2009/5/26 Marti, Rob <RJM002@xxxxxxxx> > _______________________________________ > From: redhat-list-bounces@xxxxxxxxxx [redhat-list-bounces@xxxxxxxxxx] On > Behalf Of ESGLinux [esggrupos@xxxxxxxxx] > Sent: Tuesday, May 26, 2009 07:02 > To: General Red Hat Linux discussion list > Subject: Re: Default params for quorumdisk > > Hello, > thanks for your answer. The first link is really usefull for me. > > In short my problem is that when I unplug the network cable of the service > interface of one of the node, the two nodes halts. I have looking for an > explanation and I get this: > > "CMAN keeps track of cluster quorum by monitoring the count of cluster > nodes. If more than half the nodes are active, the cluster has quorum. If > half the nodes (or fewer) are active, the cluster does not have quorum, and > all cluster activity is stopped." > > With a two nodes cluster it seams that it one fails, then the cluster > hasn´t > quorum and the cluster stops. > > So I decide to use quorum disk to see if this solve my problem, but I´m not > sure if this is the problem, > > Greetings, > > ESG > > > > > 2009/5/26 Schaffrath, Rene <Rene.Schaffrath@xxxxxxxxxxxxxx> > > > Hi ESG, > > > > I'm definitely no fan of the RHEL Cluster Suite, so I'll keep it short: > > > > must read if you use multipathing (timings): > > http://kbase.redhat.com/faq/docs/DOC-2882 > _______________________________________ > > Do you have it set for a two node cluster? Hello, not at all, I haven´t configured multipathing. I´m now testing on a Xen enviroment with this: I create the quorom disk: mkqdisk -L mkqdisk v0.6.0 /dev/disk/by-id/scsi-S_beaf11-part1: /dev/disk/by-path/ip-192.168.1.191:3260-iscsi-iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz-lun-1-part1: /dev/sda1: Magic: eb7a62c2 Label: quorumdisk Created: Tue May 26 13:31:19 2009 Host: NODOXEN1 Kernel Sector Size: 512 Recorded Sector Size: 512 and in the cluster.conf I have setup this: <quorumd interval="10" label="quorumdisk" min_score="1" tko="10" votes="1"/> the values are to test, I don´t know if they are good. What I have seen is that it works better than when I havent this quoromdisk. look what happened now when I unplug one node: May 26 14:16:15 NODOXEN1 openais[1802]: [CLM ] got nodejoin message 192.168.100.2 May 26 14:16:15 NODOXEN1 openais[1802]: [CLM ] got nodejoin message 192.168.100.3 May 26 14:16:15 NODOXEN1 openais[1802]: [CPG ] got joinlist message from node 1 May 26 14:16:21 NODOXEN1 qdiskd[1830]: <info> Node 2 shutdown May 26 14:16:22 NODOXEN1 openais[1802]: [CMAN ] lost contact with quorum device May 26 14:16:26 NODOXEN1 kernel: dlm: connecting to 2 May 26 14:18:01 NODOXEN1 qdiskd[1830]: <info> Assuming master role I think this can be the answer to all of my problems, I´ll continue investigating, Greetings ESG > > > -- > redhat-list mailing list > unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe > https://www.redhat.com/mailman/listinfo/redhat-list > -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subjecthttps://www.redhat.com/mailman/listinfo/redhat-list
