Hi,

On Tue, May 06, 2008 at 07:40:41PM +0400, Alex Orlov wrote:
> stonithd[18930]: 2008/05/06_19:22:37 info: Node xnfs1 try to help node xnfs2 
> to fence node xnfs2.
> 
> 8) 
> 
> its fantastic.  kills himself 
> 
> 
>        <primitive class="stonith" type="external/riloe" provider="heartbeat" 
> id="ST1">
>          <meta_attributes id="ST1_meta_attrs">
>            <attributes>
>              <nvpair id="ST1_metaattr_target_role" name="target_role" 
> value="started"/>
>            </attributes>
>          </meta_attributes>
>          <instance_attributes id="ST1_instance_attrs">
>            <attributes>
>              <nvpair id="5bbeadf5-1c9f-42bb-a0a7-841033cc5eb2" 
> name="hostlist" value="xnfs2"/>
>              <nvpair id="90a68bcf-b577-496e-973a-ef854f968606" 
> name="ilo_hostname" value="xnfs2-rm"/>
>              <nvpair id="41f49b53-d6eb-4646-bc4d-5a53f7d0015b" 
> name="ilo_user" value="Administrator"/>
>              <nvpair id="335cb1ac-ae5b-4590-90ad-1382d0127c1e" 
> name="ilo_password" value="Administrator"/>
>              <nvpair id="c5ec172e-fe27-4581-8c5a-ad475fcf3aa7" 
> name="ilo_can_reset" value="1"/>
>              <nvpair id="0cfa50f1-5e91-4de1-9a78-dbef0626ff51" 
> name="ilo_protocol" value="2.0"/>
>              <nvpair id="1d09d1e5-0d02-4a76-b470-f8e0247a0055" 
> name="ilo_powerdown_method" value="button"/>
>            </attributes>
>          </instance_attributes>
>          <operations>
>            <op name="monitor" timeout="15" start_delay="0" disabled="false" 
> role="Started" id="13de3090-1207-4f8f-9f4a-308929cecd10" interval="30"/>
>            <op id="370a83e7-5052-4c01-a382-3395bce8a72f" name="start" 
> timeout="15" start_delay="0" disabled="false" role="Started"/>
>          </operations>
>        </primitive>
>        <primitive id="ST2" class="stonith" type="external/riloe" 
> provider="heartbeat">
>          <meta_attributes id="ST2_meta_attrs">
>            <attributes>
>              <nvpair id="ST2_metaattr_target_role" name="target_role" 
> value="started"/>
>            </attributes>
>          </meta_attributes>
>          <instance_attributes id="ST2_instance_attrs">
>            <attributes>
>              <nvpair id="064e2267-a857-4bc3-9549-d7924ff6c4e0" 
> name="hostlist" value="xnfs1"/>
>              <nvpair id="dc307275-92c6-474c-9fc0-7e4d8de13170" 
> name="ilo_hostname" value="xnfs1-rm"/>
>              <nvpair id="071b8f84-edbd-4016-bab4-a4f969e625c1" 
> name="ilo_user" value="Administrator"/>
>              <nvpair id="6cca8058-973e-4320-aeb7-260f27eb77e4" 
> name="ilo_password" value="Administrator"/>
>              <nvpair id="1eadd314-da6d-4058-a2c4-c23b597558c5" 
> name="ilo_can_reset" value="1"/>
>              <nvpair id="285c5feb-2b15-4c76-8f9b-642c9d24d63f" 
> name="ilo_protocol" value="2.0"/>
>              <nvpair id="9291eb7c-32e1-4c94-99a1-c3987f0ecf75" 
> name="ilo_powerdown_method" value="button"/>
>            </attributes>
>          </instance_attributes>
>          <operations>
>            <op id="56ace93d-b669-4452-8b85-3da9ba03f053" name="monitor" 
> interval="30" timeout="15" start_delay="0" disabled="false" role="Started"/>
>            <op id="73694b16-e80f-4bcb-a6c7-6db5b35a69e2" name="start" 
> timeout="15" start_delay="0" disabled="false" role="Started"/>
>          </operations>
>        </primitive>
> 
> 
>      <constraints>
>        <rsc_location id="location_st1" rsc="ST1">
>          <rule id="prefered_location_st1" score="INFINITY">
>            <expression attribute="#uname" 
> id="6362da76-af09-403e-8a84-cdc5549c2856" operation="eq" value="xnfs1"/>
>          </rule>
>        </rsc_location>
>        <rsc_location id="location_st2" rsc="ST2">
>          <rule id="prefered_location_st2" score="INFINITY">
>            <expression attribute="#uname" 
> id="1274a005-5612-4123-85f2-da7035ef518b" operation="eq" value="xnfs2"/>
>          </rule>
>        </rsc_location>
>      </constraints>
> 
> Where a mistake?

Can't say without logs. Please use hb_report to capture the
incident and send the resulting tarball.

Thanks,

Dejan

> -- Dejan Muhamedagic, ??????? 06 ??? 2008:
> > Hi,
> > 
> > On Tue, May 06, 2008 at 06:27:44PM +0400, Alex Orlov wrote:
> > > thnx.
> > > I clear config. Make new, old be from v2.0.8 ...
> > > and all ok.
> > > 
> > > but, if i shutdown eth(heartbeat) on active node - passive goes to 
> > > reboot. 
> > > after boot it not see other node and startup all resources... split 
> > > brain...
> > 
> > Fence unseen nodes? Set startup-fencing to true in crm_config.
> > 
> > Thanks,
> > 
> > Dejan
> > 
> > > 
> > > 
> > > -- Dejan Muhamedagic, ??????? 06 ??? 2008:
> > > > Hi,
> > > > 
> > > > This looks OK to me. You can also try "off" instead of
> > > > "reset"--that's what you have in the CIB.
> > > > 
> > > 
> > > 
> > > _______________________________________________
> > > Linux-HA mailing list
> > > [email protected]
> > > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > > See also: http://linux-ha.org/ReportingProblems
> > _______________________________________________
> > Linux-HA mailing list
> > [email protected]
> > http://lists.linux-ha.org/mailman/listinfo/linux-ha
> > See also: http://linux-ha.org/ReportingProblems
> > 
> 
> 
> _______________________________________________
> Linux-HA mailing list
> [email protected]
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to