[ovirt-users] Re: Add Storage Domain to existing Datacenter - side effect

Hello Fred, Thanks for clarification. Is there a RedHat or oVirt link related to this behavior ? If some hosts have access to the volume and some another have no access, what the behavior ? Is there a percentage calculation like “Skip fencing on cluster connectivity issues” ? Regards, De : Fred Rolland <frolland(a)redhat.com&gt; Envoyé : dimanche 24 mars 2019 11:40 À : jeanbaptiste.coupiac(a)nfrance.com Cc : users <users(a)ovirt.org&gt; Objet : Re: [ovirt-users] Add Storage Domain to existing Datacenter - side effect If none of the nodes can access the storage domain, then the storage domain will move to inactive status and the nodes will stay in UP. If one of the node cannot access the storage domain and all the other nodes can access, then the storage domain will stay in UP and the node will not. The "Skip fencing on cluster connectivity issues " is not related to node <-> storage domain connectivity but related to oVirt engine(manager)<-> nodes connectivity. On Fri, Mar 22, 2019 at 3:51 PM <jeanbaptiste.coupiac(a)nfrance.com <mailto:jeanbaptiste.coupiac@nfrance.com> > wrote: Hello Guys, We are evaluating oVirt , and this morning I had to face to a side effect ( small impact because oVIrt Datacenter is actually a small one). 1. I added a new Storage domain from iSCSI SAN 2. Not all hosts HBA were authorized to access to this LUN (lack of configuration on SAN side) 3. One of ovirt Node on four nodes into my datacenter couldn’t access to the LUN => this ovirt node has gone into “Non-responsible” state. => All VMs running into this host has gone to another My question is regarding the “operational mode” : * If I add a Storage Domain which is not already fully configured / badly configured (“regarding HBA acces granted on SAN side), when I add the Storage Domain to my DC, each oVirt node into my datacenter will move to “non-operationnal” state after fews seconds since no one will be able to log into the new LUN ? (This state change can have some “major effect” (vm migration between hosts without any logic) ?) Does I miss something ? Is there a guard ? Is Cluster Fencing policy option: Skip fencing on cluster connectivity issues => Threshold: xx% can protect this type of error ? Regards, Jean-Baptiste _______________________________________________ Users mailing list -- users(a)ovirt.org <mailto:users@ovirt.org> To unsubscribe send an email to users-leave(a)ovirt.org <mailto:users-leave@ovirt.org> Privacy Statement: https://www.ovirt.org/site/privacy-policy/ oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/AIKWEISCUFX...