Re: 2.5.PRE9 released

I can probably buy that. We're playing with an external_acl_type inside
of a no_cache directive...maybe we're always matching on no_cache, even
though that wasn't the plan. The logic of external_acl OK|ERR when
matched with no_cache with its negative logic is too much for my little
head to wrap around, I guess. ;-)

I'll reverse the test in the helper and see what happens.

Henrik Nordström wrote:
> On Sat, 20 Jul 2002, Joe Cooper wrote:
>
>
>>Might need to add to the list 'mysterious' disappearence of aufs
>>threads. I've got three boxes running it, two of which lost their
>>threads during a restart. Strangely, on shutdown, we get 'aioSync:
>>done' messages, and the storedir cachemgr output reports the directories
>>as aufs storage type. But no threads are started.
>
>
> Hmm.. the threads are only started at the first I/O operation. Maybe these
> boxes have simply not touched any of the aufs directories after the
> restart? Got fooled by this myself when upgrading from my patchy 2.2
> version to 2.5..
>
> Regards
> Henrik
>

-- 
Joe Cooper <joe@swelltech.com>
Web caching appliances and support.
http://www.swelltech.com