Thank you for help. Now it's fixed by modifying the libvirtd and heartbeat starting and stopping order.
On Sat, Mar 13, 2010 at 9:26 AM, שלום קלמר <[email protected]> wrote: > Hi. > > I never tried that but mybe this article can help you: > > How do I create a clustered virtual service for my Xen guest using > system-config-cluster? Article ID: 5911 - Created on: Oct 28, 2007 7:00 PM > - Last Modified: Oct 28, 2007 7:00 PM > > The virtual service component of Red Hat Cluster Suite allows Xen guests to > be relocated or failed over among cluster members, providing high > availability of that guest. In order to use virtual services the cluster > must be running the Xen kernel and have shared storage, such as GFS on iSCSI > or SAN. This storage will be used to share the guest image(s) and > configuration files. > > After storage has been set up and mounted on all nodes, the guest must be > created on one of the nodes using the normal methods such as virt-manager or > virt-install. See the following link for more information on setting up > virtualized guests: > > > www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Virtualization-en-US/index.html > > Once the guest has been created and is completely shutdown, its disk image > and configuration file need to be moved to the shared storage. By default, > configuration files are stored in /etc/xen/ and images are stored in > /var/lib/xen/images. Make sure to edit the configuration file to point to > the new path for the guest image. > > Now, using system-config-cluster go to "Virtual Services" and click the > "Create a Virtual Service" button on the right side. In the dialog box, the > name of the Virtual Service should be the name of the guest that was > created. The path should point to the directory containing the guest > configuration file. > > Next, click the "Send to Cluster" button to save and propagate the > configuration changes to the cluster members. The guest should start > automatically on one of the nodes in the failover domain > > [r...@cluster2-1 ~]# xm list > > Name ID Mem(MiB) VCPUs State Time(s) > > clustered_guest 5 511 1 -b---- 11.5 > > Domain-0 0 1509 2 r----- 6077.3 > > It should only be running on one node at any given time. Note that without > further configuration, guests will be restarted rather than live migrated > upon relocation. > > > > > On Sat, Mar 13, 2010 at 9:36 AM, cheibi welid <[email protected]>wrote: > >> Hi, >> >> I currently testing a two-nodes HA-cluster with DRBD (Pacemaker, >> Heartbeat, DRBD). When the virtual machines (with KVM) are running on one >> node and I halt or reboot this node, heartbeat cannot migrate the resources >> to the other node. Then the virtual machines are unmanaged in both nodes. I >> think it's related to heartbeat init scripts or some configuration is >> missing ? >> >> Thank you in advance. >> >> Best regards. >> Chaibi. >> >> -- >> Linux-cluster mailing list >> [email protected] >> https://www.redhat.com/mailman/listinfo/linux-cluster >> > > > -- > Linux-cluster mailing list > [email protected] > https://www.redhat.com/mailman/listinfo/linux-cluster >
-- Linux-cluster mailing list [email protected] https://www.redhat.com/mailman/listinfo/linux-cluster
