[drbd-mc] another apiauth needed?

Jerry Amundson jamundso at gmail.com
Tue Dec 7 21:28:47 CET 2010


On Tue, Dec 7, 2010 at 11:41 AM, Rasto Levrinc <rasto.levrinc at linbit.com> wrote:
> On Sun, December 5, 2010 3:00 pm, Lars Ellenberg wrote:
>> On Fri, Dec 03, 2010 at 11:55:41AM -0600, Jerry Amundson wrote:
>>
>>> Fyi...
>>> I was getting this message logged every second...
>>> ERROR: te_connect_stonith: Sign-in failed: triggered a retry
>>>
>>> It went away once I added
>>> apiauth stonith-ng uid=root to ha.cf.
>>
>> That's implicit with heartbeat 3.0.4.
>> It does not affect pacemaker 1.0 users,
>> only 1.1.x/1.2 with stonith-ng.
>
> Hmm, since combination Hb 3.0.3 and Pacemaker > 1.1.x will be quite rare
> in the future, I'll not clutter the config with it, unless this
> combination of versions will be popular as it sometimes happens. :)
>
> Rasto

Well, you used Heartbeat in your screencast "DRBD cluster in 10
minutes", which made it "popular" to me. :)

Mainly, though, the original problem is kind of intense, and the large
messages file hinders other troubleshooting.

I'm willing to switch to Corosync, if that's a better fit for two
Fedora 14 nodes.

thanks,
jerry


More information about the drbd-mc mailing list