[DRBD-user] DRBD write throughput very very bad

Lars Ellenberg lars.ellenberg at linbit.com
Fri Apr 24 11:49:02 CEST 2009

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


On Fri, Apr 24, 2009 at 11:01:00AM +0200, Tobias Appel wrote:
> Lars Ellenberg wrote:
> 
> > the users guide is always for the current release of DRBD,
> > which is 8.3.1 at the time of writing this email.
> > 
> > the "wo:" thing is only available in 8.3 (since 8.2.something) it was
> > introduced when the barrier method has been introduced, at the same time
> > as the explicit "drain" and "none" methods have been introduced.
> > 
> > drbd 8.0 has only the concept of drain,
> > and since 8.0.9 the "flush" concept.
> > in 8.0 you can only disable the "flush",
> > which means you get "drain".
> > 
> > the documentation is correct, though maybe misleading.
> > drain was the only concept implemented before 8.0.9.
> > that does not mean all other methods have been introduced with 8.0.9.
> > e.g. options to disable the flush method again have only been
> > introduced with 8.0.12.  (as written in the changelog)
> > 
> 
> Thanks for the clarification. I will try to upgrade then to 8.3.1. Wish 
> me luck!

I don't think you need it ;)

ah. and btw. do not use "none",
that is to say, do NOT use "no-disk-drain".  in general, that setting
will violate write ordering constraints, and gives very little benefit.
it is useful only in very special circumstances.

-- 
: Lars Ellenberg
: LINBIT | Your Way to High Availability
: DRBD/HA support and consulting http://www.linbit.com

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.
__
please don't Cc me, but send to list   --   I'm subscribed



More information about the drbd-user mailing list