Loggerhead usage statistics

Michael Hudson michael.hudson at canonical.com
Wed Apr 21 07:02:09 BST 2010


On 21/04/10 17:56, Martin Pool wrote:
> On 21 April 2010 11:31, Robert Collins<robertc at robertcollins.net>  wrote:
>> repeated hits to a single url should be cachable;
>
> They should be in principle;

The answer is a bit "sort of": revision/100 is stable so long as no-one 
does push --overwrite.

The +filediff pages are utterly cacheable based on URL (well, and 
loggerhead version I guess -- the URLs are based on revids).

> I don't know at the moment whether it
> does set cacheability headers.  It would be good to try.

I had a branch that added etags to most pages but I don't seem to have 
landed it:

bzr+ssh://bazaar.launchpad.net/~mwhudson/loggerhead/etag-support

I don't really know why -- I remember failing to get squid to cache 
based solely on etag and getting overtaken by other things...

>> is there a squid in front
>> of loggerhead, and if so, did you get its logs too?
>
> spm says there is one but it doesn't cache anything.  That would be an
> interesting next step.

Yeah.

Cheers,
mwh



More information about the bazaar mailing list