<div dir="ltr"><div>Hello Jan,</div><div><br></div><div dir="ltr">>Please block both input and output. Corosync isn't able to handle <br>>byzantine faults. </div><div dir="ltr"><br></div><div>Thanks. It results in clean partition if I block both outgoing and incoming udp traffic to and from a given node.</div><div><br></div><div>However, could you suggest me what is the best way to handle any real world production scenarios that may result in just one way traffic loss ?</div><div><br></div><div>Thanks again.</div><div>Prasad </div><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Tue, Apr 30, 2019 at 5:26 PM Jan Friesse <<a href="mailto:jfriesse@redhat.com">jfriesse@redhat.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Prasad,<br>
<br>
> Hello :<br>
> <br>
> I have a 3 node corosync and pacemaker cluster and the nodes are:<br>
> Online: [ SG-azfw2-189 SG-azfw2-190 SG-azfw2-191 ]<br>
> <br>
> Full list of resources:<br>
> <br>
> Master/Slave Set: ms_mysql [p_mysql]<br>
> Masters: [ SG-azfw2-189 ]<br>
> Slaves: [ SG-azfw2-190 SG-azfw2-191 ]<br>
> <br>
> For my network partition test, I created a firewall rule on Node<br>
> SG-azfw2-190 to block all incoming udp traffic from node SG-azfw2-189<br>
> /sbin/iptables -I INPUT -p udp -s 172.19.0.13 -j DROP<br>
<br>
Please block both input and output. Corosync isn't able to handle <br>
byzantine faults.<br>
<br>
Honza<br>
<br>
> <br>
> I dont think corosync is correctly detecting the partition as I am getting<br>
> different membership information from different nodes.<br>
> On node SG-azfw2-189, I still see the members as:<br>
> <br>
> Online: [ SG-azfw2-189 SG-azfw2-190 SG-azfw2-191 ]<br>
> <br>
> Full list of resources:<br>
> <br>
> Master/Slave Set: ms_mysql [p_mysql]<br>
> Masters: [ SG-azfw2-189 ]<br>
> Slaves: [ SG-azfw2-190 SG-azfw2-191 ]<br>
> <br>
> whereas, on the node SG-azfw2-190, I see membership as<br>
> <br>
> Online: [ SG-azfw2-190 SG-azfw2-191 ]<br>
> OFFLINE: [ SG-azfw2-189 ]<br>
> <br>
> Full list of resources:<br>
> <br>
> Master/Slave Set: ms_mysql [p_mysql]<br>
> Slaves: [ SG-azfw2-190 SG-azfw2-191 ]<br>
> Stopped: [ SG-azfw2-189 ]<br>
> <br>
> I expected that on node SG-azfw2-189, it should have detected that other 2<br>
> nodes have left. In the corosync logs for this node, I continuously see the<br>
> below messages:<br>
> Apr 30 11:00:03 corosync [TOTEM ] entering GATHER state from 4.<br>
> Apr 30 11:00:03 corosync [TOTEM ] Creating commit token because I am the<br>
> rep.<br>
> Apr 30 11:00:03 corosync [MAIN ] Storing new sequence id for ring 2e64<br>
> Apr 30 11:00:03 corosync [TOTEM ] entering COMMIT state.<br>
> Apr 30 11:00:33 corosync [TOTEM ] The token was lost in the COMMIT state.<br>
> Apr 30 11:00:33 corosync [TOTEM ] entering GATHER state from 4.<br>
> Apr 30 11:00:33 corosync [TOTEM ] Creating commit token because I am the<br>
> rep.<br>
> Apr 30 11:00:33 corosync [MAIN ] Storing new sequence id for ring 2e68<br>
> Apr 30 11:00:33 corosync [TOTEM ] entering COMMIT state.<br>
> Apr 30 11:01:03 corosync [TOTEM ] The token was lost in the COMMIT state.<br>
> <br>
> On the other nodes - I see messages like<br>
> notice: pcmk_peer_update: Transitional membership event on ring 11888:<br>
> memb=2, new=0, lost=0<br>
> Apr 30 11:06:10 corosync [pcmk ] info: pcmk_peer_update: memb:<br>
> SG-azfw2-190 301994924<br>
> Apr 30 11:06:10 corosync [pcmk ] info: pcmk_peer_update: memb:<br>
> SG-azfw2-191 603984812<br>
> Apr 30 11:06:10 corosync [TOTEM ] waiting_trans_ack changed to 1<br>
> Apr 30 11:06:10 corosync [pcmk ] notice: pcmk_peer_update: Stable<br>
> membership event on ring 11888: memb=2, new=0, lost=0<br>
> Apr 30 11:06:10 corosync [pcmk ] info: pcmk_peer_update: MEMB:<br>
> SG-azfw2-190 301994924<br>
> Apr 30 11:06:10 corosync [pcmk ] info: pcmk_peer_update: MEMB:<br>
> SG-azfw2-191 603984812<br>
> Apr 30 11:06:10 corosync [SYNC ] This node is within the primary component<br>
> and will provide service.<br>
> Apr 30 11:06:10 corosync [TOTEM ] entering OPERATIONAL state.<br>
> <br>
> Can the corosync experts please guide me on probable root cause for this or<br>
> ways to debug this further ? Help much appreciated.<br>
> <br>
> corosync version: 1.4.8.<br>
> pacemaker version: 1.1.14-8.el6_8.1<br>
> <br>
> Thanks!<br>
> <br>
> <br>
> <br>
> _______________________________________________<br>
> Manage your subscription:<br>
> <a href="https://lists.clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">https://lists.clusterlabs.org/mailman/listinfo/users</a><br>
> <br>
> ClusterLabs home: <a href="https://www.clusterlabs.org/" rel="noreferrer" target="_blank">https://www.clusterlabs.org/</a><br>
> <br>
<br>
</blockquote></div></div>