I have recently taken over a new network and found something odd when I arrived. I can draw an actual Visio if needed but let me try it in text, we have:
(Internal Network) -->--(Inline LAN WAAS WAE#1 Inline WAN)-->--(Inline LAN WAAS WAE#2 Inline WAN)-->--(WAN Router)-->--(WAN)
Has anyone seen this in production before?? We are seeing some odd WAN speeds between this site and another large site, both with 50Mbps connections. I am not asking for help on the overall speed issues, just a validation that double inline WAE devices isn't hurting me more than helping. Both do have active connections if I do "Show Stat Connection".
I haven't seen this in real life before, but it should be possible to serially cluster two inline WAEs :
But it will required peer configuration as described in above link.
It pussles me though that both have active connections, unless WAE#1 is TFO overloaded.
From the link you posted, it doesn't appear that either of these devices are peered with each other or anything. Should I go ahead and follow the serial cluster guide in the link?? Or should I go ahead and open a TAC case and let someone validate these settings?
The "peer" setting is only used to prevent that e.g. traffic to a site without WAAS doesn't gets optimized on the LAN cable between WAE#1 and WAE#2 and thus spilling ressources on those two WAAS boxes.
Normally optimization (and auto-discovery) bypasses intermediate WAAS devices, and setup optimization between the two most far WAASes, i.e. closest to the server and client.
Does WAE#2 have optimizations to the same peers as WAE #1 ?
I think you should just follow the guide, but feel free to open a TAC case.