RE: [squid-users] How ti confirm of cache_peering is working between siblings?

Hi All,
I applied the configuration given below to get my siblings working , but unfortunately it did not.Now when I start the squid on sproxy1.example.com, I can see the following information in cache.log

2005/07/19 16:35:59| DNS Socket created at 0.0.0.0, port 32779, FD 6
2005/07/19 16:35:59| Adding nameserver x.x.x.x from /etc/resolv.conf
2005/07/19 16:35:59| Adding nameserver x.x.x.x from /etc/resolv.conf
2005/07/19 16:35:59| helperOpenServers: Starting 20 'squirm' processes
2005/07/19 16:36:02| Accepting HTTP connections at 0.0.0.0, port 8080, FD 7.
2005/07/19 16:36:02| Accepting ICP messages at 0.0.0.0, port 3130, FD 40.
2005/07/19 16:36:02| WCCP Disabled.
2005/07/19 16:36:02| Configuring Sibling sproxy2.example.com/8080/3130
2005/07/19 16:36:02| Configuring Sibling sproxy3.example.com/8080/3130
2005/07/19 16:36:02| Loaded Icons.
2005/07/19 16:36:02| Ready to serve requests.

But when I run
tcpdump src port 3130 on sproxy1
I can't see any communication between sproxy1 and sproxy2 or sproxy1 and sproxy3.However I can see the ICP communication between sproxy1 and a downstream proxy. And cache manager for sproxy1 says

Peer Digests:
no guess stats for all peers available

Per-peer statistics:

peer digest from sproxy2.example.com
no guess stats for sproxy2.example.com available

event timestamp secs from now secs from init
initialized 1121754959 -623 +0
needed 0 +0 +0
requested 0 +0 +0
received 0 +0 +0
next_check 0 +0 +0

peer digest state:
        needed: no, usable: no, requested: no

        last retry delay: 0 secs
        last request response time: 0 secs
        last request result: (none)

peer digest traffic:
        requests sent: 0, volume: 0 KB
        replies recv: 0, volume: 0 KB

peer digest structure:
        no in-memory copy

peer digest from sproxy3.example.com
no guess stats for sproxy3.example.com available

event timestamp secs from now secs from init
initialized 1121754959 -623 +0
needed 0 +0 +0
requested 0 +0 +0
received 0 +0 +0
next_check 0 +0 +0

peer digest state:
        needed: no, usable: no, requested: no

        last retry delay: 0 secs
        last request response time: 0 secs
        last request result: (none)

peer digest traffic:
        requests sent: 0, volume: 0 KB
        replies recv: 0, volume: 0 KB

peer digest structure:
        no in-memory copy

Algorithm usage:
Cache Digest: 0 ( -1%)
Icp: 0 ( -1%)
Total: 0 ( -1%)

Local Digest:
store digest: size: 7384615 bytes
         entries: count: 1705642 capacity: 11815384 util: 14%
         deletion attempts: 0
         bits: per entry: 5 on: 6442800 capacity: 59076920 util: 11%
         bit-seq: count: 11478902 avg.len: 5.15
         added: 1705642 rejected: 4186495 ( 71.05 %) del-ed: 0
         collisions: on add: 0.00 % on rej: 0.00 %

Can anyone please help which bit I am missing? or is there there any additional configuration(presumbly not) to get sibling cache peering working. Just to remind the objective is to configure either of three proxies as a sibling to each other so that thay can share each other's cache and either of these should be able to go to Internet in cacse if an object is not in any of cache.(As these proxies are at same level).

Regards,

Zia-ul-Hassan Zia
RMIT Infrastructure Services
UNIX Systems
Tel: 992 51964

>>> "Chris Robertson" <crobertson@gci.com> 07/08/05 8:27 am >>>
> -----Original Message-----
> From: Zia-ul-Hassan Zia [mailto:zia@rmit.edu.au]
> Sent: Wednesday, July 06, 2005 8:52 PM
> To: squid-users@squid-cache.org
> Cc: Chris Robertson
> Subject: [squid-users] How ti confirm if cache_peering is working
> between siblings?
>
>
> Hi Guys,
> How can I test if cache peering is working between three sdquid proxies.The configuration for peering on three sibling squid proxies is as follows
>
> on sproxy1
> -------------
> cache_peer sproxy2.example.com sibling 8080 3130 proxy-only
> cache_peer sproxy3.example.com sibling 8080 3130 proxy-only
> icp_access allow sproxy2.example.com
> icp_access allow sproxy3.example.com
> icp_access deny all
>
> on sproxy2
> -------------
> cache_peer sproxy1.example.com sibling 8080 3130 proxy-only
> cache_peer sproxy3.example.com sibling 8080 3130 proxy-only
> icp_access allow sproxy1.example.com
> icp_access allow sproxy3.example.com
> icp_access deny all
>
> on sproxy3
> -------------
> cache_peer sproxy1.example.com sibling 8080 3130 proxy-only
> cache_peer sproxy2.example.com sibling 8080 3130 proxy-only
> icp_access allow sproxy1.example.com
> icp_access allow sproxy2.example.com
> icp_access deny all
>
>
> Note that these are RedHat Linux boxes running Squid-2.5-STABLE5.Can some please advise what is the best way to confirm if peering is working between siblings?I tried to run
> tcpdump src 3130
> on either of three proxies, but got no output.
>
> TIA,
>
> Zia-ul-Hassan Zia
> RMIT Infrastructure Services
> UNIX Systems
> Tel: 992 51964

Try "tcpdump src port 3130".

Chris
Received on Tue Jul 19 2005 - 00:53:08 MDT