Bug 1049028

Summary: [locking] dnf is not blocking
Product: [Fedora] Fedora Reporter: Nicolas Mailhot <nicolas.mailhot>
Component: dnfAssignee: Packaging Maintenance Team <packaging-team-maint>
Status: CLOSED NOTABUG QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: low    
Version: rawhideCC: akozumpl, nicolas.mailhot, packaging-team-maint, pnemade, rholy
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2014-09-09 08:37:35 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Nicolas Mailhot 2014-01-06 19:59:25 UTC
Description of problem:
# rm -fr /var/lib/spamassassin/3.003002/ ; dnf -y reinstall spamassassin ; systemctl restart spamassassin.service amavisd.service postfix.service 
metadata-cache already locked by 2190
  The application with PID 2190 is: dnf
    Mémoire : 190 M RSS (720 MB VSZ)
    Débuté  : Mon Jan  6 20:11:43 2014 - il y a 23:54
    State  : Mise en attente


Expected result: services restarted after sa fix

Actual result services restarted at the wrong time

dnf-0.4.9-1.fc21.noarch

Comment 1 Ales Kozumplik 2014-01-07 07:05:39 UTC
Nicolas, can you reproduce this consistently? What is the command line of the process with PID 2190?

Thanks.

Comment 2 Nicolas Mailhot 2014-01-07 09:21:44 UTC
(In reply to Ales Kozumplik from comment #1)
> Nicolas, can you reproduce this consistently? 

Haven't tried

> What is the command line of
> the process with PID 2190?

It was a general dnf - y update that was slowly getting metadata and had not finished yet

Comment 3 Ales Kozumplik 2014-01-07 12:02:59 UTC
Allright, so you'd expect it to wait until the lock is released, is that right?

Comment 4 Nicolas Mailhot 2014-01-07 14:36:01 UTC
Exactly. Wait till what was requested is done

Comment 5 Ales Kozumplik 2014-01-07 17:28:58 UTC
I see now, thanks. I kind of like the current behavior more but let's see how this plays out.

Comment 6 Ales Kozumplik 2014-09-09 08:37:35 UTC
Closing this---the issue has not generated much interest among other users which tells us the current behavior is preferred. A simple shell wrapper could be build that restarts the DNF process after the previous one is finished.