Re: Release Schedule

From: Robert Collins <robert.collins@dont-contact.us>
Date: 22 Aug 2001 10:52:06 +1000

On 22 Aug 2001 02:54:22 +0200, Henrik Nordstrom wrote:
> Robert Collins wrote:
>
> > > logd defenitely is NOT a 2.5 candidate in my opinion.
> >
> > Because it's too new and untried?
>
> Exactly. And it is only required by CYGWIN/Windows ports. For the UNIX ports,
> NTLM should still work fine. There is no more reason why a logrotate should

logrotate restarts the helpers synchronously, so that their stderr is
the new cache.log file.

Restarting the helpers synchronously kills all in-progress NTLM
authentication requests.

NTLM *NEEDS* asynchronous helper restarts. How this is done is
irrelevant, it _must_ have it.

> mess up NTLM than why it should leak NTLM helpers on aborted requests.

Its not necesarily aborted requests. Theres a race somewhere related to
the comm_close_handler stuff AFAICT (Not in that code, but in the
assocation of the user stuct to the conenction - I think). It took a
printf bomb to find that, but oh well.

Also worth noting, if NTLM leaked helpers on every aborted request, it
wouldn't hit 5.5 Million requests before 10 got leaked - thats 1 leak
every 550,000 requests.

Rob
Received on Tue Aug 21 2001 - 19:04:28 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:15 MST