I tried doing this as a test:
refresh_pattern . 0 20% 4320
override-expire reload-into-ims ignore-reload
And then:
refresh_pattern . 60 20% 4320
override-expire reload-into-ims ignore-reload
Only with this last option will it cache anything. I am not doing a
forced reload (shift-reload), but my browser is set to check for a new
version on each load.
I don't want to force it to cache though.. shouldn't this be working
differently?
Hugh
Henrik Nordstrom wrote:
>On Sunday 02 March 2003 02.36, Hugh Buchanan wrote:
>
>
>
>>Basically, nothing is getting cached. Everything gets reported as
>>a TCP_MISS/200 in the access log. Nothing is getting written to
>>the cache.log. My store.log reports info like this:
>>
>>1046563848.097 RELEASE 00000035 200 1046563706 993686750 -1
>>image/gif 525/525 GET
>>http://www.mydomain.com/images/admin/header-bg.gif 1046563848.097
>>SWAPOUT 00000044 200 1046563848 993686750 -1 image/gif
>>525/525 GET http://www.mydomain.com/images/admin/header-bg.gif
>>
>>
>
>This tells that the previous object was considered stale, and a fresh
>copy was retrieved.
>
>As there is a identical last-modified timestamp this most likely is a
>forced reload by the client.
>
>If unsure enable log_mime_hdrs and compare two requests for the same
>object which you think should have been cached..
>
>
>You probably want to play a little with the
>ignore_reload/reload_into_ims and refresh_pattern settings to tune
>the caching of your accelerator.
>
>
>Regards
>Henrik
>
>
>
>
Received on Sun Mar 02 2003 - 21:02:17 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:13:53 MST