Two redirects on one interface?

[lucperreau]

Hi Guys,

I already have a proxy i am redirecting to for a specific product. but now i have a second proxy i have put in and i want to redirect another ACL to it.

When i do:

interface GigabitEthernet7/1

ip wccp 90 redirect in

ip wccp 100 redirect in

Only ip wccp 90 works and the ip wccp 100 doesnt.

When i put only one ip wccp command, meaning either 90 or 100 at a time. They both work individually. When when both are on the interface, only one works, the ip wccp 90 one.

Any idea why this is so?

Thanks.

[Felix Arrieta · Accepted solution]

Hi,

What platform are you redirecting the traffic from ? can you enter the following command  ip wccp check services all on the global config mode ?

The Service Groups are considered in priority order (not applicable to Cat6K) until a Service Group is found that matches the IP packet and redirection in or out ( so that is the normal process)

The problem comes when you have an ACL configure for the group or groups, and the redirection is the same (on your case "redirect in" )

so please try the command and if it does not like it then post your current version and platform and will see if there's any code that supports it and hopefully that's going to fix it!

Saludos,

Felix Arrieta

[lucperreau]

Felix,

The command you gave me worked. May i know what this command is actually doing?

Thanks for helping btw.

regards,

Luc

[Felix Arrieta]

Hello Luc,

Glad it worked, in regards the command specific job, is just to avoid one packet to be drop after not matching the first WCCP group, lets say wccp service 90 is the one with higher priority if one packet does not match the service 90 and its ACL then it gets drop.

When the ip wccp check services all command is configured, WCCP will continue to attempt to match the packet against any remaining lower priority services configured on the interface so next will be service 100.

this can also help to clarify

https://supportforums.cisco.com/message/3917591#3917591

Regards,

Felix

[lucperreau]

Felix, gotcha! Thanks again for helping. I've learnt somethign new today after all those years of using WCCP

[Felix Arrieta]

I know the feeling =)  when I saw your question and recall one big headache I had about a year ago working  on it

see you !

[Dr.X]

hi  ,

i want to ask ,

i have the same question .

i have platform cisco 7604 .

i have two services groups with acl on the same interface .

my problem is :

both of wccp works , but i note degredation in one of the squid servers  !!!!!

i dont know if it has relation to the command

ip wccp check services all

i have put it , but i didnt note any improvement in squid cache ??

im sure that my ACL is ok !

regards

[Felix Arrieta]

Hi,

What I get from degradation is that the traffic is being properly routed  to the Squid-server but the performance is not what it used to be  ... right ?

if that is the case then start with  troubleshooting  the squid server itself .. look for error logs, overload conditions . CPU utilization charts ( month .weeks. per days )

on the router you can get wccp statistics and open a TAC case for review..

show ip wccp 

show ip wccp  detail 

show egress-methods 

regards,

[Dr.X]

hi , i dont think it is squid issue ,

i want to say somthing ,

there is a router before the 7600  MLS.

if i put the squid on that router "7200 rouiter"

i note that the "in" traffic to squid  is about 200

and "out" traffic  from squid 270

which mean that suqid is making 70 M in band .

if i put the squid server on the 7604  where it will be hit with the same traffic .

i note that the traffic is about 200 M

but the out traffic is 240 M

which mean that suqid is making 40  M in band . !!!

it mean that i lose 30 M band utilization ??!!!!!

note that squid still has the same ip ??!!! but i dont know why  that degradation occurs !!!

regards

[Dr.X]

it may be somethign wrong in the  redirection from squid to clients , not in redirection from clients to squid !!!

thats which i doubt !!

[Felix Arrieta]

Hi,

Sounds to me like a bandwidth consuming troubleshooting on the server itself then WCCP redirection.

let's see if  there's anybody else with more ideas..

regards,

[Dr.X]

hi ,

here is my topology with config on it .

note the wccp configs under the interfaces below ,

note that i added the global command

ip wccp check services all

  and it seems that same issue ,

one of the two servers get some degradation !!!!

here is my wccp config from squid , wish it help :

squid1

########### WCCP2 Config#############

wccp2_router x.x.x.x

wccp_version 2

wccp2_forwarding_method 2             

wccp2_return_method 2

wccp2_assignment_method 2

wccp2_service dynamic 60

wccp2_service_info 60 protocol=tcp flags=src_ip_hash priority=240 ports=80

wccp2_service dynamic 70

wccp2_service_info 70 protocol=tcp flags=dst_ip_hash,ports_source priority=240 ports=80

=============================================================================

squid 2

########### WCCP2 Config############# 

wccp2_router x.x.x.x

wccp_version 2

wccp2_forwarding_method 2

wccp2_return_method 2

wccp2_assignment_method 2                 

wccp2_service dynamic 80

wccp2_service_info 80 protocol=tcp flags=src_ip_hash priority=240 ports=80

wccp2_service dynamic 90 

wccp2_service_info 90 protocol=tcp flags=dst_ip_hash,ports_source priority=240 ports=80