rashbre central: drive failure on backup day?

Wednesday 4 April 2018

drive failure on backup day?


Someone said it was 'backup day' or 'backup week' or something similar. Co-incidentally, I had a drive fail in my backup disk array. The picture shows the disk out of its normal NAS enclosure and being examined in one of those hard drive docking stations.

I've already replaced it with a fresh disk in the Drobo and didn't even need to take the server offline.

The last 6 Terabyte drive failed about a year ago, and was in the same unit, but a different slot.

I decided to look at mean time between failure on these devices. I've quite a few Western Digital RED drives for servers and have a pretty good experience with them. Mine are mainly 3 Terabyte (15T packs) and 6 Terabyte (30T packs). By the time they have a single or dual disk redundancy, the available space diminishes somewhat.

It still seems crazy large compared with my oldest disk computers which had something like 2x32Mb disk drives.

Nowadays, I should probably be storing all of that data in the cloud (it's not like I need it all at once), although I'm still somewhat nervous about the way various services just come to an end. A recent example is the termination of the Amazon cloud storage of personal music libraries, which isn't allowing any new things to be uploaded after the end of 2018.

I like the stats that backblaze keep on disk failure rates, from their server farms, as a quick way to get a sense of relative failure rates, from a population of about 90,000 drives.

No comments: