Re: LVM packages up for adoption
On Tue, 18 Jan 2005 17:11:50 +0000, Tim Cutts wrote:
>
> On 18 Jan 2005, at 4:06 pm, Patrick Caulfield wrote:
>
>> On Tue, Jan 18, 2005 at 03:46:18PM +0000, Tim Cutts wrote:
>>>
>>> On 17 Jan 2005, at 5:42 pm, Bastian Blank wrote:
>>>
>>>> On Mon, Jan 17, 2005 at 09:28:56AM +0000, Patrick Caulfield wrote:
>>>>> lvm2 - in active development, upstream helpful but often
>>>>> busy.
>>>>> device-mapper - largely stable. occasional releases.
>>>>> lvm10 - stable. no more upstream development at all.
>>>>> lvm-common - native package. small number of bugs need sorting
>>>>> out
>>>>> multipath-tools - in active development, upstream very helpful.
>>>>
>>>> I'm interrested onm co-maintaining lvm2 and device-mapper.
>>>
>>> As am I - we use these heavily on some fairly serious kit at work, so
>>> I
>>> can justify the time... co-maintaining sounds like a sensible thing to
>>> do.
>>>
>>
>> So how about you three co-maintain lvm2 & devmapper (and maybe
>> lvm-common ? it's
>> as much part of LVM as the lvm2 package really), and I'll hang onto
>> lvm10 &
>> multipath.
>
> Sounds good to me. I'll be able to help you with testing
> multipath-tools too; that and lvm2 are the principal bits we use (we
> don't use Debian device-mapper stuff because we build our own kernels
> from scratch)
>
> We use this stuff on both IA64 (HP rx4640) and i386 (HP DL360/380,
> mostly) architectures to talk to our dual-fabric SAN (HP StorageWorks
> HSV110 controllers on the back)
>
> How should we coordinate this?
>
My recommendation would be an LVM alioth project, w/ a svn or arch
(preferred) repository. I've kept track of lvm2 stuff in arch for a
number of years, it has worked well.
Patrick, it might even be worth all 4 of us maintaining all the LVM
related packages (throwing lvm10 in with the rest), since Tim uses
multipath-tools, and none of us care much for lvm10.
Reply to: