[DRBD-user] Concurrent remote write detected?

drbd at bobich.net drbd at bobich.net
Mon Mar 3 18:28:23 CET 2008

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


On Mon, 3 Mar 2008, drbd at bobich.net wrote:

> Using DRBD 8.0.8, the performance is quite acceptable in single-node case.

[...]

> I tried 8.0.11, and the performance of that seems to be very poor. So poor 
> that I've not yet had the patience to wait for the first node to boot. (I'm 
> using Open Shared Root on this, so my DRBD file system is actually the root 
> file system.) So, for now, I'm back to using 8.0.8.

Another thing worthy of note is that when using 8.0.8, the part in 
rc.sysinit that mounts the udef file system just completes - in 2 seconds 
or so. When using 8.0.11 (no other changes, just updated the rpms and 
re-made the initrd with the new packages), it fails / times out.

Since nothing else on the setup changes inbetween these attempts (I 
upgraded forward and backward a few times now just to make sure), it is a 
consistent problem. 8.0.8 works, 8.0.11 doesn't. Nothing of note in the 
logs, and /proc/drbd seems to be showing nothing wrong when the node 
actually finishes booting (which takes a lot of patience). The peer node 
is powered off and physically disconnected, so unless it's constantly 
trying to talk to it and timing out, I can't see why this is happening. 
8.0.8 certainly doesn't do this.

Gordan



More information about the drbd-user mailing list