[ClusterLabs] Connectivity is degraded (Expected=300)

[Ken Gaillot]

On 03/01/2016 04:33 PM, Rafał Sanocki wrote:
> but my score from pingd should be 200 because my config is
> 
> node nodeA
> node nodeB
> primitive failover-ip1 ocf:pacemaker:node \
>         op monitor interval="100s" \
>         op start interval="0" timeout="2m" \
>         op stop interval="0" timeout="1m" \
>         meta target-role="Started"
> primitive my-arping ocf:pacemaker:arping \
>         params name="arping" dampen="30s" multiplier="100"
> host_list="neteth1;192.168.1.10" \
>         op monitor interval="5s"
> primitive my-pingd ocf:pacemaker:pingd \
>         params name="pingd" dampen="15s" multiplier="100"
> host_list="192.168.2.10  193.168.10.50 " interval="2"
> clone my-conn my-arping \
>         meta globally-unique="false" clone-max="2" clone-node-max="1"
> target-role="Started"
> clone my-connp my-pingd \
>         meta globally-unique="false" clone-max="2" clone-node-max="1"
> dampen="15s"
> location pref-ip1 failover-ip1 \
>         rule $id="pref-failover-ip1-rule" pingd: defined pingd \
>         rule $id="pref-failover-ip1-rule2" arping: defined arping
> property $id="cib-bootstrap-options" \
>         dc-version="1.0.11-614a4bc7ea748b7583fc7fe164eddbcdf33b03ba" \

Hi Rafał,

1.0.11 is old (2011) and no longer supported, and there have been 14
releases in the 1.1 line. I have no idea whether the pingd agent from
that works or not. In recent versions, the ocf:pacemaker:pingd agent is
obsolete, and is now just a wrapper for ocf:pacemaker:ping.

If you have any possibility of upgrading, I would strongly recommend it.
You would also need to switch from openais to corosync >=2, and libqb >=
0.17.0. Most Linux distributions have recent versions in their regular
package repositories.

>         cluster-infrastructure="openais" \
>         stonith-enabled="false" \
>         no-quorum-policy="ignore" \
>         expected-quorum-votes="2" \
>         cluster-delay="30s" \
>         symmetric-cluster="true" \
>         last-lrm-refresh="1373962109"
> rsc_defaults $id="rsc-options" \
>         resource-stickiness="0"
> op_defaults $id="op_defaults-options" \
>         timeout="20s"
> 
> 2xhost x 100 = 200 score soo why he expect 300, is that 100 from arping
> expecting?
>
> Rafał
> 
> 
> 
> 
> 
> 
> W dniu 2016-03-01 o 16:09, Ken Gaillot pisze:
>> On 03/01/2016 08:24 AM, Rafał Sanocki wrote:
>>> Hello
>>> Can you tell if that message is correct "
>>>
>>>   #crm_mon -A
>>> Online: [ nodeA nodeB ]
>>> failover-ip1    (ocf::pacemaker:wall):  Started nodeB
>>>   Clone Set: my-conn
>>>       Started: [ nodeA nodeB ]
>>>   Clone Set: my-connp
>>>       Started: [ nodeA nodeB ]
>>> Node Attributes:
>>> * Node nodeA:
>>>      + arping                            : 0
>>>      + pingd                             : 200           : Connectivity
>>> is degraded (Expected=300)
>> crm_mon prints messages like this for ping/pingd resources. The expected
>> value is the number of nodes in the resource's host_list parameter
>> multiplied by its multiplier parameter (default 1 if not specified). If
>> the actual value is less than the expected value, it will print the
>> message.
>>
>> Whether that's "correct" depends on what you mean by that. It's correct
>> in that the actual value is less than the expected value.
>>
>>> * Node nodeB:
>>>      + arping                            : 100
>>>      + pingd                             : 200           : Connectivity
>>> is degraded (Expected=300)
>>
>> _______________________________________________
>> Users mailing list: Users at clusterlabs.org
>> http://clusterlabs.org/mailman/listinfo/users
>>
>> Project Home: http://www.clusterlabs.org
>> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>> Bugs: http://bugs.clusterlabs.org
>