Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
On Wed, Jun 10, 2009 at 01:58:06PM +0200, Reinier Haasjes wrote:
> > hmmm. too bad.
> > the combination of after-sb- configuration you chose
> > leads into a deadlock within the drbd state handling on ha1 :(
> > we'll fix that.
>
> What after-sb- configuration do you advise on a 2 box, multipath
> configuration (both boxes are primary and used as a primary)?
I don't.
that is definetely nothing for a "short email".
> > what would have been intended for this combination of after-sb-
> > configuration is ha1 tries to become Secondary,
> > if that succeeds it becomes normal SyncTarget.
> > if that does not succeed, it would call the lost-after-sb handler,
> > which you need to define, and which is supposed to hard reboot the box.
> >
> > because that handler is not configured, but the rr-conflict is on the
> > default disconnect, you'd then get on ha1 "I shall become SyncTarget,
> > but I am primary!", and ha1 would go StandAlone.
> >
>
> I changed my config a little (rr-conflict and pri-lost-after-sb) and
> did the test again, same results:
the "we'll fix that" above should be read as
"thank you for triggering and indicating this bug,
we think we have understand the problem and are working on it"
>
> ----------------------------------------
> root at ha2:/etc# drbdadm dump
> # /etc/drbd.conf
> common {
> syncer {
> rate 10M;
> }
> }
>
> # resource shared on ha2: not ignored, not stacked
> resource shared {
> protocol C;
> on ha2 {
> device /dev/drbd0;
> disk /dev/data/shared;
> address ipv4 192.168.0.2:7788;
> meta-disk internal;
> }
> on ha1 {
> device /dev/drbd0;
> disk /dev/data/shared;
> address ipv4 192.168.0.1:7788;
> meta-disk internal;
> }
> net {
> allow-two-primaries;
> cram-hmac-alg sha256;
> shared-secret DRBD2KVM;
> after-sb-0pri discard-least-changes;
> after-sb-1pri call-pri-lost-after-sb;
> after-sb-2pri call-pri-lost-after-sb;
> rr-conflict call-pri-lost;
pri-lost and pri-lost-after-sb are _different_,
and call _different_ handlers.
you'd need to also configure pri-lost "..."; below.
> }
> disk {
> on-io-error pass_on;
> }
> syncer {
> rate 300M;
> al-extents 257;
> }
> startup {
> degr-wfc-timeout 120;
> become-primary-on both;
> }
> handlers {
> pri-lost-after-sb "echo o > /proc/sysrq-trigger ; halt -f";
> }
> }
--
: 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