Recovering from hardware failures, M2002 model

In the situation where you experience a hardware failure on an appliance, there are various steps that you can take to get your queue managers running again as soon as possible.

If you detect a hardware failure, for example, indicated by the test hardware command (see Hardware testing from the CLI), contact IBM support for assistance. However, in some cases and depending on availability of spare components, there might be local actions that you can take to quickly restore your appliance queue managers to operation.

The M2002 appliance has four solid-state disks (SSDs), in a RAID10 configuration. Data is striped and mirrored across the four disks, as illustrated in the following diagram.

This configuration means that sometimes data is retrievable if two of the four SSDs fail, and sometimes not, depending on which disks fail. For example, in the configuration in the illustration, data could be retrieved if disks 1 and 3 fail, but not if disks 1 and 2 fail.

The following topics describe scenarios that might occur and give step-by-step instructions to recover from them.