Regarding Sergei's second question: > There's other very much related question I would like to find answer for. > Right now we have Failback set as False for all our RGs. If you could shed > some light what's going to happen if Failback is set to True, i.e. first > node rebooted and RGs started going back before admins had chance to do > pairresync I'd greatly appreciate that. >
I can't answer this specifically for TruCopy, however, I can describe the general effect of Failback==TRUE in the RGM. When a node joins the cluster membership (i.e., when the node boots up into an existing cluster), RGs may be brought online on the booted node. If a resource group has Failback==TRUE, it will be switched from its current master onto the newly-booted node, if the booted node is more preferred -- which usually means that it occurs earlier in the RG's Nodelist. If Failback==FALSE, the resource group would remain on its current master even if the booted node is more preferred. However, if the RG is offline when the node boots, the RG will go online on the booted node regardless of the Failback setting. If the RG is scalable-mode (in other words, "multi-mastered") with Desired_primaries is > 1, then Failback applies if the RG is already online on its Desired number of primaries when the node joins. If the RG is online on fewer than Desired_primaries, this is like the offline case -- it would come online on the booted node regardless of the setting of Failback. If the RG is suspended, then that overrides the above behavior. A suspended RG will not automatically do anything when a node joins the cluster. If the RG is offline, it would remain offline. --Marty
