I've been corresponding with poc about Evolution crashing on my system, which
due to a misunderstanding I never posted to the list. Here is a sort of digest.
The most important messages are the first and last, though the ones in the
middle contain a few useful nuggets.
jon
-------- Forwarded Message --------
From: Patrick O'Callaghan <p...@usb.ve>
To: Jonathan Ryshpan <jonr...@pacbell.net>, evolution-list@gnome.org
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 00:00:14 +0100
On Mon, 2016-07-25 at 13:15 -0700, Jonathan Ryshpan wrote:
>
> On Mon, 2016-07-25 at 19:37 +0100, Patrick O'Callaghan wrote:
> >
> > On Mon, 2016-07-25 at 08:48 -0700, Jonathan Ryshpan wrote:
> > >
> > > Evolution has been crashing frequently for the last few days, generally
> > > about
> > > once an hour when the system load is heavy, generally immediately after
> > > requesting some action like moving a message from one folder to another.
> > > The
> > > evolution window becomes unresponsive, then disappears after about 5 or 10
> > > seconds. Has anyone else seen this behavior?
> > >
> > > System: 4-processor true Intel
> > > Fedora-24, all updates installed
> > > KDE Frameworks 5.24.0
> > > Evolution 3.20.4
> >
> > What kind of folders are involved? Local, remote, virtual, one of each?
>
> All folders local, some as large as 75,000 messages, about 170 folders in all.
mbox or maildir? Moving messages between huge mbox folders can
certainly take a long time, and potentially cause other problems if
space is low on the filesystem where they are stored.
-------- Forwarded Message --------
From: Patrick O'Callaghan <p...@usb.ve>
To: Jonathan Ryshpan <jonr...@pacbell.net>, evolution-list@gnome.org
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 10:06:04 +0100
On Mon, 2016-07-25 at 18:50 -0700, Jonathan Ryshpan wrote:
>
> On Tue, 2016-07-26 at 00:00 +0100, Patrick O'Callaghan wrote:
> >
> > On Mon, 2016-07-25 at 13:15 -0700, Jonathan Ryshpan wrote:
> > >
> > > On Mon, 2016-07-25 at 19:37 +0100, Patrick O'Callaghan wrote:
> > > >
> > > > On Mon, 2016-07-25 at 08:48 -0700, Jonathan Ryshpan wrote:
> > > >
> > > > > Evolution has been crashing frequently for the last few days,
> > > > > generally about
> > > > > once an hour when the system load is heavy, generally immediately
> > > > > after requesting some action like moving a message from one folder to
> > > > > another. The evolution window becomes unresponsive, then disappears
> > > > > after
> > > > > about 5 or 10 seconds. Has anyone else seen this behavior?
> > > > >
> > > > > System: 4-processor true Intel
> > > > > Fedora-24, all updates installed
> > > > > KDE Frameworks 5.24.0
> > > > > Evolution 3.20.4
> > > >
> > > > What kind of folders are involved? Local, remote, virtual, one of each?
> > > > >
> > > > > All folders local, some as large as 75,000 messages, about 170 folders
> > > > > in all.
> >
> > mbox or maildir? Moving messages between huge mbox folders can
> > certainly take a long time, and potentially cause other problems if
> > space is low on the filesystem where they are stored.
>
> System is maildir. The available space is large:
> $ df -h $HOME
>
> Filesystem Size Used Avail Use% Mounted on
>
> /dev/mapper/fedora00-home 1.3T 345G 918G 28% /home
In that case I suggest running Evo with debugging enabled and posting
the tail of output when a hang occurs (not directly here please, use a
pastebin and post the URL here). Remember to remove any sensitive
information such as passwords.
-------- Forwarded Message --------
From: Jonathan Ryshpan <jonr...@pacbell.net>
To: Patrick O'Callaghan <p...@usb.ve>
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 08:32:41 -0700
On Tue, 2016-07-26 at 10:06 +0100, Patrick O'Callaghan wrote:
>
> mbox or maildir? Moving messages between huge mbox folders can
> > >
> > > certainly take a long time, and potentially cause other problems if
> > > space is low on the filesystem where they are stored.
> >
> > System is maildir. The available space is large:
> > $ df -h $HOME
> >
> > Filesystem Size Used Avail Use% Mounted on
> >
> > /dev/mapper/fedora00-home 1.3T 345G 918G 28% /home
>
> In that case I suggest running Evo with debugging enabled and posting
> the tail of output when a hang occurs (not directly here please, use a
> pastebin and post the URL here). Remember to remove any sensitive
> information such as passwords.
"Pastebin". Thats a word I've never seen before. I assume you mean upload the
output to Google docs (say) and post a link.
I have started it. BTW: do you have a better suggestion than
CAMEL_DEBUG=all evolution >& logfile
This produces lots of output; I've had some trouble with the very complex and
poorly documented evolution debug system.
-------- Forwarded Message --------
From: Patrick O'Callaghan <p...@usb.ve>
To: Jonathan Ryshpan <jonr...@pacbell.net>
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 16:54:43 +0100
On Tue, 2016-07-26 at 08:32 -0700, Jonathan Ryshpan wrote:
>
> On Tue, 2016-07-26 at 10:06 +0100, Patrick O'Callaghan wrote:
> >
> > mbox or maildir? Moving messages between huge mbox folders can
> > certainly take a long time, and potentially cause other problems if
> > space is low on the filesystem where they are stored.
> >
> > > System is maildir. The available space is large:
> > > $ df -h $HOME
> > >
> > > Filesystem Size Used Avail Use% Mounted on
> > >
> > > /dev/mapper/fedora00-home 1.3T 345G 918G 28% /home
> >
> > In that case I suggest running Evo with debugging enabled and posting
> > the tail of output when a hang occurs (not directly here please, use a
> > pastebin and post the URL here). Remember to remove any sensitive
> > information such as passwords.
>
> "Pastebin". Thats a word I've never seen before. I assume you mean upload
> the output to Google docs (say) and post a link.
> http://pastebin.com/ (and lots of others; Google "pastebin")
> I have started it. BTW: do you have a better suggestion than
> CAMEL_DEBUG=all evolution >& logfile
> This produces lots of output; I've had some trouble with the very complex and
> poorly documented evolution debug system.
Not really, since it's unclear where the problem might be. You don't
have to paste an entire session, just the last page or two whenever the
freeze or crash occurs.
BTW is this happening even when you move a single message, or only when
it's a large number?
-------- Forwarded Message --------
From: Jonathan Ryshpan <jonr...@pacbell.net>
To: Patrick O'Callaghan <p...@usb.ve>
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 18:25:52 -0700
On Tue, 2016-07-26 at 16:54 +0100, Patrick O'Callaghan wrote:
> >
> > > In that case I suggest running Evo with debugging enabled and posting
> > > the tail of output when a hang occurs (not directly here please, use a
> > > pastebin and post the URL here). Remember to remove any sensitive
> > > information such as passwords.
> >
> > "Pastebin". Thats a word I've never seen before. I assume you mean upload
> > the output to Google docs (say) and post a link.
>
> http://pastebin.com/ (and lots of others; Google "pastebin")
>
> > I have started it. BTW: do you have a better suggestion than
> > CAMEL_DEBUG=all evolution >& logfile
> > This produces lots of output; I've had some trouble with the very complex
> > and poorly documented evolution debug system.
>
> Not really, since it's unclear where the problem might be. You don't
> have to paste an entire session, just the last page or two whenever the
> freeze or crash occurs.
>
> BTW is this happening even when you move a single message, or only when
> it's a large number?
Crashes happen randomly, sometimes when I'm doing nothing much, say scrolling
through a message list. I think one happened while I was doing nothing at all.
Of course now that I'm investigating there have been no problems for a day. The
log is now 90M. I'll keep you posted.
-------- Forwarded Message --------
From: Jonathan Ryshpan <jonr...@pacbell.net>
To: Patrick O'Callaghan <p...@usb.ve>
Subject: Re: [Evolution] Crashes
Date: Tue, 26 Jul 2016 18:44:12 -0700
On Tue, 2016-07-26 at 18:25 -0700, Jonathan Ryshpan wrote:
>
> On Tue, 2016-07-26 at 16:54 +0100, Patrick O'Callaghan wrote:
> > >
> > > > In that case I suggest running Evo with debugging enabled and posting
> > > > the tail of output when a hang occurs (not directly here please, use a
> > > > pastebin and post the URL here). Remember to remove any sensitive
> > > > information such as passwords.
> > >
> > > "Pastebin". Thats a word I've never seen before. I assume you mean
> > > upload the output to Google docs (say) and post a link.
> >
> > http://pastebin.com/ (and lots of others; Google "pastebin")
> > >
> > >
> > > I have started it. BTW: do you have a better suggestion than
> > > CAMEL_DEBUG=all evolution >& logfile
> > > This produces lots of output; I've had some trouble with the very complex
> > > and
> > > poorly documented evolution debug system.
> >
> > Not really, since it's unclear where the problem might be. You don't
> > have to paste an entire session, just the last page or two whenever the
> > freeze or crash occurs.
> >
> > BTW is this happening even when you move a single message, or only when
> > it's a large number?
>
> Crashes happen randomly, sometimes when I'm doing nothing much, say scrolling
> through a message list. I think one happened while I was doing nothing at
> all.
>
> Of course now that I'm investigating there have been no problems for a day.
> The log is now 90M. I'll keep you posted.
Of course, no sooner was this written than evolution crashed. I have put the
last 1000 lines of the log file in:
https://drive.google.com/file/d/0B96WA_V8sR9bV0Jfb1lJNEdaX1U/view?usp=sharing
I was scrolling through a message list when the crash happened.
If you want more, just ask.
Enjoy - jon
_______________________________________________
evolution-list mailing list
evolution-list@gnome.org
To change your list options or unsubscribe, visit ...
https://mail.gnome.org/mailman/listinfo/evolution-list
