Corrupted sizes in cache once again

[Marc]

> 
> Maybe I was a bit unclear: I have about 1000 error messages per day from
> random accounts (about 500 in total so far) on all clusters. These are
> transparent to the user, so it's more like background noise at the
> moment.

Do you have ecc memory?

> No VM involved. All machines are baremetal DRBD two-node clusters.

How old are your drives?
Do you scrub the raids? 
How reliable is your drbd setup? Does drbd even sync raid fixes?
Do you have networking issues?

Connect a hdd directly and use that for a few accounts, do you still have the problem? I will bet not.

Why do you have this bare metal? Do you need performance? Otherwise switch to reliable storage that is a bit less performing. You will get headaches from these multiple drbd setups.

> As far as I see it I can not nail it down to specific accounts, POP3 vs.
> IMAP, LMTP delivery vs. IMAP store or Sieve vs. non-Sieve etc.
> 

It is not dovecot, otherwise it would be here more often listed.