P4500 Lefthand or vSphere 4 issue ?

[SoftComSC]

I have discovered an issue at which i am now at a loss about where the issue might lie.

Im in the process of setting up a complete new vSphere site with a P4500 Lefthand from HP as the common ISCSI storage.

Heres the information to decipher:

The left hand has been set up as a multi site storage with 2 lefthands in each site.

They replicate fine between each other, and the primary vSphere (Production site) servers connect fine to the Virtual Ip adress of the Lefthands. So does the secondary vSphere (Disaster Recovery site) servers also connect fine to the secondary left hands.

Now Ive emulated a disaster where in my Primary left hands dies, and so does the vSphere servers in production, and the secondary site vSphere's start up the Vm's perfectly.

Now here's the catch:

When my production site comes back up with ttheir lefthands, all of my vSphere hosts reconfigure themselves to look at the secondary lefthands instead of their own lefthands. Now this will mean that if I have a double disaster im completely down (wherein my production site goes down, and my secondary takes over, the installation will die if something happens to the secondary site eventhough the primary site is back online)..

Im at an end at what I should in this case, and as to whether this is a Left hand or a vSphere issue.

Any one have any ideas as to what I should look for ?

[weinstein5]

Are you managing the fail over manually? Have you thought about looking at SRM? SRM will handle this failure and manage the ESX hosts connecting back to the right SANs -

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

[SoftComSC]

The failover is handled perfectly by vSphere HA. Now SRM works with shadow copies, and a schedule. But the neat thing here is that the Lefthand automatically replicates my storage between the two sites on the fly and in synch.

[SoftComSC]

Well I almost forgot about this post which is now 8 months old, and I thought Id share the solution with you all.

Here's how it functions: The lefthand works with a specific manager that you installa as a virtual application and that one now handles the paths to your storage by a qurom calculation, meaning that he discoveres the fastest route that a host can take to the storage and then applies it to the hosts, so to summarize:

If you want to make sure that your hosts connect to the Lefthand from HP in a particualr way, the only way of achieveing that is by haveing a slow connection to your disaster recovery site.

Thank you for reading.

SC

[SoftComSC]

Old post that got solved over a year ago.The solution lies in the Lefthand virtual app which handles the fail over fatures and directs the host's to the proper destination in case of a system break down. Proper configuration of that is vital to the solution.

[idle-jam]

thanks for sharing the solution.