Re: [squid-users] Re: can't failover in round-robin

Mark Nottingham wrote:
> 4) use the forward_max_tries directive (currently on squid-2 HEAD and -3
> HEAD):
> http://www.squid-cache.org/bugs/show_bug.cgi?id=2632

And 3.1.0.8+
http://www.squid-cache.org/Versions/v3/3.1/changesets/b9541.patch

>
>
> On 30/06/2009, at 6:43 AM, Henrik Nordstrom wrote:
>
>> fre 2009-06-26 klockan 10:18 +0400 skrev Ken Peng:
>>
>>> After long time (it's about 3 minutes), Squid found that peer was dead.
>>> So my question is, by what way squid will find a peer is dead or alive?
>>> Why it takes so long time?
>>
>> It's not time, it's 10 requests forwarded to the peer. The more traffic
>> you have the faster this is, but when testing with a single user it may
>> feel slow.
>>
>> There is three things you can do
>>
>> a. Tune down the connect timeout. Can be set either globally or per
>> peer. This speeds up detection when a host is shut down.
>>
>> b. Enable retry_on_error making Squid automatically try the next
>> possible path on failure.
>>
>> c. Enable the per-peer monitorurl feature (not yet available in Squid-3)
>> telling Squid to actively monitor the health of the peer.
>>
>> All can be combined freely.
>>
>> Regards
>> Henrik
>>
>
> --
> Mark Nottingham mnot_at_yahoo-inc.com
>

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE6 or 3.0.STABLE16
   Current Beta Squid 3.1.0.9