When storage problems are suspected, run the verify_pdisks utility on all nodes in the system to find the nodes and pdisks in the clique that reference the failing device using the pcl -s verify_pdisks command. When a pdisk problem is reported, run verify_pdisks again in verbose mode (-v) to list the pdisks that are failing. When running in verbose mode, you may want to capture the output to a file for easier viewing.
When a fatal storage problem exists, it affects the pdisks on every node in the clique connected to the array with the problem. Once all failing global pdisks are known, system and storage information must be determined so that storage repairs can be made.
The following is an example of output obtained with the verify_pdisks -v command on a single node. This output is a good example of how the verify_pdisks command output shows that pdisk /dev/pdisk/dsk28 is failing due to a failed read. At this point, more information must be gathered in order to repair the failing storage device.
