1

I have an IBM SAS drive that has enough faults to push the Predictive Failure counter past the threshold. I have removed the drive from the array (state is Unconfigured Good), but decided to keep it in one of the drive bays for now.

I have Megaraid Storage Manager 17.05 configured to send e-mail alerts, and it's doing it every day for this errored drive. Does anyone know if there's a state I can put the drive in that will make the warnings stop? I tried 'prepare for removal', which spins it down but keeps sending the alerts. I have not yet tried pulling it out and back in to make it go into foreign/bad/offline state, since it's in a remote location.

theultramage
  • 393
  • 1
  • 4
  • 14
  • I'll counter with a question: Is there any good reason to keep a drive that has been predicted to fail in a bay? That sounds like a good way to enable potential data loss due to operator error. – Mikael H Feb 15 '19 at 08:49
  • It's not going to be used anymore (except maybe as scratch space messing around), I just don't have an empty bracket to plug the slot with, don't really want to unscrew the bracked off of a working disk that can't be plugged in anywhere else, and don't want to leave the disk lying around, adding to the clutter. – theultramage Feb 15 '19 at 12:23

1 Answers1

2

Getting the drive into an 'unconfigured bad' state did the trick.
I created a dummy virtual disk, then ejected the drive and plugged it back in. Now MSM is unhappy about there being a bad disk, but it is not sending alerts anymore. Curiously, the disk's Predictive Failure counter is showing 0, as if MSM hasn't even tried reading that information yet. It's probably why this move works.

theultramage
  • 393
  • 1
  • 4
  • 14