<div dir="ltr"><div><div>On 06/09/2017 07:43 AM, ashutosh tiwari wrote:</div><div>> Hi,</div><div>></div><div>> We have two node cluster(ACTIVE/STANDBY).</div><div>> Recently we moved these nodes to KVM.</div><div>></div><div>> When we create a private virtual network and use this vnet for</div><div>> assigning cluster interfaces then things work as expected and both the</div><div>> nodes are able to form the cluster.</div><div>></div><div>> Nodes are not able to form cluster when we use macvtap(bridge)</div><div>> interfaces for cluster.</div><div>></div><div>> came across this:</div><div>></div><div>> <a href="https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650">https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650</a></div><div>> <<a href="https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650">https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650</a>></div><div>></div><div><br></div><div>Seems to deal with multicast-issues...</div><div>Wouldn't using corosync with unicast be a possibility?</div><div><br></div><div>Regards,</div><div>Klaus</div></div><div><br></div><div><br></div>>>>Hi Klaus,<div><br></div><div>>>>Once sure that it can not be achieved with multicast, probably we can look at using unicast for corosync communication.<br><br></div><div>>>>Regards,</div><div>>>>Ashutosh </div></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Jun 9, 2017 at 3:30 PM, <span dir="ltr"><<a href="mailto:users-request@clusterlabs.org" target="_blank">users-request@clusterlabs.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Send Users mailing list submissions to<br>
<a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
<br>
To subscribe or unsubscribe via the World Wide Web, visit<br>
<a href="http://lists.clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://lists.clusterlabs.org/<wbr>mailman/listinfo/users</a><br>
or, via email, send a message with subject or body 'help' to<br>
<a href="mailto:users-request@clusterlabs.org">users-request@clusterlabs.org</a><br>
<br>
You can reach the person managing the list at<br>
<a href="mailto:users-owner@clusterlabs.org">users-owner@clusterlabs.org</a><br>
<br>
When replying, please edit your Subject line so it is more specific<br>
than "Re: Contents of Users digest..."<br>
<br>
<br>
Today's Topics:<br>
<br>
1. Re: Node attribute disappears when pacemaker is started<br>
(Ken Gaillot)<br>
2. Re: IPMI and APC switched PDUs fencing agents<br>
(Jean-Francois Malouin)<br>
3. cluster setup for nodes at KVM guest (ashutosh tiwari)<br>
4. Re: cluster setup for nodes at KVM guest (Klaus Wenninger)<br>
<br>
<br>
------------------------------<wbr>------------------------------<wbr>----------<br>
<br>
Message: 1<br>
Date: Thu, 8 Jun 2017 09:42:53 -0500<br>
From: Ken Gaillot <<a href="mailto:kgaillot@redhat.com">kgaillot@redhat.com</a>><br>
To: <a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
Subject: Re: [ClusterLabs] Node attribute disappears when pacemaker is<br>
started<br>
Message-ID: <<a href="mailto:5db7ebbe-be2a-c626-932e-27b56969b045@redhat.com">5db7ebbe-be2a-c626-932e-<wbr>27b56969b045@redhat.com</a>><br>
Content-Type: text/plain; charset=utf-8<br>
<br>
Hi,<br>
<br>
Looking at the incident around May 26 16:40:00, here is what happens:<br>
<br>
You are setting the attribute for rhel73-2 from rhel73-1, while rhel73-2<br>
is not part of cluster from rhel73-1's point of view.<br>
<br>
The crm shell sets the node attribute for rhel73-2 with a CIB<br>
modification that starts like this:<br>
<br>
++ /cib/configuration/nodes: <node uname="rhel73-2" id="rhel73-2"/><br>
<br>
Note that the node ID is the same as its name. The CIB accepts the<br>
change (because you might be adding the proper node later). The crmd<br>
knows that this is not currently valid:<br>
<br>
May 26 16:39:39 rhel73-1 crmd[2908]: error: Invalid node id: rhel73-2<br>
<br>
When rhel73-2 joins the cluster, rhel73-1 learns its node ID, and it<br>
removes the existing (invalid) rhel73-2 entry, including its attributes,<br>
because it assumes that the entry is for an older node that has been<br>
removed.<br>
<br>
I believe attributes can be set for a node that's not in the cluster<br>
only if the node IDs are specified explicitly in corosync.conf.<br>
<br>
You may want to mention the issue to the crm shell developers. It should<br>
probably at least warn if the node isn't known.<br>
<br>
<br>
On 05/31/2017 09:35 PM, ?? ?? wrote:<br>
> Hi Ken,<br>
><br>
> I'm sorry. Attachment size was too large.<br>
> I attached it to GitHub, so look at it.<br>
> <a href="https://github.com/inouekazu/pcmk_report/blob/master/pcmk-Fri-26-May-2017.tar.bz2" rel="noreferrer" target="_blank">https://github.com/inouekazu/<wbr>pcmk_report/blob/master/pcmk-<wbr>Fri-26-May-2017.tar.bz2</a><br>
><br>
>> -----Original Message-----<br>
>> From: Ken Gaillot [mailto:<a href="mailto:kgaillot@redhat.com">kgaillot@redhat.com</a>]<br>
>> Sent: Thursday, June 01, 2017 8:43 AM<br>
>> To: <a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
>> Subject: Re: [ClusterLabs] Node attribute disappears when pacemaker is started<br>
>><br>
>> On 05/26/2017 03:21 AM, ?? ?? wrote:<br>
>>> Hi Ken,<br>
>>><br>
>>> I got crm_report.<br>
>>><br>
>>> Regards,<br>
>>> Kazunori INOUE<br>
>><br>
>> I don't think it attached -- my mail client says it's 0 bytes.<br>
>><br>
>>>> -----Original Message-----<br>
>>>> From: Ken Gaillot [mailto:<a href="mailto:kgaillot@redhat.com">kgaillot@redhat.com</a>]<br>
>>>> Sent: Friday, May 26, 2017 4:23 AM<br>
>>>> To: <a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
>>>> Subject: Re: [ClusterLabs] Node attribute disappears when pacemaker is started<br>
>>>><br>
>>>> On 05/24/2017 05:13 AM, ?? ?? wrote:<br>
>>>>> Hi,<br>
>>>>><br>
>>>>> After loading the node attribute, when I start pacemaker of that node, the attribute disappears.<br>
>>>>><br>
>>>>> 1. Start pacemaker on node1.<br>
>>>>> 2. Load configure containing node attribute of node2.<br>
>>>>> (I use multicast addresses in corosync, so did not set "nodelist {nodeid: }" in corosync.conf.)<br>
>>>>> 3. Start pacemaker on node2, the node attribute that should have been load disappears.<br>
>>>>> Is this specifications ?<br>
>>>><br>
>>>> Hi,<br>
>>>><br>
>>>> No, this should not happen for a permanent node attribute.<br>
>>>><br>
>>>> Transient node attributes (status-attr in crm shell) are erased when the<br>
>>>> node starts, so it would be expected in that case.<br>
>>>><br>
>>>> I haven't been able to reproduce this with a permanent node attribute.<br>
>>>> Can you attach logs from both nodes around the time node2 is started?<br>
>>>><br>
>>>>><br>
>>>>> 1.<br>
>>>>> [root@rhel73-1 ~]# systemctl start corosync;systemctl start pacemaker<br>
>>>>> [root@rhel73-1 ~]# crm configure show<br>
>>>>> node 3232261507: rhel73-1<br>
>>>>> property cib-bootstrap-options: \<br>
>>>>> have-watchdog=false \<br>
>>>>> dc-version=1.1.17-0.1.rc2.el7-<wbr>524251c \<br>
>>>>> cluster-infrastructure=<wbr>corosync<br>
>>>>><br>
>>>>> 2.<br>
>>>>> [root@rhel73-1 ~]# cat rhel73-2.crm<br>
>>>>> node rhel73-2 \<br>
>>>>> utilization capacity="2" \<br>
>>>>> attributes attrname="attr2"<br>
>>>>><br>
>>>>> [root@rhel73-1 ~]# crm configure load update rhel73-2.crm<br>
>>>>> [root@rhel73-1 ~]# crm configure show<br>
>>>>> node 3232261507: rhel73-1<br>
>>>>> node rhel73-2 \<br>
>>>>> utilization capacity=2 \<br>
>>>>> attributes attrname=attr2<br>
>>>>> property cib-bootstrap-options: \<br>
>>>>> have-watchdog=false \<br>
>>>>> dc-version=1.1.17-0.1.rc2.el7-<wbr>524251c \<br>
>>>>> cluster-infrastructure=<wbr>corosync<br>
>>>>><br>
>>>>> 3.<br>
>>>>> [root@rhel73-1 ~]# ssh rhel73-2 'systemctl start corosync;systemctl start pacemaker'<br>
>>>>> [root@rhel73-1 ~]# crm configure show<br>
>>>>> node 3232261507: rhel73-1<br>
>>>>> node 3232261508: rhel73-2<br>
>>>>> property cib-bootstrap-options: \<br>
>>>>> have-watchdog=false \<br>
>>>>> dc-version=1.1.17-0.1.rc2.el7-<wbr>524251c \<br>
>>>>> cluster-infrastructure=<wbr>corosync<br>
>>>>><br>
>>>>> Regards,<br>
>>>>> Kazunori INOUE<br>
<br>
<br>
<br>
------------------------------<br>
<br>
Message: 2<br>
Date: Thu, 8 Jun 2017 16:40:59 -0400<br>
From: Jean-Francois Malouin <<a href="mailto:Jean-Francois.Malouin@bic.mni.mcgill.ca">Jean-Francois.Malouin@bic.<wbr>mni.mcgill.ca</a>><br>
To: Cluster Labs - All topics related to open-source clustering<br>
welcomed <<a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a>><br>
Subject: Re: [ClusterLabs] IPMI and APC switched PDUs fencing agents<br>
Message-ID: <<a href="mailto:20170608204059.GC11019@bic.mni.mcgill.ca">20170608204059.GC11019@bic.<wbr>mni.mcgill.ca</a>><br>
Content-Type: text/plain; charset=us-ascii<br>
<br>
* Jean-Francois Malouin <<a href="mailto:Jean-Francois.Malouin@bic.mni.mcgill.ca">Jean-Francois.Malouin@bic.<wbr>mni.mcgill.ca</a>> [20170607 13:09]:<br>
> Hi,<br>
<br>
..snip...<br>
<br>
> I'm having some difficulties understanding the fencing_topology syntax.<br>
> Making the changes adapted for my local configuration I get the error<br>
> when trying to add the fence topology:<br>
><br>
> ~# crm configure fencing_topology antonio: fence_antonio_ipmi fence_antonio_psu1_off,fence_<wbr>antonio_psu2_off,fence_<wbr>antonio_psu1_on,fence_antonio_<wbr>psu2_on \<br>
> leonato: fence_leonato_ipmi fence_leonato_psu1_off,fence_<wbr>leonato_psu2_off,fence_<wbr>leonato_psu1_on,fence_leonato_<wbr>psu2_on<br>
> WARNING: fencing_topology: target antonio not a node<br>
> WARNING: fencing_topology: target antonio not a node<br>
> WARNING: fencing_topology: target leonato not a node<br>
> WARNING: fencing_topology: target leonato not a node<br>
<br>
In retrospect I should have used the long (fqdn) of the nodes rather<br>
than the short names, which as reportd, are not node names.<br>
<br>
> What should I use for the <node> in:<br>
><br>
> fencing_topology <stonith_resources> [<stonith_resources> ...]<br>
> fencing_topology <fencing_order> [<fencing_order> ...]<br>
><br>
> fencing_order :: <target> <stonith_resources> [<stonith_resources> ...]<br>
><br>
> stonith_resources :: <rsc>[,<rsc>...]<br>
> target :: <node>: | attr:<node-attribute>=<value><br>
<br>
I tried to use the node names (fqdn) when adding the fence_topology<br>
resource to the cib but I always got an error from crm telling me of an<br>
invalid DTD/schema. In the end I had to add a 'dummy' attributes to each<br>
nodes (dummy=<node_name>) and use them to create the fence levels, as<br>
in:<br>
<br>
crm configure fencing_topology \<br>
attr:dummy=node1 fence_node1_ipmi fence_node1_psu1_off,fence_<wbr>node1_psu2_off,fence_node1_<wbr>psu1_on,fence_node1_psu2_on \<br>
attr:dummy=node2 fence_node2_ipmi fence_node2_psu1_off,fence_<wbr>node2_psu2_off,fence_node2_<wbr>psu1_on,fence_node2_psu2_on<br>
<br>
Is this the expected behaviour?<br>
<br>
Thanks,<br>
jf<br>
<br>
<br>
<br>
------------------------------<br>
<br>
Message: 3<br>
Date: Fri, 9 Jun 2017 11:13:36 +0530<br>
From: ashutosh tiwari <<a href="mailto:ashutosh.kvas@gmail.com">ashutosh.kvas@gmail.com</a>><br>
To: <a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
Subject: [ClusterLabs] cluster setup for nodes at KVM guest<br>
Message-ID:<br>
<<a href="mailto:CA%2BvEgjj1a7gKYReY-YcFootNHmkAzGJf_3XHxJehW2S0vMrgaA@mail.gmail.com">CA+vEgjj1a7gKYReY-<wbr>YcFootNHmkAzGJf_<wbr>3XHxJehW2S0vMrgaA@mail.gmail.<wbr>com</a>><br>
Content-Type: text/plain; charset="utf-8"<br>
<br>
Hi,<br>
<br>
We have two node cluster(ACTIVE/STANDBY).<br>
Recently we moved these nodes to KVM.<br>
<br>
When we create a private virtual network and use this vnet for assigning<br>
cluster interfaces then things work as expected and both the nodes are able<br>
to form the cluster.<br>
<br>
Nodes are not able to form cluster when we use macvtap(bridge) interfaces<br>
for cluster.<br>
<br>
came across this:<br>
<br>
<a href="https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650" rel="noreferrer" target="_blank">https://github.com/vagrant-<wbr>libvirt/vagrant-libvirt/<wbr>issues/650</a><br>
<br>
tried the suggested workaround in the threadbut(trustGuetRxFilters='<wbr>yes')<br>
but of no help.<br>
<br>
<br>
Thanks and Regards,<br>
Ashutosh<br>
-------------- next part --------------<br>
An HTML attachment was scrubbed...<br>
URL: <<a href="http://lists.clusterlabs.org/pipermail/users/attachments/20170609/2305453f/attachment-0001.html" rel="noreferrer" target="_blank">http://lists.clusterlabs.org/<wbr>pipermail/users/attachments/<wbr>20170609/2305453f/attachment-<wbr>0001.html</a>><br>
<br>
------------------------------<br>
<br>
Message: 4<br>
Date: Fri, 9 Jun 2017 07:49:48 +0200<br>
From: Klaus Wenninger <<a href="mailto:kwenning@redhat.com">kwenning@redhat.com</a>><br>
To: <a href="mailto:users@clusterlabs.org">users@clusterlabs.org</a><br>
Subject: Re: [ClusterLabs] cluster setup for nodes at KVM guest<br>
Message-ID: <<a href="mailto:bcaac851-2c1f-3204-2839-9145a1b2acae@redhat.com">bcaac851-2c1f-3204-2839-<wbr>9145a1b2acae@redhat.com</a>><br>
Content-Type: text/plain; charset=utf-8<br>
<br>
On 06/09/2017 07:43 AM, ashutosh tiwari wrote:<br>
> Hi,<br>
><br>
> We have two node cluster(ACTIVE/STANDBY).<br>
> Recently we moved these nodes to KVM.<br>
><br>
> When we create a private virtual network and use this vnet for<br>
> assigning cluster interfaces then things work as expected and both the<br>
> nodes are able to form the cluster.<br>
><br>
> Nodes are not able to form cluster when we use macvtap(bridge)<br>
> interfaces for cluster.<br>
><br>
> came across this:<br>
><br>
> <a href="https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650" rel="noreferrer" target="_blank">https://github.com/vagrant-<wbr>libvirt/vagrant-libvirt/<wbr>issues/650</a><br>
> <<a href="https://github.com/vagrant-libvirt/vagrant-libvirt/issues/650" rel="noreferrer" target="_blank">https://github.com/vagrant-<wbr>libvirt/vagrant-libvirt/<wbr>issues/650</a>><br>
><br>
<br>
Seems to deal with multicast-issues...<br>
Wouldn't using corosync with unicast be a possibility?<br>
<br>
Regards,<br>
Klaus<br>
<br>
> tried the suggested workaround in the<br>
> threadbut(trustGuetRxFilters='<wbr>yes') but of no help.<br>
><br>
><br>
> Thanks and Regards,<br>
> Ashutosh<br>
><br>
><br>
><br>
> ______________________________<wbr>_________________<br>
> Users mailing list: <a href="mailto:Users@clusterlabs.org">Users@clusterlabs.org</a><br>
> <a href="http://lists.clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://lists.clusterlabs.org/<wbr>mailman/listinfo/users</a><br>
><br>
> Project Home: <a href="http://www.clusterlabs.org" rel="noreferrer" target="_blank">http://www.clusterlabs.org</a><br>
> Getting started: <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/<wbr>doc/Cluster_from_Scratch.pdf</a><br>
> Bugs: <a href="http://bugs.clusterlabs.org" rel="noreferrer" target="_blank">http://bugs.clusterlabs.org</a><br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
------------------------------<br>
<br>
______________________________<wbr>_________________<br>
Users mailing list<br>
<a href="mailto:Users@clusterlabs.org">Users@clusterlabs.org</a><br>
<a href="http://lists.clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://lists.clusterlabs.org/<wbr>mailman/listinfo/users</a><br>
<br>
<br>
End of Users Digest, Vol 29, Issue 10<br>
******************************<wbr>*******<br>
</blockquote></div><br></div>