Drobo detected a drive was removed - No, it definitely wasn't!

I just got 3 alerts from my Drobopro.

1st one telling me a drive had been removed - WTF?

2nd one telling me it cannot protect my data.

3rd one telling me it is rebuilding drive protection - not to remove any drives/power off etc.

So how did a drive go from “being removed” (which is physically not possible as no one is home!) to now being re-inserted and fine (i.e. it’s now fine and is being re-written with data and parity - aka rebuild.

This is rather alarming to put it mildly!!!

the drive stopped responding - so drobo thought it had gone

it could not protect your data as it did not have enough space to rebuild onto / it was no longer redundant

when the drive started responding again drobo thought it had been re-added and began rebuilding onto it

its either a loose connection or a drive starting to fail.

send your logs to support and they will tell you which drive.

Thanks for your response Docchris.

I have to send logs to support to find which drive caused the issue? Shouldn’t I have access to that information from the UI? What do I do at 3am when the support is closed and I have an issue with my Drobo?

A loose connection wouldn’t wash with me (no offense meant) - how would it have moved to become loose suddenly? I would hope this can be safely ruled out.

And if a drive is bad, why re-introduce it to the array? Surely it should be excluded and marked as bad?

I’m astounded by the lack of factual information I get from the UI and device itself - I’m sat here with 3 critical alerts and absolutely no idea what has or is currently happening.

Well its hard for drobo to tell between a drive which has stopped responding it one which is missing
But yes, the UI needs some kind of log / history so you can see what happened and when and what drive caused it

Well I certainly can’t do anything BUT send the log file off - it’s completely unreadable. That’s a shame.

Well this would be a feature request - give us FAR better access to clear and concise information, please. (obviously not directed at you, Docchris)


i think they are working on /looking at splitting the log file in two - a user readable part - and a part which contains all the proprietary stuff about how drobo actually lays out its data.

Yeah that would be good. Or maybe a section in the UI, much like a Windows Event Log - something tracking all information, reboots, drive status, everything.

I.e. we can have email alerts set to inform us of any meaningful information. Having a viewable log with ‘any meaningful information’ in it would be great.

I’ll wait for the next dashboard release with baited breath :stuck_out_tongue:

Oh and I’ve just logged a support call with the log attached. Fingers crossed I get some good information back :slight_smile:

well in some respects the e-mails are kind of like an ongoing event log - all they need to say is “drive 6 failed” rather than “A” drive failed which is about as much use as a snooze button on a smoke alarm

Yeah that’s exactly how I see it. Though knowing the device rebooted while you were away would be nice. Arguably you could just setup snmp traps on the iSCSI host I guess.

I should add - my Windows 2008 server event logs are full of iSCSI issues. For example Event ID 9 - iSCSIPrt - Target did not respond in time for a SCSI request. The CDB is given in the dump data.

I will do some investigating…

And add the name of the Drobo unit for those of us with multiple units. :slight_smile:

It’d save me from having to mount a webcam on the shelf above my Drobo.