[Linux-HA] pingd does not work for connection loss

[Yann Zurcher]

Hi,

I had some other priorities the last couple of days...
But now I am back and still have the same problem. At least I figuered out,
that there seems to be a misconfiguration of the pingd, but I'm not able to
figur out, where it is.
I attached the (correct :-) )log.
Some of the configuration changed since my last post, but the main stuff in
HA didn't change.
- ha80/ha81 --> openint1/openint2
- IP-addresses changed

I found some wired info in the logfile:

 Aug 16 15:16:13 openint1 lrmd: [6010]: info: RA output:
(pingd:0:start:stdout) This account is currently not available.

Aug 16 15:16:13 openint1 lrmd: [6010]: info: Resource Agent (operation
start[49] on ocf::pingd::pingd:0 for client 6013, its parameters:
multiplier=[100] CRM_meta_prereq=[nothing] CRM_meta_op_target_rc=[7]
dampen=[5s] CRM_meta_id=[pingd_start] CRM_meta_timeout=[5000]
crm_feature_set=[1.0.6] user=[hacluster] CRM_me): pid [7192] exited with
return code 1

Aug 16 15:16:13 openint1 lrmd: [6010]: info: Resource Agent output: []

Aug 16 15:16:13 openint1 crmd: [6013]: WARN: process_lrm_event:lrm.c LRM
operation (49) start_0 on pingd:0 Error: (1) unknown error

What kind of Account do I need for pingd? Unknown Error confuses me...?!

Thanks,
Yann


On 8/9/06, Andrew Beekhof <beekhof at gmail.com> wrote:
>
> I'll know more once you sort out the logging.
> Its not 100% obvious, but you're attaching the wrong files :-)
>
> If you specify: use_logd on
> Then these entries have no effect:
>    logfile         /var/log/ha_log
>    debugfile       /var/log/ha_dlog
>
> However because ha.cf is parsed sequentially, you'll still get *some*
> logs in those files before it switches over to ha_logd.  Hence the
> confusion.
>
> http://linux-ha.org/ha.cf#use_logd  - you've got case 1b.
> Quote: If the logging daemon is used, logfile/debugfile/logfacility in
> this file are not meaningful any longer. You should check the config
> file for logging daemon (the default is /etc/logd.cf).
>