Clustering in Azure
are there any recommendations how ClusterXL settings should be modified, when it is used in Azure environment? (modifying delays or some timeout values)
I'm curious, because i've noticed couple of times that fail over happen, but without any sign of link down/up state or VM reboot/shutdown.
ClusterXL in general works differently in Azure.
- Sync happens over unicast (due to the fact Azure doesn't support Multicast or Broadcast)
- Monitoring and failover happens using API calls to Azure
Which means most of the standard tuning really doesn't apply.
Hopefully someone from R&D can chime in on troubleshooting this since I'm less familiar with those steps.
In sk110194 Check Point provides an Example Environment that details on failover dependencies regarding to the cluster resources used, times, Azure API calls, reconfigurations made and more. Maybe this sk is of help for you.