Re: [Evolution] Search Folders on IMAP+

[Patrick O'Callaghan <poc usb ve>]

On Wed, 2011-01-05 at 14:54 -0500, Philippe LeCavalier wrote:
> On Wed, 2011-01-05 at 20:29 +0100, Milan Crha wrote: 
> > On Wed, 2011-01-05 at 13:29 -0500, Philippe LeCavalier wrote:
> > > Yes. I noticed that so at least it's not hard to recover. Forcing a
> > > reset could be dangerous considering Evo could be intimately connected
> > > to other sockets during which time it's hanging on others...In my case
> > > it's not inconsistant at all. I can count on it hanging at 0%. I think
> > > Adam's suggestion of turning some additional logging on is likely the
> > > only way to "see" more info.
> > >
> > > Thanks again. I'll reply if I find anything.
> >
> >     Hi,
> > what about trying to see what is evolution doing that the process got
> > stuck at 0%? Please install debug info packages for
> > evolution-data-server and evolution. Then enter your search folder (the
> > best when the initial update of folders after start will be done), and
> > then run this command on a console:
> >    $ gdb --batch --ex "t a a bt" -pid=PID &>/tmp/bt.txt
> > where PID is a process ID of a running evolution (you can get it by, for
> > example, ps -A | grep evo).
> >
> > Then get the /tmp/bt.txt file, check if it doesn't contain any private
> > data, like passwords, usernames, server addresses you do not want to
> > share with others, and then post it here, though dealing with this
> > through https://bugzilla.gnome.org/enter_bug.cgi?product=Evolution would
> > be better, to not bother other users/readers of this list.
> >
> > You can also try to get backtrace after some time, say a minute or five,
> > to see whether anything changes. Please check your CPU usage, which may
> > indicate some processing and make sure you'll write exact versions of
> > evolution-data-server and evolution you are using.
> >     Thanks and bye,
> >     Milan
> Those are certainly the best technical approches :) Just out of
> curiosity I tried removing the "is not" conditions(2) and the search
> completed in no time and has all the relevant mail in it. So there
> appears to be something with that condition...
> Maybe the above would help flush that out...I always try the lists
> before posting bugs just to avoid user issue clogging the system. If
> this does turn out to be a bug I will certainly report it. Besides I
> don't think anyone here minds? Is this in part what this list is for?
> If not then my understanding of mailing lists is very inaccurate.

Use your judgement. A long and very technical thread may not be
appropriate (the evolution-hackers list could be better) but the
occasional trace is OK. Do take the trouble to send it as a compressed
attachment though, or post it to a website, instead of creating huge
messages people have to download.

poc