Bug 677605

Summary: Traceback when shutting down
Product: Red Hat Enterprise Linux 6 Reporter: Alexander Todorov <atodorov>
Component: anacondaAssignee: Ales Kozumplik <akozumpl>
Status: CLOSED ERRATA QA Contact: Release Test Team <release-test-team>
Severity: urgent Docs Contact:
Priority: urgent    
Version: 6.1CC: akozumpl, ddumas, jzeleny, syeghiay
Target Milestone: beta   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: anaconda-13.21.99-1 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2011-05-19 12:37:40 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Bug Depends On:    
Bug Blocks: 670159    
Attachments:
Description Flags
the traceback
none
install.log
none
install.syslog
none
anaconda.log
none
anaconda.program.log
none
anaconda.storage.log
none
anaconda.syslog
none
anaconda.xlog
none
anaconda.yum.log
none
screenshot (possibly misleading)
none
new traceback with updates.img none

Description Alexander Todorov 2011-02-15 10:54:33 UTC
Created attachment 478852 [details]
the traceback

Description of problem:
There's a traceback when anaconda is shutting down:
IOError: [Errno 4] Interrupted system call. See attached screenshot for more details.


Version-Release number of selected component (if applicable):
anaconda-13.21.97-1.el6.x86_64

How reproducible:
always

Steps to Reproduce:
1. perform install in GUI mode and reboot
2.
3.
  
Actual results:
some traceback on the console

Expected results:
clean shutdown

Additional info:
There's more of the traceback/error messages but I'm unable to catch all of them. I'll try setting up serial console to grab all the messages.

Comment 1 Chris Lumens 2011-02-15 16:17:35 UTC
Yeah, we could definitely use the other log files for this one.  Also, do you know what tree this started showing up in?  Nothing changed from .96 to .97 that should cause this sort of problem.

Comment 2 Alexander Todorov 2011-02-15 16:34:24 UTC
Created attachment 478909 [details]
install.log

Comment 3 Alexander Todorov 2011-02-15 16:34:32 UTC
Created attachment 478910 [details]
install.syslog

Comment 4 Alexander Todorov 2011-02-15 16:34:44 UTC
Created attachment 478911 [details]
anaconda.log

Comment 5 Alexander Todorov 2011-02-15 16:34:53 UTC
Created attachment 478912 [details]
anaconda.program.log

Comment 6 Alexander Todorov 2011-02-15 16:35:03 UTC
Created attachment 478913 [details]
anaconda.storage.log

Comment 7 Alexander Todorov 2011-02-15 16:35:21 UTC
Created attachment 478914 [details]
anaconda.syslog

Comment 8 Alexander Todorov 2011-02-15 16:35:29 UTC
Created attachment 478915 [details]
anaconda.xlog

Comment 9 Alexander Todorov 2011-02-15 16:35:40 UTC
Created attachment 478916 [details]
anaconda.yum.log

Comment 10 Ales Kozumplik 2011-02-15 16:56:39 UTC
Created attachment 478922 [details]
screenshot (possibly misleading)

I have a screenshot of some traceback. Alex, is that what you saw?

Comment 11 Alexander Todorov 2011-02-16 07:11:19 UTC
Ales,
in short YES.

The screenshot from comment #0 is the start of the traceback. Then there is a little more text which I couldn't capture and then comes the blue background with the text captured in comment #10. 

IMHO the last lines are just consequence of the first traceback but I could be wrong.

Comment 12 Ales Kozumplik 2011-02-16 13:49:21 UTC
Related: bug 604614. The way processes are shutdown upon reset has changed slightly. The tracebacks are from the second anaconda fork that continues after metacity has died (because our handling of metacity termination is broken at this point).

Alex, can you please retest with this updates image:
http://akozumpl.fedorapeople.org/bz677605.img

Comment 13 Alexander Todorov 2011-02-17 07:48:38 UTC
Created attachment 479277 [details]
new traceback with updates.img

This is new traceback when using the updates.img. It says:

ImportError: No module named ui.gui

Comment 14 Ales Kozumplik 2011-02-17 09:13:06 UTC
Alex I am unable to reproduce the traceback from comment 13 on a 0215 nightly compose. Can you please retry on a recent compose and see if the problem is gone with the updates image?

Comment 15 Alexander Todorov 2011-02-17 14:30:46 UTC
I get the same error as in comment #13 with the 0217.n.0 nightly and the updates.img from comment #12.

This is on KVM guest.

Comment 16 Ales Kozumplik 2011-02-18 12:27:34 UTC
Patch posted for review:
https://www.redhat.com/archives/anaconda-devel-list/2011-February/msg00167.html

I was not able to hit the tracebacks from comment 13, but using the patch I was repeatedly able to finish GUI installation on virt and bare-metal. I consider the comment 13 traceback unrelated.

Alex, if you see the comment 13 traceback in an anaconda where the current bug is fixed please open a new BZ, with anaconda.log attached (because of the command line), describing what phase of install you were able to spot the error (you can switch to tty1 one or two times during the installation to check the errors do not actually appear before restart) and whether you used F12 or a mouse click to activate the "Reboot" button on the last screen.

Thanks.

Comment 17 Ales Kozumplik 2011-02-21 08:44:58 UTC
Fixed by e72a83ff83918ef809eeba78c9c490d77900c89d.

Comment 18 Ales Kozumplik 2011-02-21 09:25:36 UTC
Fixed on master by 58e76d37ed8e55e4ac8f5e02febbc242a6a3bc4e.

Comment 21 Alexander Todorov 2011-03-01 12:56:03 UTC
I'm unable to reproduce on 0224.2 build. Tried several times but no traceback. Moving to VERIFIED for now.

Comment 22 errata-xmlrpc 2011-05-19 12:37:40 UTC
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2011-0530.html