What does Greyhole do when a drive goes bad?

[crayserver]

Hi - I am new to all of this (Linux, Amahi and Greyhole). I have made the switch from WHS 2011 to Amahi, and so far I am very happy. I have almost 3TB of data to protect. I have setup Greyhole with the LZ on root (but plenty of space for it), and 4 SATA drives as the storage pool totaling 9TB. Everything has been copied over successfully, and Greyhole has made the extra copy.

What I want to understand is what would happen if a drive in the storage pool goes bad? Does Amahi or Greyhole let me know that a drive has gone bad? Does Greyhole then recreate the files that were on the bad drive and move them to good drives in the pool? Or would this be a manual recovery of sorts?

Thanks!

~Jay

[reezin68]

If a drive goes bad in the greyhole pool you'd lose that data on that drive only.This is why it's recommended to keep 2 copies in the pools of data you don't want to lose.So you'd still have access to that data(different drive). I'm unaware of Greyhole/Amahi giving out warning that drive going/gone bad.But it could be something that I don't know about.

Once you replace the bad drive with a new one Greyhole will automatically re-balance the data across the disks in the pools.Recreating copies on the new disk if necessary.You can do this manually also. It is always a good idea to have current backups of your data stored else where.

Info can be found here. https://github.com/gboudreau/Greyhole/wiki

[jkbuckethead]

Unfortunately, I found out first hand that amahi does not provide indication of drive errors.

I use my HDA primarily as a media server, and I noticed files would mysteriously disappear then reappear. The amahi dashboard didn't provide any error indications,. However, there were loads of errors when I connected a monitor to the server. These errors do not show when connecting by SSH, only with a physically connected monitor.

[bigfoot65]

Unfortunately, I found out first hand that amahi does not provide indication of drive errors.

That is correct. You won't find many headless servers that do provide drive monitoring by default. Usually have to add that piece manually.

We do have a feature request to add that functionality, but it's a bit complicated as a one-click app.

The Disk Error Alerts wiki guidance might help. I use it myself and also phpSysInfo with the drive monitoring piece as well.