<div dir="ltr">It is Freescale e6500 processor. Nobody here has tried running it in LE mode so it is going to take some doing.<div>We are going to add some debug logs to figure out where does corosync initialization get stalled.</div><div>If you have have suggestions, pls let us know.</div><div><br></div><div>-Thanks</div><div>Nikhil<br><div><br></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 2, 2016 at 1:00 PM, Nikhil Utane <span dir="ltr"><<a href="mailto:nikhil.subscribed@gmail.com" target="_blank">nikhil.subscribed@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">So what I understand what you are saying is, if the HW is bi-endian, then enable LE on PPC. Is that right?<div>Need to check on that. </div></div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 2, 2016 at 12:49 PM, Nikhil Utane <span dir="ltr"><<a href="mailto:nikhil.subscribed@gmail.com" target="_blank">nikhil.subscribed@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Sorry about my ignorance but could you pls elaborate what do you mean by "try to ppcle"?<div><br></div><div>Our target platform is ppc so it is BE. We have to get it running only on that.</div><div>How do we know this is LE/BE issue and nothing else?</div><div><br></div><div>-Thanks</div><span><font color="#888888"><div>Nikhil</div><div><br></div></font></span></div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 2, 2016 at 12:24 PM, Jan Friesse <span dir="ltr"><<a href="mailto:jfriesse@redhat.com" target="_blank">jfriesse@redhat.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
As your hardware is probably capable of running ppcle and if you have an<br>
environment<br>
at hand without too much effort it might pay off to try that.<br>
There are of course distributions out there support corosync on<br>
big-endian architectures<br>
but I don't know if there is an automatized regression for corosync on<br>
big-endian that<br>
would catch big-endian-issues right away with something as current as<br>
your 2.3.5.<br>
</blockquote>
<br></span>
No we are not testing big-endian.<br>
<br>
So totally agree with Klaus. Give a try to ppcle. Also make sure all nodes are little-endian. Corosync should work in mixed BE/LE environment but because it's not tested, it may not work (and it's a bug, so if ppcle works I will try to fix BE).<br>
<br>
Regards,<br>
  Honza<div><div><br>
<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Regards,<br>
Klaus<br>
<br>
On 05/02/2016 06:44 AM, Nikhil Utane wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Re-sending as I don't see my post on the thread.<br>
<br>
On Sun, May 1, 2016 at 4:21 PM, Nikhil Utane<br>
<<a href="mailto:nikhil.subscribed@gmail.com" target="_blank">nikhil.subscribed@gmail.com</a> <mailto:<a href="mailto:nikhil.subscribed@gmail.com" target="_blank">nikhil.subscribed@gmail.com</a>>> wrote:<br>
<br>
     Hi,<br>
<br>
     Looking for some guidance here as we are completely blocked<br>
     otherwise :(.<br>
<br>
     -Regards<br>
     Nikhil<br>
<br>
     On Fri, Apr 29, 2016 at 6:11 PM, Sriram <<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a><br>
     <mailto:<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a>>> wrote:<br>
<br>
         Corrected the subject.<br>
<br>
         We went ahead and captured corosync debug logs for our ppc board.<br>
         After log analysis and comparison with the sucessful logs(<br>
         from x86 machine) ,<br>
         we didnt find *"[ MAIN  ] Completed service synchronization,<br>
         ready to provide service.*" in ppc logs.<br>
         So, looks like corosync is not in a position to accept<br>
         connection from Pacemaker.<br>
         Even I tried with the new corosync.conf with no success.<br>
<br>
         Any hints on this issue would be really helpful.<br>
<br>
         Attaching ppc_notworking.log, x86_working.log, corosync.conf.<br>
<br>
         Regards,<br>
         Sriram<br>
<br>
<br>
<br>
         On Fri, Apr 29, 2016 at 2:44 PM, Sriram <<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a><br>
         <mailto:<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a>>> wrote:<br>
<br>
             Hi,<br>
<br>
             I went ahead and made some changes in file system(Like I<br>
             brought in /etc/init.d/corosync and /etc/init.d/pacemaker,<br>
             /etc/sysconfig ), After that I was able to run  "pcs<br>
             cluster start".<br>
             But it failed with the following error<br>
              # pcs cluster start<br>
             Starting Cluster...<br>
             Starting Pacemaker Cluster Manager[FAILED]<br>
             Error: unable to start pacemaker<br>
<br>
             And in the /var/log/pacemaker.log, I saw these errors<br>
             pacemakerd:     info: mcp_read_config:  cmap connection<br>
             setup failed: CS_ERR_TRY_AGAIN.  Retrying in 4s<br>
             Apr 29 08:53:47 [15863] node_cu pacemakerd:     info:<br>
             mcp_read_config:  cmap connection setup failed:<br>
             CS_ERR_TRY_AGAIN.  Retrying in 5s<br>
             Apr 29 08:53:52 [15863] node_cu pacemakerd:  warning:<br>
             mcp_read_config:  Could not connect to Cluster<br>
             Configuration Database API, error 6<br>
             Apr 29 08:53:52 [15863] node_cu pacemakerd:   notice:<br>
             main:     Could not obtain corosync config data, exiting<br>
             Apr 29 08:53:52 [15863] node_cu pacemakerd:     info:<br>
             crm_xml_cleanup:  Cleaning up memory from libxml2<br>
<br>
<br>
             And in the /var/log/Debuglog, I saw these errors coming<br>
             from corosync<br>
             20160429 085347.487050 <tel:085347.487050> airv_cu<br>
             daemon.warn corosync[12857]:   [QB    ] Denied connection,<br>
             is not ready (12857-15863-14)<br>
             20160429 085347.487067 <tel:085347.487067> airv_cu<br>
             <a href="http://daemon.info" rel="noreferrer" target="_blank">daemon.info</a> <<a href="http://daemon.info" rel="noreferrer" target="_blank">http://daemon.info</a>> corosync[12857]:   [QB<br>
             ] Denied connection, is not ready (12857-15863-14)<br>
<br>
<br>
             I browsed the code of libqb to find that it is failing in<br>
<br>
             <a href="https://github.com/ClusterLabs/libqb/blob/master/lib/ipc_setup.c" rel="noreferrer" target="_blank">https://github.com/ClusterLabs/libqb/blob/master/lib/ipc_setup.c</a><br>
<br>
             Line 600 :<br>
             handle_new_connection function<br>
<br>
             Line 637:<br>
             if (auth_result == 0 &&<br>
             c->service->serv_fns.connection_accept) {<br>
                     res = c->service->serv_fns.connection_accept(c,<br>
                                              c->euid, c->egid);<br>
                 }<br>
                 if (res != 0) {<br>
                     goto send_response;<br>
                 }<br>
<br>
             Any hints on this issue would be really helpful for me to<br>
             go ahead.<br>
             Please let me know if any logs are required,<br>
<br>
             Regards,<br>
             Sriram<br>
<br>
             On Thu, Apr 28, 2016 at 2:42 PM, Sriram<br>
             <<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a> <mailto:<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a>>> wrote:<br>
<br>
                 Thanks Ken and Emmanuel.<br>
                 Its a big endian machine. I will try with running "pcs<br>
                 cluster setup" and "pcs cluster start"<br>
                 Inside cluster.py, "service pacemaker start" and<br>
                 "service corosync start" are executed to bring up<br>
                 pacemaker and corosync.<br>
                 Those service scripts and the infrastructure needed to<br>
                 bring up the processes in the above said manner<br>
                 doesn't exist in my board.<br>
                 As it is a embedded board with the limited memory,<br>
                 full fledged linux is not installed.<br>
                 Just curious to know, what could be reason the<br>
                 pacemaker throws that error.<br>
<br>
                 /"cmap connection setup failed: CS_ERR_TRY_AGAIN.<br>
                 Retrying in 1s"<br>
<br>
                 /<br>
                 Thanks for response.<br>
<br>
                 Regards,<br>
                 Sriram.<br>
<br>
                 On Thu, Apr 28, 2016 at 8:55 AM, Ken Gaillot<br>
                 <<a href="mailto:kgaillot@redhat.com" target="_blank">kgaillot@redhat.com</a> <mailto:<a href="mailto:kgaillot@redhat.com" target="_blank">kgaillot@redhat.com</a>>> wrote:<br>
<br>
                     On 04/27/2016 11:25 AM, emmanuel segura wrote:<br>
                     > you need to use pcs to do everything, pcs<br>
                     cluster setup and pcs<br>
                     > cluster start, try to use the redhat docs for<br>
                     more information.<br>
<br>
                     Agreed -- pcs cluster setup will create a proper<br>
                     corosync.conf for you.<br>
                     Your corosync.conf below uses corosync 1 syntax,<br>
                     and there were<br>
                     significant changes in corosync 2. In particular,<br>
                     you don't need the<br>
                     file created in step 4, because pacemaker is no<br>
                     longer launched via a<br>
                     corosync plugin.<br>
<br>
                     > 2016-04-27 17:28 GMT+02:00 Sriram<br>
                     <<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a> <mailto:<a href="mailto:sriram.ec@gmail.com" target="_blank">sriram.ec@gmail.com</a>>>:<br>
                     >> Dear All,<br>
                     >><br>
                     >> I m trying to use pacemaker and corosync for<br>
                     the clustering requirement that<br>
                     >> came up recently.<br>
                     >> We have cross compiled corosync, pacemaker and<br>
                     pcs(python) for ppc<br>
                     >> environment (Target board where pacemaker and<br>
                     corosync are supposed to run)<br>
                     >> I m having trouble bringing up pacemaker in<br>
                     that environment, though I could<br>
                     >> successfully bring up corosync.<br>
                     >> Any help is welcome.<br>
                     >><br>
                     >> I m using these versions of pacemaker and corosync<br>
                     >> [root@node_cu pacemaker]# corosync -v<br>
                     >> Corosync Cluster Engine, version '2.3.5'<br>
                     >> Copyright (c) 2006-2009 Red Hat, Inc.<br>
                     >> [root@node_cu pacemaker]# pacemakerd -$<br>
                     >> Pacemaker 1.1.14<br>
                     >> Written by Andrew Beekhof<br>
                     >><br>
                     >> For running corosync, I did the following.<br>
                     >> 1. Created the following directories,<br>
                     >>     /var/lib/pacemaker<br>
                     >>     /var/lib/corosync<br>
                     >>     /var/lib/pacemaker<br>
                     >>     /var/lib/pacemaker/cores<br>
                     >>     /var/lib/pacemaker/pengine<br>
                     >>     /var/lib/pacemaker/blackbox<br>
                     >>     /var/lib/pacemaker/cib<br>
                     >><br>
                     >><br>
                     >> 2. Created a file called corosync.conf under<br>
                     /etc/corosync folder with the<br>
                     >> following contents<br>
                     >><br>
                     >> totem {<br>
                     >><br>
                     >>         version: 2<br>
                     >>         token:          5000<br>
                     >>         token_retransmits_before_loss_const: 20<br>
                     >>         join:           1000<br>
                     >>         consensus:      7500<br>
                     >>         vsftype:        none<br>
                     >>         max_messages:   20<br>
                     >>         secauth:        off<br>
                     >>         cluster_name:   mycluster<br>
                     >>         transport:      udpu<br>
                     >>         threads:        0<br>
                     >>         clear_node_high_bit: yes<br>
                     >><br>
                     >>         interface {<br>
                     >>                 ringnumber: 0<br>
                     >>                 # The following three values<br>
                     need to be set based on your<br>
                     >> environment<br>
                     >>                 bindnetaddr: 10.x.x.x<br>
                     >>                 mcastaddr: 226.94.1.1<br>
                     >>                 mcastport: 5405<br>
                     >>         }<br>
                     >>  }<br>
                     >><br>
                     >>  logging {<br>
                     >>         fileline: off<br>
                     >>         to_syslog: yes<br>
                     >>         to_stderr: no<br>
                     >>         to_syslog: yes<br>
                     >>         logfile: /var/log/corosync.log<br>
                     >>         syslog_facility: daemon<br>
                     >>         debug: on<br>
                     >>         timestamp: on<br>
                     >>  }<br>
                     >><br>
                     >>  amf {<br>
                     >>         mode: disabled<br>
                     >>  }<br>
                     >><br>
                     >>  quorum {<br>
                     >>         provider: corosync_votequorum<br>
                     >>  }<br>
                     >><br>
                     >> nodelist {<br>
                     >>   node {<br>
                     >>         ring0_addr: node_cu<br>
                     >>         nodeid: 1<br>
                     >>        }<br>
                     >> }<br>
                     >><br>
                     >> 3.  Created authkey under /etc/corosync<br>
                     >><br>
                     >> 4.  Created a file called pcmk under<br>
                     /etc/corosync/service.d and contents as<br>
                     >> below,<br>
                     >>       cat pcmk<br>
                     >>       service {<br>
                     >>          # Load the Pacemaker Cluster Resource<br>
                     Manager<br>
                     >>          name: pacemaker<br>
                     >>          ver:  1<br>
                     >>       }<br>
                     >><br>
                     >> 5. Added the node name "node_cu" in /etc/hosts<br>
                     with 10.X.X.X ip<br>
                     >><br>
                     >> 6. ./corosync -f -p & --> this step started<br>
                     corosync<br>
                     >><br>
                     >> [root@node_cu pacemaker]# netstat -alpn | grep<br>
                     -i coros<br>
                     >> udp        0      0 10.X.X.X:61841     0.0.0.0:*<br>
                     >> 9133/corosync<br>
                     >> udp        0      0 10.X.X.X:5405      0.0.0.0:*<br>
                     >> 9133/corosync<br>
                     >> unix  2      [ ACC ]     STREAM     LISTENING<br>
                        148888 9133/corosync<br>
                     >> @quorum<br>
                     >> unix  2      [ ACC ]     STREAM     LISTENING<br>
                        148884 9133/corosync<br>
                     >> @cmap<br>
                     >> unix  2      [ ACC ]     STREAM     LISTENING<br>
                        148887 9133/corosync<br>
                     >> @votequorum<br>
                     >> unix  2      [ ACC ]     STREAM     LISTENING<br>
                        148885 9133/corosync<br>
                     >> @cfg<br>
                     >> unix  2      [ ACC ]     STREAM     LISTENING<br>
                        148886 9133/corosync<br>
                     >> @cpg<br>
                     >> unix  2      [ ]         DGRAM<br>
                       148840 9133/corosync<br>
                     >><br>
                     >> 7. ./pacemakerd -f & gives the following error<br>
                     and exits.<br>
                     >> [root@node_cu pacemaker]# pacemakerd -f<br>
                     >> cmap connection setup failed:<br>
                     CS_ERR_TRY_AGAIN.  Retrying in 1s<br>
                     >> cmap connection setup failed:<br>
                     CS_ERR_TRY_AGAIN.  Retrying in 2s<br>
                     >> cmap connection setup failed:<br>
                     CS_ERR_TRY_AGAIN.  Retrying in 3s<br>
                     >> cmap connection setup failed:<br>
                     CS_ERR_TRY_AGAIN.  Retrying in 4s<br>
                     >> cmap connection setup failed:<br>
                     CS_ERR_TRY_AGAIN.  Retrying in 5s<br>
                     >> Could not connect to Cluster Configuration<br>
                     Database API, error 6<br>
                     >><br>
                     >> Can you please point me, what is missing in<br>
                     these steps ?<br>
                     >><br>
                     >> Before trying these steps, I tried running "pcs<br>
                     cluster start", but that<br>
                     >> command fails with "service" script not found.<br>
                     As the root filesystem<br>
                     >> doesn't contain either /etc/init.d/ or<br>
                     /sbin/service<br>
                     >><br>
                     >> So, the plan is to bring up corosync and<br>
                     pacemaker manually, later do the<br>
                     >> cluster configuration using "pcs" commands.<br>
                     >><br>
                     >> Regards,<br>
                     >> Sriram<br>
                     >><br>
                     >> _______________________________________________<br>
                     >> Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
                     <mailto:<a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a>><br>
                     >> <a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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:<br>
                     <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/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>
                     Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
                     <mailto:<a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a>><br>
                     <a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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:<br>
                     <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/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>
         Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
         <mailto:<a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a>><br>
         <a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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:<br>
         <a href="http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf" rel="noreferrer" target="_blank">http://www.clusterlabs.org/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>
Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
<a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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/doc/Cluster_from_Scratch.pdf</a><br>
Bugs: <a href="http://bugs.clusterlabs.org" rel="noreferrer" target="_blank">http://bugs.clusterlabs.org</a><br>
</blockquote>
<br>
<br>
_______________________________________________<br>
Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
<a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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/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>
</blockquote>
<br>
<br>
_______________________________________________<br>
Users mailing list: <a href="mailto:Users@clusterlabs.org" target="_blank">Users@clusterlabs.org</a><br>
<a href="http://clusterlabs.org/mailman/listinfo/users" rel="noreferrer" target="_blank">http://clusterlabs.org/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/doc/Cluster_from_Scratch.pdf</a><br>
Bugs: <a href="http://bugs.clusterlabs.org" rel="noreferrer" target="_blank">http://bugs.clusterlabs.org</a><br>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>