[Twisted-Python] trac's reliability
corydodt at gmail.com
Mon Nov 27 18:12:50 EST 2006
I support using foom's /usr/local compiled SVN 1.4 on wolfwood as-is, and
migrating to fsfs. I have used fsfs a bit here, and I'm confident that it
will not lose data. If we migrate the SVN database, it might actually make
a server-wide upgrade to a supported (recent) OS a little easier.
I wouldn't recommend using a /usr/local compiled postgresql, however. If we
switch to a pg backend for trac, I have a hunch things will go smoother if
the server-wide upgrade is done first.
In my head the basic plan looks like:
1- migrate to fsfs on a recent SVN. Watch for improvements in trac. If
trac starts to behave, stop here.
2- upgrade wolfwood server. Watch for improvments in trac. Upgrade pyramid
while you're in there. ;-) If trac improves, stop here.
3- Migrate backend to pgsql. Watch for improvements in trac. If trac
doesn't improve, migrate to something not-trac.
Yes, I am volunteering my time. I don't currently have any permissions on
these servers apart from commits.
On 11/27/06, Jonathan Lange <jml at mumak.net> wrote:
> On 11/28/06, Phil Mayers <p.mayers at imperial.ac.uk> wrote:
> > exarkun at divmod.com wrote:
> > > If we convert the repository to fsfs, we might find the segfaults from
> > > the svn bindings disappear (of course, we might not - I think we can
> > > recognize the quality of this sort of bug stomping). In addition to
> > Whilst I agree generally those kinds of statements are disturbing, in
> > this particular case I can offer my 100%-independent support for that
> > action.
> For more information, see the Subversion book's entry on the two
> Twisted-Python mailing list
> Twisted-Python at twistedmatrix.com
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Twisted-Python