survive because the other node will wait 15 seconds before initiating fencing. If a particular resource is more important than the rest, you can give it a higher priority.

The node running the master role of a promotable clone gets an extra 1 point if a priority has been configured for that clone.

Any delay set with the priority-fencing-delay property will be added to any delay from the pcmk_delay_base and pcmk_delay_max fence device properties. This behavior allows some delay when both nodes have equal priority, or both nodes need to be fenced for some reason other than node loss (for example, if on-fail=fencing is set for a resource monitor operation). If used in combination, it is recommended that you set the priority-fencing-delay property to a value that is significantly greater than the maximum delay from pcmk_delay_base and pcmk_delay_max, to be sure the prioritized node is preferred (twice the value would be completely safe).

Only fencing scheduled by Pacemaker itself will observe priority-fencing-delay. Fencing scheduled by external code such as dlm_controld will not provide the necessary information to the fence device.

node-health-strategy

none

When used in conjunction with a health resource agent, controls how Pacemaker responds to changes in node health. Allowed values:

* none - Do not track node health.

* migrate-on-red - Resources are moved off any node where a health agent has determined that the node’s status is red, based on the local conditions that the agent monitors.

* only-green - Resources are moved off any node where a health agent has determined that the node’s status is yellow or red, based on the local conditions that the agent monitors.

* progressive, custom - Advanced node health strategies that offer finer-grained control over the cluster’s response to health conditions according to the internal numeric values of health attributes.

63.2. Setting and removing cluster properties

To set the value of a cluster property, use the following pcs command. 

pcs property set property=value

For example, to set the value of symmetric-cluster to false, use the following command. 

# pcs property set symmetric-cluster=false

You can remove a cluster property from the configuration with the following command. 

pcs property unset property

Alternately, you can remove a cluster property from a configuration by leaving the value field of the pcs property set command blank. This restores that property to its default value. For example, if you have previously set the symmetric-cluster property to false, the following command removes the value you have set from the configuration and restores the value of symmetric-cluster to true, which is its default value. 

# pcs property set symmetic-cluster=

63.3. Querying cluster property settings

In most cases, when you use the pcs command to display values of the various cluster components, you can use pcs list or pcs show interchangeably. In the following examples, pcs list is the format used to display an entire list of all settings for more than one property, while pcs show is the format used to display the values of a specific property.

To display the values of the property settings that have been set for the cluster, use the following pcs command. 

pcs property list

To display all of the values of the property settings for the cluster, including the default values of the property settings that have not been explicitly set, use the following command. 

pcs property list --all

To display the current value of a specific cluster property, use the following command. 

pcs property show property

For example, to display the current value of the cluster-infrastructure property, execute the following command: 

# pcs property show cluster-infrastructure
Cluster Properties:
 cluster-infrastructure: cman

For informational purposes, you can display a list of all of the default values for the properties, whether they have been set to a value other than the default or not, by using the following command. 

pcs property [list|show] --defaults