Re: [Linux-HA] Lots of configuration changes
The problem seems to be fixed. I have changed nothing in my configuration, but we had a problem in our network (it seems to be a loop in the spanning tree of some vlans). Once the network problem was fixed, the cluster run without problems. El 06/06/13 10:26, Angel L. Mateo escribió: Hello, I have a two node cluster with cman (v3.1.7) and pacemaker (v1.1.6) running in ubuntu 12.04 (as recommended in pacemaker documentation). This cluster has been running without problems for more than a month (since I isntalled it), but I'm having problems since I rebooted one node yesterday (I rebooted the "standby" node). The problem I'm having is that cman is continously changing its configuration. I can see with cman_tool status that the cluster generation value is continously increased. In the logs there a lot of messages of the form: Jun 6 09:29:49 myotis52 corosync[32048]: [TOTEM ] A processor joined or left the membership and a new membership was formed. Jun 6 09:29:49 myotis52 corosync[32048]: [CPG ] chosen downlist: sender r(0) ip(155.54.211.166) ; members(old:1 left:0) Jun 6 09:29:49 myotis52 corosync[32048]: [MAIN ] Completed service synchronization, ready to provide service. Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] CLM CONFIGURATION CHANGE Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] New Configuration: Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] r(0) ip(155.54.211.166) Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Left: Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Joined: Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] CLM CONFIGURATION CHANGE Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] New Configuration: Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] r(0) ip(155.54.211.166) Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Left: Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Joined: This is happening even when I run cman in just one node. I have attached my cluster.conf file. In this file I'm using udpu transport protocol because nodes are running in vmware infrastructure where I had problems with multicast. I have checked ip configuration and it is correct: ips are correctly configured in /etc/hosts and interfaces are working without any (apparent) problem. I have also checked this same configuration but without redundant ring (that is, without altname options and dlm control=detect and rrp_mode=none) but with the same problem. Any help? ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems -- Angel L. Mateo Martínez Sección de Telemática Área de Tecnologías de la Información y las Comunicaciones Aplicadas (ATICA) http://www.um.es/atica Tfo: 868887590 Fax: 86337 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] Lots of configuration changes
Hello,
I have a two node cluster with cman (v3.1.7) and pacemaker (v1.1.6)
running in ubuntu 12.04 (as recommended in pacemaker documentation).
This cluster has been running without problems for more than a month
(since I isntalled it), but I'm having problems since I rebooted one
node yesterday (I rebooted the "standby" node).
The problem I'm having is that cman is continously changing its
configuration. I can see with cman_tool status that the cluster
generation value is continously increased.
In the logs there a lot of messages of the form:
Jun 6 09:29:49 myotis52 corosync[32048]: [TOTEM ] A processor joined
or left the membership and a new membership was formed.
Jun 6 09:29:49 myotis52 corosync[32048]: [CPG ] chosen downlist:
sender r(0) ip(155.54.211.166) ; members(old:1 left:0)
Jun 6 09:29:49 myotis52 corosync[32048]: [MAIN ] Completed service
synchronization, ready to provide service.
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] CLM CONFIGURATION
CHANGE
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] New Configuration:
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] r(0)
ip(155.54.211.166)
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Left:
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Joined:
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] CLM CONFIGURATION
CHANGE
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] New Configuration:
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] r(0)
ip(155.54.211.166)
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Left:
Jun 6 09:29:51 myotis52 corosync[32048]: [CLM ] Members Joined:
This is happening even when I run cman in just one node.
I have attached my cluster.conf file. In this file I'm using udpu
transport protocol because nodes are running in vmware infrastructure
where I had problems with multicast.
I have checked ip configuration and it is correct: ips are correctly
configured in /etc/hosts and interfaces are working without any
(apparent) problem. I have also checked this same configuration but
without redundant ring (that is, without altname options and dlm
control=detect and rrp_mode=none) but with the same problem.
Any help?
--
Angel L. Mateo Martínez
Sección de Telemática
Área de Tecnologías de la Información
y las Comunicaciones Aplicadas (ATICA)
http://www.um.es/atica
Tfo: 868889150
Fax: 86337
___
Linux-HA mailing list
Linux-HA@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Is CLVM really needed in an active/passive cluster?
El 22/04/13 13:01, Lars Marowsky-Bree escribió: > On 2013-04-22T11:14:14, "Angel L. Mateo" wrote: > >> The problem I have is that I have firstly configured the cluster with >> CLVM, but with this I can't create snapshots of my volumes, which is >> required for backups. >> >> But is this CLVM really necessary? Or it is enough to configure LVM >> with fencing and stonith? > > If you're happy with having each volume group being active/passive, yes, > a normal LVM resource with fencing etc is sufficient. > > cLVM is only needed if you need to concurrently manage a volume group > cluster-wide. The choice can be made per VG. > Then, I want to configure cman + corosync + pacemaker (as described at 1). With this in mind which fence mechanism should I use? fence_scsi (my shared storage is in fibre channel scsi san)? Or the pcmk fence is enough? Where can I found documentation for fence mechanisms? In other configuration scenarios (2) using clvm, a dlm and sfex (3) resource is also created. Should I configure it too? Thanks for your help. 1. http://clusterlabs.org/quickstart-ubuntu.html 2. https://www.suse.com/documentation/sle_ha/book_sleha/?page=/documentation/sle_ha/book_sleha/data/book_sleha.html 3. http://www.linux-ha.org/wiki/Sfex_%28resource_agent%29 -- Angel L. Mateo Martínez Sección de Telemática Área de Tecnologías de la Información y las Comunicaciones Aplicadas (ATICA) http://www.um.es/atica Tfo: 868887590 Fax: 86337 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
Re: [Linux-HA] Is CLVM really needed in an active/passive cluster?
El 22/04/13 11:20, emmanuel segura escribió: > Hello Angel > > In this thread > http://comments.gmane.org/gmane.linux.redhat.release.rhel5/6395 you can > find the answer to your question > In that thread I can see that cLVM does not support snapshot. There is a link pointing to dd as a workaround for the snapshot (but I have a 10TB FS and dd is not an option). So my question continue, is clvm really needed? Or "legacy" LVM with fencing and stonich mechanism would be enough? -- Angel L. Mateo Martínez Sección de Telemática Área de Tecnologías de la Información y las Comunicaciones Aplicadas (ATICA) http://www.um.es/atica Tfo: 868889150 Fax: 86337 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
[Linux-HA] Is CLVM really needed in an active/passive cluster?
Hello, I'm deploying a clustered pop/imap server with mailboxes stored in a SAN connected with fibre channel. The problem I have is that I have firstly configured the cluster with CLVM, but with this I can't create snapshots of my volumes, which is required for backups. But is this CLVM really necessary? Or it is enough to configure LVM with fencing and stonith? -- Angel L. Mateo Martínez Sección de Telemática Área de Tecnologías de la Información y las Comunicaciones Aplicadas (ATICA) http://www.um.es/atica Tfo: 868889150 Fax: 86337 ___ Linux-HA mailing list Linux-HA@lists.linux-ha.org http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
