Drobo

Got an error message, drobo volume trashed, now super slow

I didn’t get the error message, as it popped up while typing and I hit enter and cleared it as part of my typing.

However the drive disappeared, and wouldn’t mount.

I tried disk repair, it ran for a few hours and finally said it could fix it.

So since it’s my backup volume I just nuked from orbit and recreated a volume. At first I did journaled/encryption as I figured that would be nice to have.

I have 1.77TB to back up, Time machine told me it’d take 23 days. So I thought oh uh, no encryption then. Blew it away, repartitioned as no encryption.

Time machine now says 39 days

I have the Drobo Gen2. I manually updated to the latest dashboard because it never told me I needed an update (post error message) as part of my checking.

I’m using the FW800 connection on a MacPro 1,1. It’s been solid for a year or two, this is my first hiccup.

I have 4x2TB WD Drives in, shows green.

Rebooted system, shut down drobo/unplugged replugged. Still slow as snot.

Mac OS X Lion 10.7.1 upgraded from Snow Leopard. Was not slow until whatever happened that corrupted the volume.

Any ideas?[hr]
I went back in and looked at the various screen in the new (pretty) dashboard. I noticed that it said that there was still 3TB or so in use. So I nuked it from orbit and re-created using the drobo tools.

Let me mention, it’s been maybe 2 years since I’ve configured this, so I was a bit rusty. I let drobo create & format the partition.

It appears to be running a bit faster, still seems a lot slower than I remember. Now it’s saying 9 days and slowly coming down.

i’ll let it run, but this is really weird. I’ll admit I’ve not ever backed up from scratch in a year or so maybe, and data has slowly accumulated on my 2.5TB drive array inside my Mac Pro. However formatting it / reseting device to defaults seems to have improved it some. no longer just getting DSL speeds (1MB/s)

HOLD ON!!!

All your nuking and repartitioning may get you into a bad spot, especially if you had a lot of data on the volume(s) beforehand.

Drobo’s housekeeping takes a bit of time to catch up when there are large deletions. If you manage to get yourself into a situation where there are no (already) available blocks and Drobo hasn’t freed new blocks (because the housekeeping hasn’t caught up yet), then bad things could happen.

So… Wait a while. At the very least, it might just be Drobo’s garbage collection that’s slowing down the Time Machine writes.

Regardless, check your capacity lights to make sure you don’t get close to full in the meantime.

grr. I wished it’d said something like that when I nuked it. It’s performing better now, bout 8MB/s write and 16MB/s read and the backup finished. I was in a bit of a panic as I was unprotected for a bit. Wish I knew what cratered it to begin with.

It’s a non-issue in normal use, but definitely can lag when you do a multi-terabyte move or delete as I have done in the past.
Glad to hear it sorted itself out.