Bug 959595 - in 19 Beta TC1 through TC3, DVD FileConflicts test requires huge resources (CPU and RAM)
Summary: in 19 Beta TC1 through TC3, DVD FileConflicts test requires huge resources (C...
Keywords:
Status: CLOSED EOL
Alias: None
Product: Fedora
Classification: Fedora
Component: yum
Version: 19
Hardware: All
OS: Linux
unspecified
unspecified
Target Milestone: ---
Assignee: Packaging Maintenance Team
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2013-05-03 22:18 UTC by Andre Robatino
Modified: 2015-02-10 01:43 UTC (History)
7 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2015-02-10 01:43:43 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)

Description Andre Robatino 2013-05-03 22:18:55 UTC
Description of problem:
I've been running https://fedoraproject.org/wiki/QA:Testcase_Mediakit_FileConflicts on each compose for a while now. I'm currently running it on a F18 x86_64 host. It normally takes just a few seconds to complete. Suddenly, starting with 19 Beta TC1, it requires huge resources in both CPU and RAM to complete. My machine only has 4 GiB RAM, someone who was able to complete the test told me the process was using up to 6 GiB for them. I can no longer do the test, it runs for about 15 minutes, gets to around 85%, then there is a message "Killed" (probably OOM). For someone who can do the test, it does complete normally, except for the resources required. (See bug 958512.)

My HDD suddenly went bad about a day after I ran the latest attempt, and it may not be a coincidence, so I'm not willing to run the test for more than a few seconds until it's fixed.

The script used in the test (potential_conflict.py) has not changed since last year. The sister test https://fedoraproject.org/wiki/QA:Testcase_Mediakit_Repoclosure continues to work normally (takes only a few seconds to complete).

Version-Release number of selected component (if applicable):
19 Beta TC1 DVD and later (both i386 and x86_64)

How reproducible:
always

Comment 1 Andre Robatino 2013-05-03 22:29:14 UTC
Possibly relevant: the file conflict in bug 958512 also first appeared in 19 Beta TC1. There were no file conflicts in 19 Alpha RC4 (AKA Alpha Gold).

Comment 2 Andre Robatino 2013-05-04 14:36:47 UTC
Should also mention that I've been doing this test for a while, and it has flagged other file conflicts without ever causing this spike in resource usage. Either it's something unique to this particular conflict, or the conflict has nothing to do with it.

Comment 3 Mukundan Ragavan 2013-05-04 18:45:12 UTC
Just an update. Same problems exist with TC-3.

[ ... ]

python-django-1.4.5-2.fc19.noarch
python-django14-1.4.5-5.fc19.noarch

[ ... ]

$ top | sed -n 8p

20799 root      20   0 7900m 7.4g 7740 R  90.4 47.4   8:59.11 python

Comment 4 Andre Robatino 2013-05-10 19:31:18 UTC
I am able to do the fileconflicts test on the 19 Beta TC4 i386 DVD on my box with only 2 GiB of RAM, quickly. The file conflict between the python-django* packages is gone so that may have been responsible. Would still be nice to know why that happens for those two particular packages, when there have been file conflicts before without this happening.

[andre@dell-pc tmp]$ ~/potential_conflict.py --repofrompath=media,/opt/tmp/foo1 -r media
Getting complete filelist for:
file:///opt/tmp/foo1/
565947 files found.

Looking for duplicated filenames:
5178 duplicates found.

Doing more advanced checks to see if these are real conflicts:
  5% complete (   258/  5178,   287/sec),    0 found - eta 0:00:17
 10% complete (   516/  5178,  3008/sec),    0 found - eta 0:00:08
 15% complete (   774/  5178,  1281/sec),    0 found - eta 0:00:06
 20% complete (  1032/  5178,  4027/sec),    0 found - eta 0:00:05
 25% complete (  1290/  5178,  1728/sec),    0 found - eta 0:00:04
 30% complete (  1548/  5178,  2796/sec),    0 found - eta 0:00:03
 35% complete (  1806/  5178,  2605/sec),    0 found - eta 0:00:02
 40% complete (  2064/  5178,  4581/sec),    0 found - eta 0:00:02
 45% complete (  2322/  5178,  3900/sec),    0 found - eta 0:00:02
 50% complete (  2580/  5178,  4041/sec),    0 found - eta 0:00:01
 55% complete (  2838/  5178,  2692/sec),    0 found - eta 0:00:01
 60% complete (  3096/  5178,  4277/sec),    0 found - eta 0:00:01
 65% complete (  3354/  5178,  4130/sec),    0 found - eta 0:00:01
 70% complete (  3612/  5178,  2204/sec),    0 found - eta 0:00:00
 75% complete (  3870/  5178,  4303/sec),    0 found - eta 0:00:00
 80% complete (  4128/  5178,  2806/sec),    0 found - eta 0:00:00
 85% complete (  4386/  5178,  4045/sec),    0 found - eta 0:00:00
 90% complete (  4644/  5178,  4291/sec),    0 found - eta 0:00:00
 95% complete (  4902/  5178,  1311/sec),    0 found - eta 0:00:00
100% complete (  5160/  5178,  2551/sec),    0 found - eta 0:00:00
0 file conflicts found.
1 package conflicts found.

== Package conflicts ==
2:libpng-devel-1.5.13-2.fc19.i686
libpng12-devel-1.2.50-3.fc19.i686


== File conflicts, listed by conflicting packages ==
[andre@dell-pc tmp]$

Comment 5 Andre Robatino 2013-05-11 16:52:02 UTC
The file conflict, and the huge resource use, are gone in 19 Beta TC4. Would still be nice to know if the high resource use apparently caused by the conflict is expected, or a problem in the script. The affected TCs will be available for download in https://dl.fedoraproject.org/pub/alt/stage/ at least until Final Gold.

Comment 6 Fedora End Of Life 2015-01-09 18:02:18 UTC
This message is a notice that Fedora 19 is now at end of life. Fedora 
has stopped maintaining and issuing updates for Fedora 19. It is 
Fedora's policy to close all bug reports from releases that are no 
longer maintained. Approximately 4 (four) weeks from now this bug will
be closed as EOL if it remains open with a Fedora 'version' of '19'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version.

Thank you for reporting this issue and we are sorry that we were not 
able to fix it before Fedora 19 is end of life. If you would still like 
to see this bug fixed and are able to reproduce it against a later version 
of Fedora, you are encouraged  change the 'version' to a later Fedora 
version prior this bug is closed as described in the policy above.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events. Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.


Note You need to log in before you can comment on or make changes to this bug.