Bug 114131 - kio_imap4 takes 100 % cpu when imap server is unreachable
Summary: kio_imap4 takes 100 % cpu when imap server is unreachable
Status: RESOLVED WORKSFORME
Alias: None
Product: kio
Classification: Frameworks and Libraries
Component: imap (show other bugs)
Version: unspecified
Platform: Gentoo Packages Linux
: NOR normal
Target Milestone: ---
Assignee: kdepim bugs
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-10-09 19:31 UTC by SatanClaus
Modified: 2009-08-05 14:24 UTC (History)
3 users (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments
Head of ~/.xsession-errors filled up with stupid KIO errors (14.52 KB, text/plain)
2006-06-16 02:30 UTC, Julian Mehnle
Details

Note You need to log in before you can comment on or make changes to this bug.
Description SatanClaus 2005-10-09 19:31:12 UTC
Version:           1.8.1 (using KDE KDE 3.4.1)
Installed from:    Gentoo Packages
Compiler:          gcc 3.3.6 
OS:                Linux

I added this bugreport as everything I found was old stuff from kde 2.2.1 or something... (and closed/resolved)

In case the imap server is offline or unreachable for some reason I get a popup telling something like:
"Connection to host xxx is broken." (where I substituted my server name with xxx).
Until then everything is fine, but after I click OK my cpu usage goes up to 100 %. Until I close kmail that's approx. 60 % kmail and 40 % kio_imap4.
In case I stay in kmail for another 10 minutes I get the same popup again... Also another imap process now appears when using top, so that they each have 30 % while kmail now uses 40 %... (I think I have to wait for 10 minutes, as I told kmail to check mails every 10 minutes).
So in case I close kmail the kio_imap4 processes stay alive. They'll use 100 % cpu when summed up (but they're killable with SIGTERM).
Comment 1 Carsten Burghardt 2006-01-06 12:12:01 UTC
Please try 3.5 (kmail 1.9.1)
Comment 2 Chris Howells 2006-05-24 04:59:28 UTC
I cannot reproduce this. Please try Carsten's suggestion and if you still have a problem reopen the bug report.

Thanks.
Comment 3 Julian Mehnle 2006-06-16 02:29:04 UTC
I experienced the exact same effect today with KMail 1.9.1 (KDE 3.5.3).  But of course there is no way to reopen this bug, and it is likely I am going to be ignored by the developers, so I wonder if me writing a detailed report and getting KMail to work properly isn't just a gross waste of time.

Because I still like KMail better than the competition (or perhaps just because I never learn), I'll try anyway...

I have configured an IMAP server (imaps://intern.link-m.de) in KMail that has been working fine for ages.  Today the server was subjected to a major software upgrade, which caused it to go down.  KMail then said:

"The connection to the server Link-M was unexpectedly closed or timed out. It will be re-established automatically if possible."

Then the kmail process started using about 70% CPU and kio_imap4 about 25%.  After a few minutes, billions(!) of simple, useless error messages started piling up in my .xsession-errors file:

"kio (KIOConnection): ERROR: Could not write data"

Billions of them, within less than half an hour.  Until my hard disk was full.  (I had to log off and terminate the X connection in order to be able to free the disk space by deleting the .xsession-errors file.)  I am attaching the head of my .xsession-errors file.

When I renamed the server in the KMail configuration to "x-intern.link-m.de", closed KMail, `kill`ed the remaining kio processes, and restarted KMail, KMail rightly complained about the invalid server name but did not start misbehaving again.  If I reinstitute the old (correct) server name, it all starts again.

Now, if you want to kill off the last bit of motivation in me for reporting bugs to KDE, feel free to again brush off this report.
Comment 4 Julian Mehnle 2006-06-16 02:30:19 UTC
Created attachment 16630 [details]
Head of ~/.xsession-errors filled up with stupid KIO errors
Comment 5 Julian Mehnle 2006-06-16 02:40:52 UTC
(No, I'm not SatanClaus <satan.claus@gmx.net>.  It just occurred to me my posting could be mistaken to that degree.  I just seem to have encountered the same problem as him.)
Comment 6 Oren Held 2009-05-16 13:35:31 UTC
This ancient bug seems to come back to live, still happens to me on kde 4.2.2
Comment 7 Rian Schmits 2009-08-05 14:24:29 UTC
happens to me too and am also on kde 4.2.2 (on kubuntu 9.04)