[DRBD-user] Discarding device blocks very slow

Robert Altnoeder robert.altnoeder at linbit.com
Mon Jul 31 11:16:56 CEST 2017

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

On 07/31/2017 10:46 AM, Eric Robinson wrote:
> But that still does not explain why discards across the network are so slow.

There are lots of possible reasons.
Discard might not be enabled on both nodes, there should be a message
about discard being available or unavailable in the logs. IIRC, if it
can't discard, it will zero-out.
The network may have too high latency or too low bandwidth.
Without a great amount of detailed information all one can do about
performance-related issues is to guess.

Generally, both nodes should be running the same and most recent version
of DRBD 9, because all those features have been added quite recently,
and lots of enhancements and fixes are still applied to the code.

Robert Altnoeder
+43 1 817 82 92 0
robert.altnoeder at linbit.com

LINBIT | Keeping The Digital World Running
DRBD - Corosync - Pacemaker
f /  t /  in /  g+

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.

More information about the drbd-user mailing list