Hi all,
congratulations on the RedHat/CentOS announcement. Sorry for responding so late on SIGs: I was in China until recently and had no time to follow up. There is definitely interest to continue the Xen (or Xen4CentOS) SIG within CentOS, building on the Xen4CentOS work of last year. I want to get the endorsement of the Xen Project Advisory Board for joining the SIG at next Tuesday's board meeting. This will give an application (if needed) extra weight. I don't expect any issues.
The other question I have is whether we do need to re-apply (http://wiki.centos.org/SpecialInterestGroup lists Xen4CentOS as a SIG already) and what the format for an application would be other than posting to this list. Maybe we ought to look at the naming of the SIG and make it more generic to cover for future version changes in Xen as well as targeting CentOS versions beyond CentOS 7 (which ought to be a lot easier than CentOS 6 because we only need Xen and no custom kernel).
Also, there would probably be practical issues on list naming, etc. (Xen4CentOS is merged with CentOS-virt from a list perspective). I don't have an issue with the status quo, but it may be cleaner for CentOS in the long run if there was a clear SIG to list mapping.
Best Regards Lars
hi Lars,
On 01/17/2014 12:45 PM, Lars Kurth wrote:
congratulations on the RedHat/CentOS announcement. Sorry for responding so late on SIGs: I was in China until recently and had no time to follow up. There is definitely interest to continue the Xen (or Xen4CentOS) SIG within CentOS, building on the Xen4CentOS work of last year. I want to get the endorsement of the Xen Project Advisory Board for joining the SIG at next Tuesday's board meeting. This will give an application (if needed) extra weight. I don't expect any issues.
thanks, and let us know how that goes.
The other question I have is whether we do need to re-apply (http://wiki.centos.org/SpecialInterestGroup lists Xen4CentOS as a SIG already) and what the format for an application would be other than posting to this list. Maybe we ought to look at the naming of the SIG and make it more generic to cover for future version changes in Xen as well as targeting CentOS versions beyond CentOS 7 (which ought to be a lot easier than CentOS 6 because we only need Xen and no custom kernel).
imho, its worth going down the route of setting up a formal SIG in the present scope of things, although the code + content and release stuff is already out there in centos.org ( which might also make it a lot faster and easier, since we just need to realign the git repo and get some build metadata around it ).
Also, there would probably be practical issues on list naming, etc. (Xen4CentOS is merged with CentOS-virt from a list perspective). I don't have an issue with the status quo, but it may be cleaner for CentOS in the long run if there was a clear SIG to list mapping.
This came up recently in other conversations as well, the thinking was that as a sig matures and has community traction we can split lists away to be more sig focused. I'd imagine the entire Cloud / virt / hypervisor SIGs could start from the centos-virt list as homebase and then branch out. Having said that, most of the content on the centos-virt list at the moment IS xen specific.
regards,
KB, let's see whether we can get any more feedback on this thread. I will pick this up after the board meeting and we can discuss some more stuff at FOSDEM too. I don't have a problem just formalizing Xen4CentOS as it is : I just wanted to raise the question, as now is probably the right time to verify assumptions. Lars
On 17/01/2014 17:07, Karanbir Singh wrote:
hi Lars,
On 01/17/2014 12:45 PM, Lars Kurth wrote:
congratulations on the RedHat/CentOS announcement. Sorry for responding so late on SIGs: I was in China until recently and had no time to follow up. There is definitely interest to continue the Xen (or Xen4CentOS) SIG within CentOS, building on the Xen4CentOS work of last year. I want to get the endorsement of the Xen Project Advisory Board for joining the SIG at next Tuesday's board meeting. This will give an application (if needed) extra weight. I don't expect any issues.
thanks, and let us know how that goes.
The other question I have is whether we do need to re-apply (http://wiki.centos.org/SpecialInterestGroup lists Xen4CentOS as a SIG already) and what the format for an application would be other than posting to this list. Maybe we ought to look at the naming of the SIG and make it more generic to cover for future version changes in Xen as well as targeting CentOS versions beyond CentOS 7 (which ought to be a lot easier than CentOS 6 because we only need Xen and no custom kernel).
imho, its worth going down the route of setting up a formal SIG in the present scope of things, although the code + content and release stuff is already out there in centos.org ( which might also make it a lot faster and easier, since we just need to realign the git repo and get some build metadata around it ).
Also, there would probably be practical issues on list naming, etc. (Xen4CentOS is merged with CentOS-virt from a list perspective). I don't have an issue with the status quo, but it may be cleaner for CentOS in the long run if there was a clear SIG to list mapping.
This came up recently in other conversations as well, the thinking was that as a sig matures and has community traction we can split lists away to be more sig focused. I'd imagine the entire Cloud / virt / hypervisor SIGs could start from the centos-virt list as homebase and then branch out. Having said that, most of the content on the centos-virt list at the moment IS xen specific.
regards,
On 17/01/2014 17:07, Karanbir Singh wrote:
hi Lars,
On 01/17/2014 12:45 PM, Lars Kurth wrote:
congratulations on the RedHat/CentOS announcement. Sorry for responding so late on SIGs: I was in China until recently and had no time to follow up. There is definitely interest to continue the Xen (or Xen4CentOS) SIG within CentOS, building on the Xen4CentOS work of last year. I want to get the endorsement of the Xen Project Advisory Board for joining the SIG at next Tuesday's board meeting. This will give an application (if needed) extra weight. I don't expect any issues.
thanks, and let us know how that goes.
That is approved. Although I have not had time to write up and post the meeting minutes
The other question I have is whether we do need to re-apply (http://wiki.centos.org/SpecialInterestGroup lists Xen4CentOS as a SIG already) and what the format for an application would be other than posting to this list. Maybe we ought to look at the naming of the SIG and make it more generic to cover for future version changes in Xen as well as targeting CentOS versions beyond CentOS 7 (which ought to be a lot easier than CentOS 6 because we only need Xen and no custom kernel).
imho, its worth going down the route of setting up a formal SIG in the present scope of things, although the code + content and release stuff is already out there in centos.org ( which might also make it a lot faster and easier, since we just need to realign the git repo and get some build metadata around it ).
OK.
I guess the next step then is to work on a formal proposal for the board. Is there a template or list of questions that need to be answered? Is there something else we need to wait for?
On the packaging/dependency between SIGs/variants side, the type of challenges are probably quite similar to the ones in the storage SIG. Maybe we can also move away from the custom kernel we had for Xen4CentOS6 and see whether the the CentOS7 kernel may work better.
Regards Lars
On 02/10/2014 10:03 AM, Lars Kurth wrote:
That is approved. Although I have not had time to write up and post the meeting minutes
awesome!
The other question I have is whether we do need to re-apply (http://wiki.centos.org/SpecialInterestGroup lists Xen4CentOS as a SIG already) and what the format for an application would be other than posting to this list. Maybe we ought to look at the naming of the SIG and make it more generic to cover for future version changes in Xen as well as targeting CentOS versions beyond CentOS 7 (which ought to be a lot easier than CentOS 6 because we only need Xen and no custom kernel).
imho, its worth going down the route of setting up a formal SIG in the present scope of things, although the code + content and release stuff is already out there in centos.org ( which might also make it a lot faster and easier, since we just need to realign the git repo and get some build metadata around it ).
OK.
I guess the next step then is to work on a formal proposal for the board. Is there a template or list of questions that need to be answered? Is there something else we need to wait for?
I will try and get that together today evening.
On the packaging/dependency between SIGs/variants side, the type of challenges are probably quite similar to the ones in the storage SIG. Maybe we can also move away from the custom kernel we had for Xen4CentOS6 and see whether the the CentOS7 kernel may work better.
right, and the layers of repo's is going to need resolving - things like xen and the storage components we'd want as low as possible so that as many people as possible can consume them from outside the xen repos.
-
Karanbir Singh -
Lars Kurth