Strawberry’s High Availability Cluster is based on the open source Linux-HA project. It uses the Pacemaker, PCS, Corosync Cluster Engine, CMAN, and CRMShell components.
The so-called „heartbeat“ is sent via UDP packets on a dedicated LAN-LAN crossover cable between both server nodes to determine the other node’s state. This connection is also used for DRBD and PostgreSQL data replication.
When the servers are running it must never be disconnected! Removing this cable will result in a split-brain condition, which should generally be avoided.
Also, you should never start or stop Strawberry services manually using init.d scripts, as these are managed by the corresponding resource agents. Manual intervention can cause data inconsistencies and other errors!
Please use the appropriate CRM shell command to start or stop the Strawberry resource group.
- Active/Passive Cluster Architecture Diagram
- Monitoring the Cluster’s Health
- Controlling the Services
- Controlling the Nodes
- Migrating the Strawberry Resource Group
- Switching the Complete Cluster into Maintenance Mode
- Troubleshooting
- Links to External Package Documentation
- Addendum
Need more help with this?
Visit the Projective Support Websites