I am new to AAM and trying to work through a POC but I'm stuck.
We have two data centers and replicate storage via iSCSI from site A to site B. Performance over our 1Gb WAN connection leaves much to be desired, and unfortunately a Compellent SAN provides almost no options in terms of tweaking TCP settings.
So I though I would give our BIG-IP 2200S a go to see if they can improve upon the current lackluster performance.
Here's my POC network:
So for my POC I have one BIG-IP 2200S and an instance of the VE to work with. My simulated WAN is a Linux box configured as a bridge which uses tc settings to introduce latency (works brilliantly). I currently have successful remote endpoint connectivity and advertised routes between the two BIG-IPs (showing just one side here):
So far, so good... but what do I do next? The documentation on symmetric optimizations is pretty sparse, and a lot of what's out there is outdated (for WOM). I'll be setting up iSCSI initiator/target on fs1 and fs2, so I'll look to tunnel and optimize TCP port 3260. I don't need encryption and the SAN already does dedup. What I do need is the TCP optimizations that I know the BIG-IP can provide.
I'm kinda stuck... any help is greatly appreciated!
