Re: [Tools-discuss] Re: Apache upgrade.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Tools-discuss] Re: Apache upgrade.




on 2006-11-14 18:15 Frank Ellermann said the following:
[...]
> Trying more old drafts, not the wanted effect, but here:
> http://www1.tools.ietf.org/html/draft-mengwong-sender-rewrite
> 
> document not found + additionally a 404 for the error document 

Aha.  Found this problem, fixed it.  A script which gradually deletes
old files in the /html/ cache also deleted the 404 handler script on
www1 [[ouch]] (but not yet on www2 and www3).

> On www2: only a document not found, no additional missing err404.
> On www3: same as www2
> 
> Now adding -01, same effect (now www1 doesn't claim that it has any
> additional issue with the err404 document).  Adding -01.txt, still
> no success.  Using www3 + id + explicit 01.txt: found.  Switching
> from id to html: works.

Hmm. Weird.  Could this be a problem with the client-side casheing
not taking the full host name into consideration??

> Okay, that's plausible, if I use an exact name with id (not html)
> on a specific server (especially www3), and then switch to ftml, 
> it works.  
> 
> But that's probably no "apache upgrade" effect, we discussed a
> similar effect some months ago, then you did something, I forgot
> what, cleared the cache maybe, and it worked again everywhere.

Right.  That was a cache issue, but I don't remember details.

> Wild guess:  Your cache is too big for some odd race condition
> somewhere, and www1 and/or www2 are more fragile than www3 wrt
> this rare condition (?)

Could be.  In this case the primary problem was the disappearance of
the 404 handler on www1, but I can't explain the problems with www2
and www3.  I'll keep an eye out for this (and if you have further
ideas, I'd be glad to hear them).

Regards,

	Henrik

_______________________________________________
Tools-discuss mailing list
Tools-discuss at ietf.org
https://www1.ietf.org/mailman/listinfo/tools-discuss




Note: Messages sent to this list are the opinions of the senders and do not imply endorsement by the IETF.