Further to this, Felix, if you decided against inline deployment, you can set up WCCP ACLs that would ensure that only traffic to/from the WAAS-enabled spoke site is redirected at the head-end.
i.e. if the spoke site is 192.168.10.0/24, you could have a config like this on the WCCP router(s) at the hub site: ip access-list extended WCCP61-LAN permit ip any 192.168.10.0 0.0.0.255 ! ip access-list extended WCCP62-WAN permit ip 192.168.10.0 0.0.0.255 any ! ip wccp 61 redirect-list WCCP61-LAN ip wccp 62 redirect-list WCCP62-WAN ! interface WANx/x description WAN side ip wccp 62 redirect in ! interface LANx/x description LAN side ip wccp 61 redirect in You can do something similar on the spoke site to ensure that you only redirect and optimise traffic that's come from the specified subnets (or whatever you choose to put in the ACL). Otherwise, everything TCP is redirected, possibly unnecessarily. Yes, it's handled transparently and passed-through, but I prefer not to add extra processing if possible. cheers, Dale On Tue, May 19, 2009 at 7:45 AM, Felix Nkansah <[email protected]> wrote: > Thanks Andrew. > Your response is appreciated. > > > On Mon, May 18, 2009 at 9:00 PM, Tolstykh, Andrew < > [email protected]> wrote: > >> >>If the WAE at the head office accelerates traffic going to a spoke >> site >> >>without a WAE, would the traffic be dropped? >> >> No >> >> >>If the hub site receives non-accelerated traffic from spoke sites >> without >> >>WAE, would the head office WAE drop the traffic? >> >> No >> >> Cisco WAAS is also transparent in the sense that accelerator appliances >> can use auto-discovery to determine whether a peer accelerator is >> available at the other end of the link. After auto-discovery, a pair of >> accelerators can auto-negotiate an acceleration policy to be applied to >> the application flow. If a peer accelerator is not discovered, the >> application flow passes through unchanged. >> >> HTH, >> Andrew >> >> -----Original Message----- >> From: [email protected] >> [mailto:[email protected]] On Behalf Of Felix Nkansah >> Sent: Monday, May 18, 2009 3:43 PM >> To: [email protected] >> Subject: [c-nsp] OT: Cisco WAAS Setup Scenario >> >> Hi Team, >> Pardon me for the OT. >> >> I want to deploy Cisco WAAS as a proof of concept to a client with >> several >> sites connected in a hub-n-spoke topology. >> >> I would deploy only one WAE (and a CM) at the hub/head office and one >> WAE at >> a selected spoke, in production. >> >> I intend on setting the WAEs Inline for simplicity. However, I have some >> doubts that I hope you could help clear. >> >> If the WAE at the head office accelerates traffic going to a spoke site >> without a WAE, would the traffic be dropped? >> >> If the hub site receives non-accelerated traffic from spoke sites >> without >> WAE, would the head office WAE drop the traffic? >> >> I am concerned because I know the acceleration process utilizes >> compression >> schemes which may require decompression at the other site by a WAE. >> >> Labbing this up would give me the answers, but I felt I could leverage >> your >> skills for quick answers to these :-) >> >> Your responses are appreciated. >> >> Felix _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
