Re: [squid-users] object cached with certainty

From: Carlos Defoe <carlosdefoe_at_gmail.com>
Date: Thu, 13 Feb 2014 19:53:44 -0200

It's squid 3.3.8.

Yeah, I had forgotten about the replacement policies... in fact, I
read the "definitive guide", but it was almost 10 years ago.

I tried with an empty disk cache, default replacement policies, clean
start, and now it gets cached, but in memory. I'll wait till both
cache_mem and cache_dir get full, to test again.

Another possiblity that comes in mind is that the aufs threads could
have been very busy, thus discarding the objects instead of writing it
to disk.

One more question: how squid decides that an object should be stored
on disk instead of RAM? Those jpeg images that are now cached on RAM,
will they ever be stored on disk?

On Wed, Feb 12, 2014 at 6:21 PM, Amos Jeffries <squid3_at_treenet.co.nz> wrote:
> On 2014-02-13 09:47, Carlos Defoe wrote:
>>
>> Hello,
>>
>> is there a way to be sure that some objects will be cached?
>>
>> I'm trying to cache this image blog:
>> http://lustik.tumblr.com
>>
>> I configured one refresh_pattern line to match all tumblr, with some
>> options that, as far as I undestood, will agressively try to cache it.
>>
>> ####
>> # REFRESH_PATTERNS
>> ####
>> refresh_pattern -i tumblr.com 2880 90% 7200 override-expire
>> override-lastmod ignore-no-store ignore-reload ignore-private
>> # Add any of your own refresh_pattern entries above these.
>> refresh_pattern ^ftp: 1440 20% 10080
>> refresh_pattern ^gopher: 1440 0% 1440
>> refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
>> refresh_pattern . 0 20% 4320
>>
>> No luck. All I get with this is always TCP_MISS/200, for all objects. E.g:
>> TCP_MISS/200 172504 GET
>>
>> http://24.media.tumblr.com/967c977f757bc64f9e10184acc934bd2/tumblr_n0qsckwQA31qztdg6o4_500.jpg
>>
>> I tried to load that page on different browsers, and different
>> machines, but the objects are never cached. Why is that? What can I
>> do?
>
>
> What Squid version? Its a HIT for me with default cache settings, with the
> current latest version (3.4.3).
>
> HTTP/1.1 200 OK
> Accept-Ranges: bytes
> Cache-Control: max-age=31536000
> Content-Type: image/jpeg
> Date: Wed, 12 Feb 2014 21:08:36 GMT
> ETag: "1b6df41f754d349a0b3d9314d71431ee"
> Last-Modified: Sun, 09 Feb 2014 18:50:50 GMT
> Server: ECS (syd/EBBD)
> <snip>
> X-Cache: HIT
> Content-Length: 171808
> Age: 11
> X-Cache: HIT from treenet.co.nz
>
> Note that the max-age (1 year) from server is significantly larger than your
> 7200 minute (5 day) limit. So your rule will be *shortening* the object
> storage time if/when it works.
>
>
>>
>> If my disk cache is already full, the behavior should be to keep the
>> objects that are already stored, or delete the oldest and store those
>> new? I mean, this could be caused by a full cache dir?
>
>
> Maybe, but the space clearing removals also depends on time since last use.
> So this should not happen during testing.
>
> Amos
Received on Thu Feb 13 2014 - 21:53:55 MST

This archive was generated by hypermail 2.2.0 : Fri Feb 14 2014 - 12:00:04 MST