Bug 1310287 - rpm update should not restart server
Summary: rpm update should not restart server
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Fedora
Classification: Fedora
Component: mariadb
Version: 24
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Jakub Dorňák
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2016-02-20 00:55 UTC by Dennis Jacobfeuerborn
Modified: 2016-02-24 15:27 UTC (History)
5 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2016-02-24 13:27:31 UTC
Type: Bug


Attachments (Terms of Use)

Description Dennis Jacobfeuerborn 2016-02-20 00:55:08 UTC
Apparently the current server rpm restarts the server when the rpm gets updated. Given that this could be a really heavy weight operation this is a really bad idea since this process can take a very long time and if something happens during that time you end up with a half finished yum transaction.
Also a call to mysql_upgrade call might be necessary between a stop and restart of the database. 
Given that only then admin has the knowledge when and how to restart the database properly and the package can only make one-size-fits-all decisions this automatic restart should be removed from the package.

Comment 1 Terje Røsten 2016-02-22 21:47:21 UTC
It's common pattern for server packages in fedora to *restart* themselves during
rpm/yum/dnf upgrades, however scripts will not *start* a stopped service. 

If you don't want the automatic restart, stop the server service before doing packages upgrade and start after package tools have done their jobs.

Comment 2 Honza Horak 2016-02-24 13:27:31 UTC
(In reply to Dennis Jacobfeuerborn from comment #0)
> Apparently the current server rpm restarts the server when the rpm gets
> updated. Given that this could be a really heavy weight operation this is a
> really bad idea since this process can take a very long time and if
> something happens during that time you end up with a half finished yum
> transaction.

I agree with Terje here, this is the behavior users are used to in Fedora and it makes sense to keep all services behave the same. Also I don't here any other admins who would have problem with it.

> Also a call to mysql_upgrade call might be necessary between a stop and
> restart of the database. 

This is usually required only when updating to new major version, which we don't change during one release, so mysql_upgrade is usually run after system upgrade (e.g. from F23 to F24).

> Given that only then admin has the knowledge when and how to restart the
> database properly and the package can only make one-size-fits-all decisions
> this automatic restart should be removed from the package.

At the same time, admin is also responsible for updating and when mariadb-server should not be updated, he must say this to DNF and run the RPM update of mariadb once database is prepared for it.

Comment 3 Jan Kurik 2016-02-24 15:27:26 UTC
This bug appears to have been reported against 'rawhide' during the Fedora 24 development cycle.
Changing version to '24'.

More information and reason for this action is here:
https://fedoraproject.org/wiki/Fedora_Program_Management/HouseKeeping/Fedora24#Rawhide_Rebase


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