I am testing the above module in a 3845 router, which is at a remote site with a dual-T1 multipoint connection back to the data center. The Core device at the Datat center is a WAE-612 appliance.
Everything is set up properly, WCCP is redirecting OK, and I'm noticing a great improvement in the latency across the dual-T1 link.
The big WAN killer is when remote users upload a new set up database tables from the data center. in testing without WAAS, this process took 25 minutes, and transferred 209MB of traffic. During that time the latency on the WAN is anywhere between 20ms and 60ms, and other remote users really notice their apps slow way down.
With WAAS installed, the uploading of the database tables still takes about 25 minutes, but there is a 10x compression for SQL, and the WAN latency remains around 4-5 ms.
It seems that DRE and TFO are really working for this SQL upload, with the 10:1 compression, but the time remains the same. I would have thought that any DRE reads off the WAE disk would be quicker that pulling the actual data across the WAN - but maybe I'm wrong>
Should I see an decrease in time due to the 10:1 SQL compression statistics? Is the WAE disk read the limiting factor?