Custom Query (894 matches)


Show under each result:

Results (100 - 102 of 894)

Ticket Owner Reporter Resolution Summary
#425 desai [email protected] fixed group/host specific :info files

Contents of configuration files can be controlled per group/host, but not the metadata associated with them, e.g. user/group/mode/mtime.

Different config files often have different owners (like apache vs httpd or exim vs mail), so it would be nice to allow different file metadata for different groups/hosts.

#813 desai Martin Schwier <[email protected]…> wontfix group tags in genshi bundles

I just migrated to bcfg2 1.0 and noticed that I can't use group tags in *.genshi Bundles. Is that intended?

I think it would be useful if I could use genshi in every Bundle without renaming it to *.genshi and adding the xmlns:py attribute.

Did you also notice that bcfg2-repo-validate has trouble with genshi Bundles?

#595 desai Jim Rowan <[email protected]…> fixed group categories not handled quite right

I tried something which turns out to be illegal in my group config, and was surprised with the results. I'm not sure which aspects of this are defective, if any.

from Metadata/groups.xml:

  <Group name='sles'>
    <Group name='suse'/>
  <Group name='sles9' category='sles-version'>
    <Group name='sles'/>
  <Group name='sles9sp1' category='sles-version'>
    <Group name='sles9'/>
  <Group name='sles9sp3' category='sles-version'>
    <Group name='sles9sp1'/>

Output from bcfg2-info groups:

 sles             | no       |                 | suse,rpm,linux,unix
 sles9            | no       | sles-version    | sles,suse,rpm,linux,unix
 sles9sp1         | no       | sles-version    | sles9,sles,suse,rpm,linux,unix
 sles9sp3         | no       | sles-version    | sles9sp1 

So, I have two issues.

1.) The configuration validated with no mention that I had this heirarchy of groups with the same category. Probably, some warning is appropriate at the very least.

2.) The doc says "A group can only contain one instance of a group in any category." Literally taken, this is ambiguous (I believe) with respect to the situation of a group including another instance of the same category. However, the intent seems clear and I think the first level inclusion of sles9 in sles9sp1 should have failed.

Note: See TracQuery for help on using queries.