[Gluster-devel] Bit-rot functionality

[Fred van Zwieten]

Hi there,

I am working for a customer right now who is considering Red Hat Storage Server. One of the sought-after features there is bit-rot detection, and even better, (semi-automatic) bit-rot restoration.

I know RHSS (nor GlusterFS) has this kind of functionality at the moment (correct me if i'm wrong!), but I would like to propose a design for this as a sort of translator that can be stacked on i.e. a (geo) replication translator.

Bit-rot detection can be done through check-summing. It should be a very low priority job running on one of the bricks. The job walks the complete file system and, per file, calculates the check-sum, compares it with the stored check-sum (if present, otherwise it stores the check-sum on all involved bricks, because it hasn't been checked before).

Bit-rot restoration could be implemented by comparing the check-sums of the replicas. If there is a mismatch, a more thorough check must be performed, like running a check-sum on all replica's for that file again, do a bit-wise compare, or whatever. If the files are still the same, the check-sum(s) must be replaced. If not, there is actual bit-rot detected. Now what to do? Which replica holds the clean version (the thruth?). With an uneven number of replicas one could simply make it a democratic process and have it fully automated. It should however save the to be replaced version in a separate store and notify the admin for verification. Another method would be to just notify the admin and do nothing.

The obvious place to store the check-sums would be in the extended attributes, but one could use a database for it.

I have watch the presentation Red Hat Summit 2012 - A Deep Dive Into Red Hat Storage by Jeff Darcy and I know he (and Red Hat) are very keen on extending the number of translators with useful functionality. I am no programmer myself, but would like to get involved in this kind of stuff.

Comments are very welcome!

Fred

PS (this was already posted on the user list, but I was advised to post it on the devel list)