Re: [squid-users] 'squid -k reconfigure' and connectivity breaking

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Wed, 20 Apr 2011 18:28:15 +1200

On 20/04/11 17:09, Eugene M. Zheganin wrote:
> Hi.
>
> On 18.04.2011 13:47, Amos Jeffries wrote:
>> The behave identically in this regard. I suspect something is causing
>> 3.1 to resume service much slower than 2.7 did. Which particular 3.1
>> release is doing this?
> That's a 3.1.12 right now.

What does "right now" mean?
  you upgraded since the message was happening? or the 3.1.12 was
showing that message on each reconfigure?

>
>>> Anyway, is there a way to do a 'soft reconfiguration' ? Without closing
>>> HTTP/ICP/SNMP connections (or at least not breaking client
>>
>> Sadly not yet. We are working towards it for future releases.
> Good news.
>
>> If that is true, then I suspect you are using one of the early 3.1
>> releases with broken LFS support. Or something is breaking/corrupting
>> the swap.state journal during a reconfigure.
>> Does your cache.log contain a warning about "version 1 LFS" detected
>> or mention a DIRTY load during reconfigure? (may need ALL,1 debug level).
> Actually it says this all the time:
>
> Version 1 of swap file with LFS support detected...
>
> # grep "Version 1 of swap file with LFS support detected" cache.log | wc -l
> 100
>
> # stat swap.state
> 95 4012314 -rw-r----- 1 squid squid 16326000 10203960 "Apr 19 14:02:21
> 2011" "Apr 20 10:53:45 2011" "Apr 20 10:53:45 2011" "Apr 19 14:02:21
> 2011" 16384 19968 0 swap.state
>
> cache_dir:
>
> cache_dir ufs /usr/local/squid/cache 1100 16 256
>
> And the solution is... ?
>
> Thanks.
> Eugene.

Well ,bug 3053 was fixed in 3.1.9. So 3.1.12 should not be hitting the
LFS message repeatedly.

If it continues the problem will be Squid not getting enough time to
rebuild and save the cache index out during shutdown. So getting stuck
with an old format one being loaded over-and-over.
  Try stopping Squid. Erasing the /usr/local/squid/cache/swap.state
completely and starting Squid again. This will take a long time to load
the cache on that first restart, but the warning should be gone in future.
  DO NOT restart or reconfigure until the cache state rebuild is completed.

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE9 or 3.1.12
   Beta testers wanted for 3.2.0.7 and 3.1.12.1
Received on Wed Apr 20 2011 - 06:28:20 MDT

This archive was generated by hypermail 2.2.0 : Wed Apr 20 2011 - 12:00:03 MDT