We're not flushing the caches at all.  I did modify AxKit to store
everything in multiple subdirectories (based on the key...so
09e301ca411f508c7e1ffe349de2cc2f would be stored in axkit/09/e3), so I
avoided the "very large directory lookup" problems we had.  

Of course, the cache is 80546 files totalling 850M...which is a little
unwieldy :)

Brian

On Thu, 2001-12-06 at 15:57, Sebastian Rahtz wrote:
> Would people like to outline their AxKit cache-housekeeping
> mechanisms? We started off with a simple reaper which
> zapped files which had not been accessed for a week, but this
> seems to break AxKit. I assume this is because a call for foo.bar.baz
> generates XXXX and XXXX.gz, which may get accessed at different times. 
> But then AxKit seems to assume XXXX.gz exists if XXXX does.
> 
> I assume that a) people *are* using caches, and b) are not
> simply flushing them by hand?
> -- 
> Sebastian Rahtz      OUCS Information Manager
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to