Cluster/Load-Balanced FCSvr

[JSamuel]

Before anyone screams at me too loudly, bare in mind that I don't claim to fully deploy FCSvr myself, I respect the boundaries of my own expertise, and generally call in the FCSvr veteran(s) to complete the user training and workflow integration with the client - I just start off the basics.

FCSvr is disappointing in some ways (fantastic in others) but from my background (long before I joined the Apple industry) I always view things in a particular way by sheer habit - a focus on HA/load-balancing/scability and DR.

How have you found FCSvr in a redundancy/HA environment? For me, I can easily create two identical boots, a shared storage system tweaked for FCSvr's database IO and stick a load balancer infront of all of it - but am I missing something?

[mw10dot1]

Hi Joel

As far as i know it is not possible to do. I do not know of any setup where you have multiple fcsvr's connecting to the same database.

I think as close as you could get to HA would be to have a second server sitting next to it. Then in the event of a failure restore the database to the new server.

Using the gui this would mean the most you would loose is 1 days work. Although i am sure you could backup the database from the command line to reduce this window.

Another option depending on how you are using fcsvr would be to have two separate instances on the same SAN with automations to sync the data between the two.

Michael

[JSamuel]

Hi Michael,

I thought as much
Scripted sync seems like the logical way of making a failover structure.

Food for thought