On Mon, Aug 17, 2015 at 02:23:48PM -0700, Dennis E. Hamilton wrote: > Thanks Tony, this reminds me of some questions that would help us understand > what is involved. These questions are anyone knowledgable of the current > arrangements: > > I see four levels of support to wikis and forums: > > 1. User Account and Content Administration > I believe this is handed. Does anyone believe it is not?
How do you mean handled? This function is managed within the two products by those who manage them already. In other words this is fully devolved to the AOO PMC to determine how this should be managed. > 2. Administering the Running Service > That is, a server administrator for the service (not necessarily the host) > Is this done? Do we need a replacement or an expansion here? What are > prerequisite qualifications for being able to do this. Again this is down to the PMC to determine how to best do this. Right now Infra provide a number of VMs for AOO to run the service on. This is mostly due to significant historical inertia when AOO first joined the ASF, but also because it is what the PMC wanted to make their lives easier. If you want these to be come more managed by Infra then a significant number of changes would need to be made to make this happen. Not least of which is moving the system into Puppet. This is a non-trivial piece of work which I'd estimate would take a member of paid staff at least 4-6 weeks to complete (minimum). > > 3. Administrating the Server that hosts the Service > I assume this is where one deals with Ubuntu upgrades and such, whether > the server is real or virtual. > Is this provided by the project? Do we need a replacement/expansion here > from within the project? Again, what are prerequisite qualifications? Again, due to historical inertia and what I suspect is fear of letting go in the past (perhaps even now in the present) this element was handled at the PMC's discretion - with the caveat that Infra would force an update where required (say SSL vulnerability etc), or if the PMC did not do it itself the service would be shutdown and the instance shutdown until such time that the problem had been resolved. Right now, I do not believe that these instances are being properly maintained, and are in need of significant TLC to make them tick along nicely once more. However, owing to the obscenely difficlut (and in my opinion frankly ridiculous) setup of the services this is near impossible to do well. > 4. I assume hardware IT support is not the business of the project, and > changing boxes is a different deal. Yes? Correct, the PMC is provided with a number of virtual machines (not native hardware, just to avoid any confusion). If you need to move (and frankly I would push for this very hard) we would need to stand up the new instances for you, based on an open discussion of your requirements. Infra are responsible for making sure that the instances are up, and that the hardware they run is available. > I suspect this is known. I don't know if the necessary information is > anywhere in project materials (haven't looked). > > This is an area of ignorance for me. I hope I have helped clarify somewhat. -- Many thanks, Tony
pgpEMbLdOugGK.pgp
Description: PGP signature
