[DRBD-user] database on drbd

Lars Ellenberg Lars.Ellenberg at linbit.com
Wed Mar 10 16:12:54 CET 2004


/ 2004-03-08 16:59:59 -0500
\ Carson Gaspar:

> >Using "B" does currently not gain anything, but degrades
> >throughput, and loses transactional semantics, so should be avoided.
> 
> Perhaps a note to that effect can be added to the docs? We went with B here 
> because if both machines went down at once, we were hosed anyway, and 
> wanted better performance. *sigh* Time for another maintenance window (and 
> I just upgraded everything yesterday!)

Hm. It is, is it not?
from the scripts/drbd.conf example config file:

  # transfer protocol to use.
  # C: write IO is reported as completed, if we know it has
  #    reached _both_ local and remote DISK.
  #    * for most cases.
  #    * the only protocol with true transactional semantics.
  # B: write IO is reported as completed, if it has reached
  #    local DISK and remote buffer cache.
  #    * should perform better than C,
  #      but according to benchmarks, does not.
  #      so use C instead for now.
  # A: write IO is reported as completed, if it has reached
  #    local DISK and local tcp send buffer. (see also sndbuf-size)
  #    * for high latency networks
  protocol = C

Yes, I know in previous releases (up to ? I forgot) "B" was
recommended "for most applications".
And btw, benchmarks are just that, and if you do performance tests
yourself, this might be different for you.

	Lars Ellenberg



More information about the drbd-user mailing list