On Jun 24, 2009, at 6:20 AM, Sven Mueller wrote: > > Luke Kanies schrieb: >> On Jun 22, 2009, at 9:54 AM, Sven Mueller wrote: >> >>> I currently have a problem setting up fstab options without >>> interfering >>> with whatever software is currently using the mounts (or not). >>> More precisely, these are devices shared between two hosts where >>> only >>> one host can have them mounted at a time (via DRBD). >>> >>> The behaviour I'm trying to achieve is similar to what the service >>> type >>> allows: Set defaults without changing its current state (ensure vs. >>> enable). Back to the mount type, this means I want the type to allow >>> updating of /etc/fstab (or netinfo) without having an effect on the >>> current status of the mount. >>> >>> Currently the type allows for ensure: >>> absent => unmount, remove from /etc/fstab >>> present => add to/update in /etc/fstab, mount >>> >>> What I would like would be (but see below): >>> >>> absent => remove from /etc/fstab (I don't care about the unmount >>> here) >>> mounted => ensure its correctly in /etc/fstab, make sure its mounted >>> present => ensure its correctly in /etc/fstab, remount if already >>> mounted and options changed >>> perhaps: unmounted => ensure its in fstab, make sure not mounted >> >> The 'present' value should mean 'in fstab but not mounted'. Are you >> not seeing that behaviour? > > That's precisely the behaviour I see (and that is causing me > problems). > >> If that is working, do you still need a behaviour change? > > Perhaps my intention was unclear: > The behaviour I need is: > > make sure the entry is in fstab as specified, but don't touch the > actual > mounting state of the mount point (i.e. leave it unmounted if it is > not > mounted, leave it mounted if it already is, optionally remount it if > it > is already mounted but with the wrong set of options). > > This is a behaviour that I currently can't achieve withe the "mount" > type.
Ah; you want to specify the presence but not specify the mount state. Hmm. > >>> So as not to change the current semantics of the type, we could >>> introduce a new parameter: handle_mount with values: >>> >>> true => unmount/mount like before >>> false => never mount/unmount/remount >>> remount_only => only remount if mounted and options changed, >>> don't mount/unmount >>> >>> I currently worked around the issue by using a rather complicated >>> exec. >>> Should I open a ticket? >>> >>> regards, >>> Sven >>> >>> PS: While my Ruby skills certainly improved over the past weeks, I >>> still >>> don't understand types/provides in puppet fully yet... I will try to >>> take a stab at implementing this next weekend perhaps... I like the idea, but I'm not too stoked with the implementation. Having a separate attribute that's only sometimes used seems clumsy. That being said, I don't have a better idea, either. Anyone have any other ideas on how to specify this? -- Somebody has to do something, and it's just incredibly pathetic that it has to be us. --Jerry Garcia --------------------------------------------------------------------- Luke Kanies | http://reductivelabs.com | http://madstop.com --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To post to this group, send email to puppet-dev@googlegroups.com To unsubscribe from this group, send email to puppet-dev+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/puppet-dev?hl=en -~----------~----~----~----~------~----~------~--~---
