Problem Wiht ISCSI FAILOVER SERVICE

I have Two Server.
DSS01 and DSS02
I have Open-e DATA STORAGE SOFTWARE V6.
i have create replications from DSS01 To DSS02 and then i have Create Virtual ip for failover service. My DSS02 Crash and i have shutdown it. Unfortunately DSS01 power-down. It is incredible the FAILOVER SERVICE is enable to strat because DSS02 is not present and then it is impossible to start the service and then i cannot ping the virtual ip. My Wmvare infrastructure not function. It' is possible to start Failover Service? There is some configuration to check? I think it is impossible that Open-e not run if DSS02 is not present and DSS01 PowerDown accidentaly. Anyone can help me???:mad:

help55,

When anything this serious happens I would call Open-E support immediately via phone since your hard down.

Just to make sure I understand the problem your DSS secondary was down, and then your primary went down, and when the primary came back up, you lost your Virtual IP.

I would hope someone at QC in Open-e tested this scenario because I could definitely see it happening.

Anybody else not yet in production have the guts to test this scenario? :o

Hi!

We have 5 pairs of DSS V6 servers running in our environment.

We didnt`t have a crash or failure of our servers but whenever we make an update of either DSS or BIOS / Firmware of the servers we first do the update on the secondary and then reboot it and the primary is running as stand-alone with the virtual-IP.

When the secondary is back up we do sync of the iSCSI Volumes and then a manual failover to the secondary and then update the primary and do the reboot of the primary. The secondary will run with virtual IP then.

When the primary is back up we do a resync of the volumes and then a manual failback and then the cluster will be running again under its virtual ip.

This scenario worked fine several times so far.

I hope this helped a little bit.

Regards
DerFux

@derfux:

Did you suspend (stop) the failover on the secondary node, before shutdown? Cause there's the option to stop the failover on the secondary node. I just want to get sure that that might be the best way in case of updates.

We tried both ways.
First with stopping the failover on the secondary node which is obviously the correct way.
And the next time without stopping it to check how the two servers behave.

I don't think a proper shutdown is the issue.

How DSS reacts when the secondary is powered down and then the primary DSS power is bounced is the issue. May try it over Christmas with a pair of PC's I have laying around and a few usb drives. Really love the 60 day trial.