Bug 1223803 - dnf seems to be sleeping forever during installation
Summary: dnf seems to be sleeping forever during installation
Keywords:
Status: CLOSED INSUFFICIENT_DATA
Alias: None
Product: Fedora
Classification: Fedora
Component: dnf
Version: 23
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Packaging Maintenance Team
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2015-05-21 13:18 UTC by Zdenek Kabelac
Modified: 2022-08-31 09:09 UTC (History)
8 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2015-07-22 13:11:49 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)
lsof | grep dnf (22.94 KB, text/plain)
2015-05-21 13:30 UTC, Zdenek Kabelac
no flags Details


Links
System ID Private Priority Status Summary Last Updated
Red Hat Bugzilla 1233134 0 unspecified CLOSED dnf update hangs the system and causes filesystem corruption? 2021-02-22 00:41:40 UTC

Internal Links: 1233134

Description Zdenek Kabelac 2015-05-21 13:18:15 UTC
Description of problem:

Todays  'dnf upgrade' seems to have lots of problem with dnf.
So where am I now - 


  Upgrading   : wine-arial-fonts-1.7.43-1.fc23.noarch   276/1108 
  Upgrading   : vte-profile-0.40.2-1.fc23.x86_64        277/1108 
  Upgrading   : vte291-0.40.2-1.fc22.x86_64             278/1108 
  Upgrading   : perl-HTTP-Tiny-0.056-1.fc23.noarch      279/1108 
  Upgrading   : mesa-filesystem-10.6.0-0.devel.6.5a55f68.fc23.x86_64  280/1108 
  Upgrading   : man-pages-4.00-1.fc23.noarch            281/1108 

**Waiting here** forever?

TOP shows nothing:
%Cpu(s):  4,8 us,  1,3 sy,  0,0 ni, 93,7 id,  0,2 wa,  0,0 hi,  0,0 si,  0,0 st

Everything in idle (except for mozilla browser).

pstree:

bash---su---bash---dnf

Nothing forked from 'dnf'  - so it sleeps INSIDE dnf.

Is this 'dnf'  tested before being pushed to rawhide?
(glad we still have at least yum-deprecated)

I wanted to attach  'gdb' to python - but with no success - freezed as well.
Thus the only thing I could see is from strace:

futex(0x7f00cf75f624, FUTEX_WAIT, 1

Here is kernel stacktrace:
dnf             S ffff880078d87c28     0 15769  10479 0x00000080
ffff880078d87c28 ffff880078c44590 ffff8800677327c0 ffffc9000066cac8
ffff880078d87fd8 ffff8800677327c0 0000000000000000 ffffc9000066cac0
ffff880078d87d00 ffff880078d87c48 ffffffff81783f77 ffffc9000066cac0
Call Trace:
[<ffffffff81783f77>] schedule+0x37/0x90
[<ffffffff811174e8>] futex_wait_queue_me+0xc8/0x140
[<ffffffff811177a9>] futex_wait+0x119/0x290
[<ffffffff810ab372>] ? get_signal+0xf2/0x610
[<ffffffff81119c55>] do_futex+0x145/0xab0
[<ffffffff81013547>] ? do_signal+0x37/0x760
[<ffffffff810ab24b>] ? ptrace_notify+0x5b/0x90
[<ffffffff81022e56>] ? tracehook_report_syscall_exit+0xd6/0x120
[<ffffffff8111a641>] SyS_futex+0x81/0x190
[<ffffffff810238db>] ? syscall_trace_enter_phase1+0x14b/0x1b0
[<ffffffff81013cef>] ? do_notify_resume+0x7f/0xa0
[<ffffffff81788089>] system_call_fastpath+0x12/0x17


I really think such important package like 'dnf' needs some testing before it lands even in rawhide - 'we don't care what happens - it's rawhide, it's your fault you are using it' is really a wrong way here...

Version-Release number of selected component (if applicable):
dnf-1.0.0-1.fc23.noarch

How reproducible:


Steps to Reproduce:
1.  dnf upgrade   (for rawhide on 2015-05-21)
2.
3.

Actual results:


Expected results:


Additional info:

Comment 1 Zdenek Kabelac 2015-05-21 13:30:44 UTC
Created attachment 1028180 [details]
lsof | grep dnf

Possibly  dnf  died because  it has been upgraded in the middle of execution ?

Is  'dnf' trying to convince user they need to upgrade  in 'special' environment.

Comment 2 Zdenek Kabelac 2015-05-21 13:44:22 UTC
So as a great bonus  after 'killing'  dnf process (since nothing else was helpful) I'm left with wonderful rpm DB state:

# LC_ALL=C rpm -qa
error: rpmdb: BDB0113 Thread/process 18156/140297016870656 failed: BDB1507 Thread died in Berkeley DB library
error: db5 error(-30973) from dbenv->failchk: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 -  (-30973)
error: cannot open Packages database in /var/lib/rpm
error: rpmdb: BDB0113 Thread/process 18156/140297016870656 failed: BDB1507 Thread died in Berkeley DB library
error: db5 error(-30973) from dbenv->failchk: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 -  (-30973)
error: cannot open Packages database in /var/lib/rpm

Comment 3 Honza Silhan 2015-06-12 13:05:28 UTC
Zdenek, can you upload `/var/log/dnf.log`, please, so we can try to reproduce it with the same package set? Still I think the most probably it was caused by some scriplet or the error inside RPM process. Post the `rpm -q rpm` too. Thanks.

Comment 4 Jan Kurik 2015-07-15 14:07:35 UTC
This bug appears to have been reported against 'rawhide' during the Fedora 23 development cycle.
Changing version to '23'.

(As we did not run this process for some time, it could affect also pre-Fedora 23 development
cycle bugs. We are very sorry. It will help us with cleanup during Fedora 23 End Of Life. Thank you.)

More information and reason for this action is here:
https://fedoraproject.org/wiki/BugZappers/HouseKeeping/Fedora23

Comment 5 Honza Silhan 2015-07-22 13:11:49 UTC
I talked to Zdenek via IRC and unfortunately the log were rotated. I am sorry but we can;t reproduce it and didn't know the cause.


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