Bug 739327

Summary: device-mapper-multipath doesn't need to be in Base
Product: [Fedora] Fedora Reporter: Adam Williamson <awilliam>
Component: compsAssignee: Bill Nottingham <notting>
Status: CLOSED CURRENTRELEASE QA Contact:
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 16CC: clumens, dlehman, rvokal
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: 2011-09-19 20:49:55 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Adam Williamson 2011-09-17 18:04:43 UTC
Currently, comps has device-mapper-multipath as 'default' in Base. However, this package is only of any use on systems with multipath storage devices (a very small minority), and anaconda already has logic to install it when necessary:

pyanaconda/storage/devices.py

...
class MultipathDevice(DMDevice):
    """ A multipath device """
    _type = "dm-multipath"
    _packages = ["device-mapper-multipath"]
    _services = ["multipathd"]
    _partitionable = True
    _isDisk = True
...

AIUI, the _packages line adds the package to the install set, and the _services line enables the service.

Given this, is there any reason the package also needs to be in base? This is a problem because the maintainer wants to enable it on upgrades when converting from sysv to systemd, so actual multipath systems will continue to work, but because the package has been in Base for a while, most people even without multipath devices will have the package installed, and hence will get a useless service starting up when they upgrade to F16.

Comment 1 Adam Williamson 2011-09-17 18:05:22 UTC
CC a couple of anaconda folks to make sure the code in question is known to *actually work* (esp with systemd).

Comment 2 Chris Lumens 2011-09-18 02:23:24 UTC
As long as chkconfig still works, this should be fine.  However as far as known to work, I don't know that anyone has sat down and verified.

Comment 3 Bill Nottingham 2011-09-19 20:49:55 UTC
Dropped. Long-term, this should go in a more general storage group, but, eh... string change.