Failure Scenarios - IBM Storwize V5010 Manual

10.12.2 Failure scenarios

If one node fails, the other node in the same I/O group takes over the responsibility for all
volumes that are owned by the affected I/O group, as shown in Figure 10-161. The system
can deactivate the write cache, which might influence the overall performance. The multipath
driver can switch the active paths to the named node. The Metro Mirror relationship continues
to operate and provides a synchronous, consistent copy at the remote I/O group.
Figure 10-161 Single node failure in an I/O group
635
Chapter 10. Copy services