Managing PCS Cluster
As a passionate DevOps Engineer with 3+ years of experience, I specialize in building robust, scalable, and secure infrastructures. My expertise spans Kubernetes, Jenkins, Docker, AWS, Ansible, Flask, Apache, Nginx, Kibana, Uyuni, Percona PMM, MySQL, and more.
Pacemaker Enable or Disable Maintenance Mode or Freeze Cluster
Managing a PCS cluster involves the coordination of a group of interconnected computers to work together as a Highly availability. It requires expertise in various areas, including hardware and software configuration, performance tuning, security, and troubleshooting. One of the main challenges of managing a PCS cluster is ensuring that all the nodes in the cluster work together seamlessly and efficiently. This can be achieved by implementing a reliable cluster management software, using automation tools to simplify management tasks, and regularly monitoring and maintaining the health of the cluster. With proper management, a PC cluster can provide organizations with a cost-effective way to tackle complex computing problems that require vast amounts of computational resources.
This write-up details the process of placing a Pacemaker cluster into maintenance mode or freezing the cluster and managing it .
Enable Maintenance Mode
1 – Run the command to place the cluster into maintenance mode.
pcs property set maintenance-mode=true
2 – Next run the command to verity that it displays maintenance-mode: true which means the cluster is in maintenance mode.
pcs property
Example Output :
Cluster Properties:
cluster-infrastructure: cman
dc-version: 1.1.15-5.el6-e174ec8
have-watchdog: false
last-lrm-refresh: 1527095308
maintenance-mode: true
no-quorum-policy: freeze
3 – Next run the command and you will see an alert at the top of the status output showing the cluster is in maintenance mode.
pcs status --full
Example Output :
Cluster name: TEST_CLUSTER
Stack: cman
Current DC: server01-cpn (version 1.1.15-5.el6-e174ec8) - partition with quorum
Last updated: Fri Jun 1 09:25:24 2018 Last change: Fri Jun 1 09:20:51 2018 by root via cibadmin on server01-cpn
*** Resource management is DISABLED ***
The cluster will not attempt to start, stop or recover services
2 nodes and 44 resources configured
Disable Maintenance Mode
1 – Run the command to take the cluster out of maintenance mode.
pcs property set maintenance-mode=false
2 – Next run the command to verity that it does not display maintenance-mode: true which means the cluster is not in maintenance mode.
pcs property
File and Directory Locations
| LOCATION | DESCRIPTION |
| /var/lib/pacemaker/cib/cib.xml | Primary cluster configuration file |
| /var/log/cluster/corosync.log | Primary cluster log file |
| /usr/lib/ocf/resource.d/heartbeat/ | Directory where resource scripts are located |
Check Cluster Status
| COMMAND | DESCRIPTION |
| pcs cluster status | Display status of cluster nodes |
| pcs status –full (double dashes) | Display detailed cluster status of nodes and resources |
| pcs resource | Display status of all resources and resource groups |
Modify Cluster Nodes
| COMMAND | DESCRIPTION |
| pcs cluster standby | Place node in standby mode |
| pcs cluster unstandby | Remove node from standby mode |
Managing Running Resources
| COMMAND | DESCRIPTION |
| pcs resource move [resource_name] node name | Move resource to another node |
| pcs resource restart [resource_name] | Restart resource on current node |
| pcs resource enable [resource_name] | Start resource on current node |
| pcs resource disable [resource_name] | Stop resource on current node |
Debugging Resources
| COMMAND | DESCRIPTION |
| pcs resource debug-start [source_name] | Force resource to start on node where command is executed showing debug information. Use --full for even more verbose output. |
| pcs resource debug-stop [source_name] | Force resource to stop on node where command is executed showing debug information. Use --full for even more verbose output. |
| pcs resource debug-monitor [source_name] | Force resource to be monitored on node where command is executed showing debug information. Use --full for even more verbose output. |
Creating and Modifying Resources
| COMMAND | DESCRIPTION |
| pcs resource agents | List available resource agents |
| pcs resource describe [resource] | List configuration setting for resource |
| pcs resource create [resource id][resource] options… | Create resource |
| pcs resource show [resource id] | Display currently configured setting of resource |
| pcs resource update [resource id] options…. | Update resource configuration |
| pcs resource delete [resource id] | Delete resource |
| pcs resource cleanup [resource id] | Cleanup resource failures |
Creating and Modifying Stonith Resources
| COMMAND | DESCRIPTION |
| pcs stonith list | List available fence agents |
| pcs stonith describe [fence agent] | List configuration settings for fence agent |
| pcs stonith describe [stonith_id] | List configuration setting for stonith agent |
| pcs stonith create [stonith_id][resource] options… | Create stonith agent |
| pcs stonith show [stonith_id] | Display currently configured setting of stonith agent |
| pcs stonith update [stonith_id] options…. | Update stonith configuration |
| pcs stonith delete [stonith_id] | Delete stonith agent |
| pcs stonith cleanup [stonith_id] | Cleanup stonith agent failures |
