aelred_1927
May 28, 2009Nimbostratus
Sticky CPU utilization (debunking info needed)
First, sorry if this is a bit off-topic.
We are having a raging debate in a large enterprise environment regarding use of Cisco's ACE product vs. F5 BigIPs for a new GTM/GLB function.
Two of the senior architecture wonks have made a statement that once sticky is turned on, only one CPU can pick up the load, which reduces throughput of the box. This is based on a presentation several years ago by some Cisco product engineers.
Does anybody have any knowledge/history of this? I've used BigIP at two large institutions previously and never run into this, but I don't have access to product or lab right now to disprove the veracity of the statement.
Any help is appreciated.
TIA,
-s