Overview of campus clustering with VCS
This overview focuses on the recovery with a VCS campus cluster. Automated recovery is handled differently in a VCS campus cluster than with a VCS local cluster.
The following table lists failure situations and the outcomes that occur with the two different settings for the ForceImport attribute of the VMDg resource. This attribute can be set to 1 (automatically forcing the import of the disk groups to the another node) or 0 (not forcing the import).
For information on how to set the ForceImport attribute:
Related information
Table: Failure situations
| Failure Situation | ForceImport set to 0 | ForceImport set to 1 |
|---|---|---|
| (import not forced) | (automatic force import) | |
| 1. Application fault | Application automatically moves to other site. | Service Group failover is automatic on the standby or preferred system or node. |
| May mean the services stopped for an application, a NIC failed, or a database table went offline. | ||
| 2. Server failure | Application automatically moves to other site. 100% of the disks are still available. | Service Group failover is automatic on the standby or preferred system or node. 100% of the mirrored disks are still available. |
| May mean that a power cord was unplugged, a system hang occurred, or another failure caused the system to stop responding. | ||
| 3. Failure of disk array or all disks | No interruption of service. Remaining disks in mirror are still accessible from other site. | The Service Group does not failover. 50% of the mirrored disk is still available at remaining site. |
| Remaining disks in mirror are still accessible from the other site. | ||
| 4. Site failure | Manual intervention required to move application. Can't import with only 50% of the disks available. | Application automatically moves to the other site. |
| All access to the server and storage is lost. | ||
| 5. Split-brain situation (loss of both heartbeats) | No interruption of service. Can't import disks because original site still has the SCSI reservation. | No interruption of service. Failover does not occur due to Service Group resources remaining online on the original nodes. Example: Online node has SCSI reservation to own disk. |
| If the public network link is used as a low-priority heartbeat, it is assumed that link is also lost. | ||
| 6. Storage interconnect lost | No interruption of service. Disks on the same node are functioning. Mirroring is not working. | No interruption of service. Service Group resources remain online, but 50% of the mirror disk becomes detached. |
| Fibre interconnect severed. | ||
| 7. Split-brain situation and storage interconnect lost | No interruption of service. Can't import with only 50% of disks available. Disks on the same node are functioning. Mirroring is not working. | Automatically imports disks on secondary site. Now disks are online in both locations - data can be kept from only one. |
| If a single pipe is used between buildings for the Ethernet and storage, this situation can occur. |