Ask, We've been running pretty much this exact patch for about 6months in production with no problems. We had a similar issue and thought that lighttpd was a great solution:)
--Justin Ask Bjørn Hansen wrote: > I'm storing some big (~100MB) files in one installation and perlbal is > eating up gobs of memory (gigabytes some of them ended up on). Rather > than tracking that down I just changed to lighttpd. > > lighttpd, correctly apparently, returns 405 when it won't create a > directory that already exists. Brad earlier - > http://lists.danga.com/pipermail/mogilefs/2007-April/000886.html - > implied that it's important to detect. Why do we care at all? If we > can't make the directory it just won't work anyway, no? Why does it > matter that we're sending extra bogus requests in that case? Or does > it just matter that the directory actually got created and we need to know? > > If the latter, wouldn't it be better to just do a PROPFIND or some such > after MKCOL to check that the directory is there? > > > - ask > > > --- lib/MogileFS/Device.pm (revision 1129) > +++ lib/MogileFS/Device.pm (working copy) > @@ -366,7 +366,7 @@ > my $ans = <$sock>; > > # if they don't support this method, remember that > - if ($ans && $ans =~ m!HTTP/1\.[01] (400|405|501)!) { > + if ($ans && $ans =~ m!HTTP/1\.[01] (400|501)!) { > $self->{no_mkcol} = 1; > # TODO: move this into method on device, which propogates to > parent > # and also receive from parent. so all query workers share > this knowledge > > >
