[TKLBAM] restore errors say "timeout" when actually caused by full disk

[JedMeister]

As per Frank's post on the forums; tklbam-restore falsely claims that the error downloading is caused by a timeout; when in fact is is because the cache location (default /tmp) is full.

It would be really good if TKLBAM checked on remaining disk size and errored accordingly (i.e. insufficient space remaining) rather than providing misleading info.

Perhaps this could also tie into a configurable tklbam-restore tmp/caching location? (as per #179 for tklbam-backup)

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

[lirazsiri]

Since Squid is doing the caching what's happening is that from TKLBAM's perspect there really is a timeout. We could add some logic to monitor diskspace usage and warn the user about that.

In the meantime, we can configure how large the cache can get with the --restore-cache-size option.
Also, the logcaion restore cache is already configurable via the --restore-cache-dir= tklbam-restore option.

[JedMeister]

Looks like we'll need to update the TKLBAM website docs e.g. tklbam-restore docs on website vs tklbam-restore docs on github.