On Mon, 2003-06-16 at 14:01, Not Zed wrote:
> >
> > > Now, there's a couple of things that I can think of, given the above,
> > > that might lend itself to this lock-up. I normally don't "auto-expunge
> > > trash on exit" with my $HOME/Maildir, which means, that I may have more
> > > than several hundred entries in that vfolder at any given point in
> > > time. It would seem that this might be problematic. What do you think?
> >
> > Naah shouldn't be. vFolders are just in-memory dynamic queries, nothing
> > gets saved to disk. The main things that get saved to disk are the
> > summary files - but these should be relatively small (compared to the
> > size of the mailbox). Although if you have a lot of folders, it could
> > write a lot of these out.
>
> Hmm, seems i may have spoken a little too soon - however, a few hundred
> deleted messages should still be no issue. Unless perhaps you have 10's
> of thousands of messages elsewhere in the system.
>
> I did some testing with large folders, copying lots of messages and the
> like, and i hit some weird shutdown stuff - lots of cpu activity, and
> then a memory blowout just before it quit. The large folder in question
> has >35000 messages. Machine is Athlon 2400+/512mb, so it didn't hit
> swap, but it gobbled up a few extra 10's of MB pretty quickly.
>
> The cpu activity seems to be due to vfolder's being closed down and some
> inefficient array manipulations therein. Code that basically does (when
> the much more complex logic around it works itself out):
>
> for (i=0;i<35000;i++)
> g_ptr_array_remove(array, 0);
> (i.e. worst case, O(N^2) stuff)
>
> I didn't manage to catch the memory blowout cause (machine is too
> fast?), and its got late again (>3am), so i might look at it tommorow.
>
> Not really happy with a lot of the vFolder stuff, but its kinda hard to
> fix properly :(
Okay, I'm replying because this just happened to me again. But this
time, it took 14 minutes(!!!!!) to get my laptop to respond again.
Here are all the particulars that I can think of....
2003-06-19
12:31:58 - clicked the close button in evolution
12:45:57 - X became responsive again
- evolution version 1.4.0
- evolution.log shows only this for this time window (evolution was
started with "CAMEL_VERBOSE_DEBUG=1 evolution")....
(evolution:7332): GLib-GObject-WARNING **: gsignal.c:2010: instance
`0x858ad40' has no handler with id `202360'
Checking expire cycle time on dir '/home/gideon/evolution/cache/http/00'
Checking '/home/gideon/evolution/cache/http/00/.' for expiry
Checking '/home/gideon/evolution/cache/http/00/..' for expiry
set language en-US en-GB en-CA
- in Maildir, I see this as the most-recently modified files....
-rw------- 1 gideon gideon 788017 Jun 19 12:30 Sent.ev-summary
-rw------- 1 gideon gideon 25970 Jun 19 12:31 Spam.ev-summary
-rw------- 1 gideon gideon 5836 Jun 19 12:31 ..ev-summary
(timestamp is before X lock)
- fetchmail and procmail were not trying to deliver any mail during this
freeze time
- nothing in $HOME/evolution was altered after 12:31
- 472 "trash" items (l $HOME/Maildir/**/*T | wc -l)
- 115 mailboxes (find $HOME/Maildir -name new -type d | wc -l)
- 6731 e-mail messages (find $HOME/Maildir -type f | wc -l)
- Linux 2.4.21-ac1, XFS filesystems all around, debian sid, 1G RAM, 2.7G
free in $HOME
This is driving me nuts. I'm putting up with a really-slow-running
evolution as I type this since I'm now running it through gdb in hopes
that if/when this happens again, I might be able to ^C in gdb to see
what's going on.
Anyone seeing this or have any theories???
--
,-----------------------------------------------------------------//
| Jason 'vanRijn' Kasper :: Numbers 6:22-26
`
| All brontosauruses are thin at one end, much MUCH thicker
| in the middle, and then thin again at the far end. That is
| the theory that I have and which is mine, and what it is too.
,
| bash$ :(){ :|:&};:
`----------------------//
__________________________________________________________________________
Disclaimer: This e-mail message is intended only for the personal use of
the recipient(s) named above. If you are not an intended recipient, you
may not review, copy or distribute this message. If you have received this
communication in error, please notify us immediately by e-mail and delete
the original message.
This e-mail expresses views only of the sender, which are not to be
attributed to Rite Aid Corporation and may not be copied or distributed
without this statement.
_______________________________________________
evolution maillist - [EMAIL PROTECTED]
http://lists.ximian.com/mailman/listinfo/evolution
