Through this action, all virtual drives assigned to a server slot can be rapidly reassigned to another server slot. Reassignment might be required if a server encounters hardware problems. Replacing it with another server would require an administrator to be physically present. Instead, you can restore services remotely by moving storage assignments to another server slot already present in the chassis.
||Servers involved in the operation should be powered off, otherwise, operating systems and services can become unstable as volumes disappear unexpectedly.|
This action does the same assigning/unassigning operations as offered through the virtual drive assign action, but greatly simplifies the task when multiple assignments are involved. All drive positions are preserved during this operation. Only server slots that have no conflicting assignments at those positions are viable candidates. You might need to reassign existing virtual drives to other positions, using the virtual drive assign action, to make a server slot available for new assignments.
Example scenarios of server failover action:
- Server 1 (has a virtual drive at position 0)
- Server 2 (has a virtual drive at position 3)
- Server 3 (has a virtual drive at position 0)
- Server 4 (has virtual drives at positions 0, 1, and 2)
- Server 5 (has virtual drives at positions 0, 1, and 2)
- Server 6 (has virtual drives at positions 0, 1, and 2)
When the user selects Server 1 and then server failover, only Server 2 is available as a destination because no other server has position 0 available. In this scenario, the user selects Server 2 and, for convenience, checks the option to reset or power it on after reassignment. When the action completes, Server 2 has a virtual drive at position 0 in addition to the pre-existing one assigned at position 3.
The server failover feature is no longer available for compute modules that have been virtualized using the Intel® Modular Server Virtualization Manager (product code: MFSVRT). Instead, virtualized compute modules use the live migration action to move virtual machines from a failed compute module to a working one.