configuring waas devices nid help

Unanswered Question
Nov 5th, 2007

i'm new in configuring waas devices. and i have many questions.

basically our network is 6513 connected to 6509 via a multilink 4xE1(see attached). with multiple VLANs. we used 612 for core and edge, and 512 for cm.

i have configured the core, edge and cm.

my 1st question is:

1. is it correct that i connected the core_wae to VLAN 144 port then connected the waas_cm to VLAN 136 port?

2. in configuring wccp on the core_router, what will be the interface where traffic will be intercepted from when leaving the data center network toward the WAN? will it be int VLAN 136? because our waas_cm is connected to VLAN 136.

3. and the interface where traffic will be intercepted from when entering the data center network from the WAN, will be our multilink 4xE1?

4. and the have created a VLAN for redirection of core_wae VLAN 144. is this correct?

5. in cofiguring wccp on the core_wae router-list. i input the ip add of VLAN 144 in the 6509. is this correct?

because the first time i run the waas devices(after configuring the WCCP) it resulted to high cpu utilization of the core_router. from what i remembered it was "input packets" that has high utilization.

rgds,

ben

Attachment: 
I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Loading.
Zach Seils Mon, 11/05/2007 - 22:12

Ben,

Here are the answers to your questions:

1. The CM can reside anywhere, so long as it has IP connectivity to the WAEs it is managing. Your placement is fine.

2. Yes, VLAN 136. It's fine that the CM is also on that subnet. Think of the CM as just another host system.

3. Yes, it should be the logical multilink interface where other L3 configuration (IP address, etc.) resides.

4. Yes, the WAEs performing acceleration should reside on a separate subnet.

5. Yes.

When configuring WCCP on a hardware-based platform such as the Catalyst, make sure you configure WCCP as follows:

- L2 Forwarding (configured using the 'l2-redirect' command on the WAE)

- Mask Assignment (configured using the 'mask-assign' command on the WAE)

- Inbound Redirection

Thanks,

Zach

hagirebench Tue, 11/06/2007 - 17:30

zach,

thanks for the clarification.

now that i know that my configuration is right, can you tell me what is the possible reason why the core_router have high cpu utilization after i configured the wccp?

rgds,

ben

Zach Seils Tue, 11/06/2007 - 17:42

Ben,

Increased CPU utilization on the 6500 after enabling WCCP is usually caused by some portion of the redirection being handled in software. I mentioned the recommended configuration options in the previous post to ensure that all redirection is handled in hardware.

Can you post some additional information from the 6500:

show version

show run

show ip wccp

show ip wccp 61 detail

show ip wccp 62 detail

Thanks,

Zach

hagirebench Fri, 11/09/2007 - 00:32

hi zach,

sorry for the late response.

after requesting for the configs that you need i was informed that the wccp command was removed from the router. all i can give you now are the show version and show run.

see attahced:

note:

my mistake. the core router was also a 6513

please post me the links that you mentioned in you previous post(about the high cpu utilization). i have some trouble finding it. tnx.

rgds,

ben

Zach Seils Fri, 11/09/2007 - 13:30

Ben,

Thanks for the follow up. When configuring WCCP on Catalyst platforms, we recommend the following guidelines are followed:

1. Use L2 Forwarding (l2-redirect)

2. Use Mask Assignment (mask-assign)

3. Only use inbound redirection (ip wccp redirect in)

(1) and (2) are configured on the WAE.

Zach

Actions

This Discussion