Soft Diagnostic Failure

A soft diagnostics Error will typically be present after an intentional or unintentional restart of one or both HSV controllers. It will be logged in the controller event log and displayed in command view Enterprise Virtual Array (EVA) as yellow exclamation mark. The EVA will be operational. This error can be present on any EVA model.

HP EVA Soft Diagnostic Failure

Solution!

I know, this product is retired but I’m sure some of you still using this product. So if the error happened for you, what’s the solution?

When any HSV controller restarts it will go through the equivalent of a Power On Self Test (POST). The POST will ensure that all required hardware is present and operational.
In case a non-essential part of the hardware fails diagnostics (e.g. one host port is not operational), the controller will start up but be marked with a Soft diagnostics error.
A typical scenario would be a host port failure that triggered a controller restart and subsequently failed diagnostics (because e.g. a bad cable). After the restart, the controller will report the soft diagnostics error.

A soft diagnostics error will only clear on the next successful POST (and, of course, if the problem is fixed at that time). So a controller restart is required. As a re-synchronization bypasses the POST, it is insufficient to clear the error.

Transient problems (e.g. a marginal switch link) can cause a soft diagnostics error on a host port while the host port might have recovered on its own thus showing no physical problem. A simple reboot of the controller will clear the condition.

As always, as a reboot takes one controller offline, present host I/O load needs to be taken into consideration before restarting. In case both controllers display this error, they can be restarted one after the other once the cause of the error is rectified.