P4000 and unexpected power downs?

[Paul Hutchings]

Hypothetical situation.

Let's say I have either two or four nodes of P4x00.

Let's say that due to a power outage that led to the UPS draining that I lost all nodes.

Once power is restored and the nodes come back up, due to it being an uncontrolled shutdown, is there any additional downtime whilst checks/rebuilds/resyncs are done?

We had a situation today where our power went and our UPS drained and fortunately the dedicated UPS on our current SAN held out, but it got me wondering what happens with more complex SANs (ours is a pretty simple controller + disk shelves model).

[Bryan McMullan]

Typically as long as there was no real damage to the drives during the outage which would corrupt the volume, you will only experience slowness due to volume resyncs (basiscally a bit of management IOP overhead). There is also some time associated with the system establishing quorum when the managers come up and the such, but that is relatively minimal.

We have unfortunately had this happen when someone overloaded a UPS on our test SAN and everything went down at once.