[DRBD-user] MySQL performance breakdown during resync to secondary node

Lars Ellenberg lars.ellenberg at linbit.com
Sat Mar 31 00:49:01 CEST 2012

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

On Tue, Mar 27, 2012 at 11:19:14AM +0200, Jakob Curdes wrote:
> Hi, we are running a primary/secondary cluster which hosts a zarafa
> database (among other things). The hardware and network (dedicated
> dual bonded GBit) should be fast enough and in fact, when the
> machines are in sync we have no performance problems whatsoever.
> However, when we need to resync the volumes, the read (!)
> performance of the MySQL database (all innodb) drops drastically. As
> this happens only on resync, I wonder where this behavior might come
> from. We already reduced the syncer rate to absurd values (like 100
> kB/s) and still get high IOWaits on the primary as soon as the
> syncing starts. At 100kbit/s I think I can rule out the disks as
> bottleneck or am I wrong there?

What I scheduler are you using? (cfq, deadline, ...)?
I suggest you use deadline.

What is the storage backend (lower level device to DRBD)?

: 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