Hi,
it would be nice if stonith-resources that only shoot one node automatically
set the anti-location constraint. But I fear you gotta do this by hand.
While an ipmi-stonith killing itself will work in that the machine dies, it
can't report back to pacemaker and then afair pacemaker hangs and
Hi.
I´m running a Pacemaker Cluster on Dell R610 machines, so I enabled the
IDrac6 and configured the following external/ipmi resources
primitive resIPMI-1 stonith:external/ipmi \
params hostname=apolo ipaddr=172.31.0.240 \
userid=root passwd=somepass interface=lan \
op mon
On Fri, 21 Sep 2012 21:16:02 +0200 Lars Marowsky-Bree
wrote:
> On 2012-09-20T10:55:59, Mia Lueng wrote:
> > There are lots of messages that shows when starting corosync, may
> > watchdog driver was tested to load into kernel.
> Load the module you want to use via initrd, or specify it in the
> mo
On 2012-09-20T10:55:59, Mia Lueng wrote:
> There are lots of messages that shows when starting corosync, may
> watchdog driver was tested to load into kernel.
Load the module you want to use via initrd, or specify it in the
modprobe.conf file.
Regards,
Lars
--
Architect Storage/HA
SUSE L
On 2012-09-19T16:35:53, Mia Lueng wrote:
> Hi All:
> I use a sbd device to be a STONITH device. Should I configure it
> as a clone resource that running on all node? I have seen a lot of
> manuals to show it only be configured on a single node. Here is my
> configure:
Do not clone it.
S
Hi all!
I've configured HA NFS storage (almost) according to the LinBIT HOWTO and,
in general it works OK for my client nodes, which mount NFS share /web as
/var/www and use it as document root for NGINX.
To get maximum performance out of the configuration two NFS exports(/web
and /img) are confi
