== Agenda == * Review ACTION points from previous meeting. * Review progress made on the [[ServerTeam/Roadmap| Roadmap]]. * [[StableReleaseUpdates|SRU]] tracking on server-related packages. * Preparing a 2009 edition of the Server Survey? - nijaba * Server tips - How to implement? - nijaba * Roadmap refresh. * Open Discussion. * Agree on next meeting date and time. == Minutes == ## Use title4 (ie ==== ) for each section of the minutes ## Only topics discussed during the meetings should be put in the minutes. ## Status reporting is done via another channel. ==== SRU tracking on server-related packages ==== ttx brought up the problem of making sure that 'fix released' bugs were properly reviewed for SRU worthiness. The missing part today is someone making sure everything relevant is nominated. The proposal is to get a list of fixed bugs in the past week, review it in the server meeting and nominate relevant ones. ACTION: mathiaz to come up with a way to generate a list of fixed bugs in the last week for packages relevant to the ubuntu server team. ==== Preparing a 2009 edition of the Server Survey ==== nijaba announced it had almost been a year since the first survey had been launched. He started to update the question set for this September. Anyone interested in helping should have a look at the [[https://wiki.ubuntu.com/ServerTeam/Survey|wiki page]] and file bugs against the [[https://bugs.launchpad.net/server-survey/|project]] with suggestions. ACTION: nijaba to write a blog post asking for review of the ubuntu-server survey questions. ==== Ubuntu HA team: Cluster stack update ==== RoAkSoAx worked on updating ipvsadm to 1.25 which includes support for ipv6. ivoks is still working on the pacemaker update. ACTION: RoAkSoAx to update the Roadmap section of the Ubuntu HA team with a short description of how the cluster stack would look like in karmic ==== Cloud power management ==== kirkland released powernap and powerwake as two opensource pieces of the puzzle. Hooks have been added to eucalyptus to use powernap to suspend/hibernate/poweroff cloud nodes when unused and powerwake to resume them when demand requires. UEC is now a "breathable" cloud. kirkland added that powernap was actually generally useful for servers: sort of like a screensaver, but instead of saving your screen, you're saving your powerbill (and the planet). He is interested in other ideas, use cases, servers where powernap can be used. ==== Agree on next meeting date and time ==== Next meeting will be on Tuesday, July 28th at 15:00 UTC in #ubuntu-meeting. == Log == {{{ [16:01] #startmeeting [16:01] Meeting started at 10:01. The chair is mathiaz. [16:01] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [16:01] o/ [16:01] Today's agenda: https://wiki.ubuntu.com/ServerTeam/Meeting [16:02] thanks ttx for running the meeting last week [16:02] ttx: and publishing the minutes [16:02] * stgraber waves [16:02] you're welcome. [16:02] Last week minutes: https://wiki.ubuntu.com/MeetingLogs/Server/20090714 [16:02] [TOPIC] # [16:02] SRU tracking on server-related packages. [16:02] New Topic: # [16:03] [TOPIC] SRU tracking on server-related packages. [16:03] New Topic: SRU tracking on server-related packages. [16:03] ttx: what's you proposal on this subject? [16:04] We need to better track Server-related SRU candidates. Basically monitor the development release updates for changes that could make sense in previous releases [16:04] to define where we want to use our resources to push them to completion. [16:05] So someone could monitor the list, look up the changes and list relevant ones on a wikipage. They would get discussed at the team meeting [16:05] I would be willing to help in that effort. [16:05] once approved, we would assign them to the ubuntu-server team and make sure they pass the process. [16:05] ttx: in my experience SRU candidates are mainly coming from bugs [16:05] I currently look through the SRU list for updates relevant to servers, mainly for LTS releases though as we mainly use LTS builds. [16:05] mathiaz: yes. [16:06] ttx: I'm not sure that monitoring karmic-changes would help [16:06] ttx: the problem we're trying to solve is to get more SRU completed or to identify more candidates for SRU? [16:07] mathiaz: I would say, to make sure we SRU all relevant bugs [16:07] and give a forum where their appropriateness would get discussed [16:07] ttx: right - so I'd rather focus on the ubuntu-server-bugs mailing list then [16:07] but you're right that -changes monitoring is not enough. [16:07] ttx: it seems that karmic-changes is to noisy [16:07] * kirkland cheers ttx on [16:07] We need to catch all FixReleased. [16:08] ttx: agreed - for a specific list of packages (relevant to the server team) [16:08] Does a list of packages relevant to the server team exist? [16:09] (Or is it just common sense?) [16:09] Omahn: https://bugs.launchpad.net/~ubuntu-server/+packagebugs [16:09] sbeattie: how do yo track relevant SRU bugs? [16:10] sbeattie: is there a process in the bug team to track which bugs are currently under an SRU process? [16:10] That doesn't mean we'll do a lot more SRUs, just give some confidence that we are not missing anything important. [16:10] ttx: ok - so there is already a list of bugs that are targeted for SRU [16:10] ttx: what's we're trying to achieve here is to get more relevant bugs on that list? [16:11] mathiaz: you mean the bugs nominated for a release ? [16:11] ttx: yes [16:12] ttx: if someone thinks a bug should be fixed in a stable release, one of the first step is to nominate that bug for the target release [16:12] Basically I want to find a way to have a team discussion on what nominated bugs should be approved. [16:13] and assign the key ones to the team rather than expect that the original poster has the nerves to push them to completion. [16:13] ttx: ok [16:13] ttx: https://bugs.launchpad.net/ubuntu/hardy/ [16:13] ttx: this is the list of bugs related to hardy [16:13] unfortunately the current nomination acceptation process is a one-man decision right now [16:14] ttx: and you can search for all bugs to which the ubuntu-server is a subscriber [16:15] ttx: http://tinyurl.com/nmxppq [16:15] mathiaz: can you see those where hardy has been nominated but not accepted ? [16:15] ttx: ^^ this is the list of bugs *accepted* for hardy to which the ubuntu-server team is subsribec [16:15] ttx: good question - sbeattie or slangasek would know the answer about it [16:16] ttx: hm - there is a review nominations link in the page above [16:16] ttx: https://bugs.launchpad.net/ubuntu/hardy/+nominations [16:17] mathiaz: so we could use the nomination process to track them. [16:17] ttx: http://tinyurl.com/lm7ogv [16:17] ttx: and ^^ this is the url relevant to the ubuntu-server tema [16:17] mathiaz: /someone/ would track bugs and make sure nominations are filed were relevant [16:18] and we would review the nominations in the team meeting [16:18] then potentially assign the key ones to the team [16:18] ttx: right. The process may be semi-automated by processing the ubuntu-server-bugs mailing list archive [16:19] ttx: and present a list of bugs marked 'Fix released' on a period of time [16:19] bdmurray: do you have a script that does this^? [16:19] The missing part today is someone making sure everything relevant is nominated. [16:19] ttx: agreed. [16:20] ttx: so what we could do is to have a weekly review in this meeting about the bugs that have been fixed released in the last week [16:20] ttx: and make sure the relevant ones are nominated, approved, assigned, etc... [16:21] mathiaz: I fear that might take too long ? [16:21] mathiaz: what if someone volunteered to track FixReleased and turn them all to nominations where appropriate ? [16:21] why not have like a weekly posting to the ubuntu server mailing list like the kernel-team [16:21] then we could review only those that make sense to be SRUed === zul_ is now known as zul [16:22] ttx: then the burden to accept the nominations would fall on the ubuntu-sru team [16:22] ttx: how about trying to setup the list and review it in the next meeting [16:22] mathiaz: you mean only ubuntu-sru should accept a nomination ? I think most devs can. [16:23] ttx: and put a maxmimum time we spend on it (say 10 mn) [16:23] ubuntu-sru accept the SRU more tan the nomination [16:23] ttx: that way we get the ball rolling and see if we need to make some adjustement in the process [16:23] mathiaz: worksforme [16:23] ttx: right - -devs can accept nominations [16:24] ttx: IIUC we're trying to come up with a way to make sure that *all* relevant bugs end up in the nomination process [16:25] yes. a bug might be valid but not make a good SRU. That's the -sru teams decisions. [16:25] accepting nominations is more ack-ing that the issue applies to a given release. [16:25] ttx: and the proposal is to get a list if fixed bugs in the last week, review it in the server meeting and nominated relevant ones [16:26] I would even nominate/accept them. [16:26] ttx: mathiaz: can we experiment on libvirt and kvm? both of these are in jaunty-proposed at the moment [16:26] I mean, experiment on the publicity, call-for-testing [16:27] kirkland: ah right. I'm not sure how that is relevant to the discussion here. [16:27] kirkland: testing comes after the SRU has been proccessed [16:27] kirkland: ie one of the very last stage of the SRU process. [16:27] [TOPIC] SRU tracking on server-related packages. [16:27] kirkland: we're focusing here on the very first step of the process [16:27] okay [16:28] i think the last step is important too :-) testing in -proposed, which I suspect very little happens [16:28] kirkland: how to make sure relevant bugs don't fall off the ubuntu-server radar. [16:28] kirkland: yes - definetly [16:28] kirkland: this is probably the subject of another topic [16:28] well, we should also weekkly-review the status of the ones we committed to do. [16:28] let's break things down in manageable chunks [16:29] ttx: right - so let's move on [16:29] but I agree that's another part of the SRU puzzle. [16:29] I think we have a plan [16:29] I'll talk to bdmurray to see if/how we can get a list of fixed bugs [16:30] [ACTION] mathiaz to come up with a way to generate a list of fixed bugs in the last week for packages relevant to the server team [16:30] ACTION received: mathiaz to come up with a way to generate a list of fixed bugs in the last week for packages relevant to the server team [16:30] anything else on this particular topic? [16:31] nope - let's move on [16:31] anything else regarding last week minutes/meeting? [16:32] nope - let's move on [16:32] [TOPIC] Preparing a 2009 edition of the Server Survey [16:32] New Topic: Preparing a 2009 edition of the Server Survey [16:32] nijaba: ^^? [16:32] yes, it has been a year since we launched our first survey [16:32] or it will be in sept [16:33] so it is time to start updating our question set for this coming sept [16:33] anyone interested in helping could have a look at the wiki [16:33] and file bugs against the project with suggestions? [16:34] nijaba: which wiki page? [16:34] https://wiki.ubuntu.com/ServerTeam/Survey [16:35] nijaba: so v1.0.4 was used as the final set of questions for the last survey? [16:35] also, the limesurvey package needs to be updated. if anyone feels up to it... [16:35] mathiaz: yep [16:36] nijaba: great - seems like everything is there to get the survey updated. [16:36] nijaba: any changes needed to the policy? [16:36] mathiaz: not that I know of [16:36] nijaba: great. [16:36] mathiaz: but it could use a review if anyone feels up to it [16:37] nijaba: any changes to the Launch plan? [16:37] nijaba: or is this something to be discussed later? [16:37] nijaba: also - what is the timeframe for the launch? [16:37] mathiaz: good question. It can be discussed later, but from the last survey, 50% of the respondent came from the banner on ubuntu.com [16:38] mathiaz: end of sept is my target [16:38] nijaba: did you get in touch with the ubuntu-news/marketing team? [16:38] mathiaz: not yet [16:39] ok - any specific action item to record on this topic? [16:40] mathiaz: just the call for review... [16:40] nijaba: writing a blog post? [16:41] mathiaz: good idea... i will [16:41] nijaba: great. [16:41] [ACTION] nijaba to write a blog post asking for review of the ubuntu-server survey questions. [16:41] ACTION received: nijaba to write a blog post asking for review of the ubuntu-server survey questions. [16:41] anything else on this topic? [16:41] not from me [16:42] ok - let's move on then [16:44] [TOPIC] Roadmap refresh [16:44] New Topic: Roadmap refresh [16:44] what have you been up to in the next week for karmic? [16:45] what have you been up to in the *last* week for karmic? [16:45] RoAkSoAx: ^^? [16:45] mathiaz, working on ipvsadm (updating it to 1.25) which includes support for ipv6 [16:46] though I have some doubts about how to handle the CFLAGS in debian/rules, so if someone is an expert on the subject and has some free time to help me, please let me know :) [16:46] nice. [16:46] RoAkSoAx: have you asked in ubuntu-motu? [16:46] RoAkSoAx: you may be able to find some help there [16:47] RoAkSoAx: what else is needed to get packages in shape for karmic? [16:47] then I already have a testing howto for heartbeat-pacemaker to do a call for testing, however I'm waiting to have latest pacemaker which I believe ivoks was working on. [16:47] mathiaz, I did asked -motu with no luck :) [16:48] RoAkSoAx: ok - is there a wiki page somewhere to track what needs to be done? [16:49] mathiaz, in the Ubuntu-HA team no. We don't have a wikipage for that [16:49] RoAkSoAx: there is a roadmap section on https://wiki.ubuntu.com/UbuntuHighAvailabilityTeam [16:49] RoAkSoAx: could this section be updated with the plan for karmic? [16:50] RoAkSoAx: just give a list of relevant packages that need to be updated in the archive [16:50] mathiaz, sure. Will work on it. [16:51] RoAkSoAx: and a short description of how the cluster stack would look like in karmic [16:51] RoAkSoAx: great thanks. [16:51] will do :) [16:51] [ACTION] RoAkSoAx to update the Roadmap section of the Ubuntu HA team with a short description of how the cluster stack would look like in karmic [16:51] ACTION received: RoAkSoAx to update the Roadmap section of the Ubuntu HA team with a short description of how the cluster stack would look like in karmic [16:53] kirkland: how is powernap going? [16:53] mathiaz: i've been working on cloud power management [16:53] powernap + powerwake are two opensource pieces of it [16:53] going well [16:53] uploaded, promoted to main [16:53] dan nurmi of eucalyptus has [16:54] added hooks into eucalyptus [16:54] to use powernap to suspend/hibernate/poweroff cloud nodes when unused [16:54] and powerwake to resume them when demand requires [16:54] kirkland: is there anything to be tested? [16:54] UEC is now a "breathable" cloud [16:54] well, i think powernap is actually generally useful for servers [16:55] sort of like a screensaver, but instead of saving your screen, you're saving your powerbill [16:56] kirkland: and the planet [16:56] i'm interested in other ideas, use cases, servers where powernap can be used [16:56] mathiaz: yes, and on a more grand and noble basis, the planet :-) [16:56] mathiaz: i'm actually using it on my mythfrontend machines [16:56] kirkland: Are there any check mechanisms that you would like help with implementing? [16:56] mathiaz: suspending those when they're not doing anything [16:56] Daviey: heh, actually, there is at least one :-) [16:56] Daviey: i have a mechanism that watches for activity on devices (so /dev/tty1 /dev/ptmx and friends) [16:56] kirkland: right - and resuming them as fast as possible when you have an urgent need to watch a movie [16:57] Daviey: and one that watches ps/2 keyboards/mice on /proc/interrupts [16:57] ^^ Also support USB keyboards? [16:57] Daviey: i need a way of watching a usb keyboard/mouse with activity *not* happening on a console [16:57] ahh.. too quick :) [16:57] Daviey: so, usb keyboard/mouse activity in gnome, for instance [16:57] low priority, since the server doesn't include X [16:57] kirkland: ok - we're running out of time [16:57] but would be nice to have [16:58] powernap is python code, btw! [16:58] kirkland: do you have a place to track the list of improvment? [16:58] kirkland: grab you later :) [16:58] all right - time to wrap up [16:58] mathiaz: launchpad [16:58] [TOPIC] Agree on next meeting date and time [16:58] New Topic: Agree on next meeting date and time [16:59] next week, same time, same place? [16:59] +1 [17:00] great then. next week, same time, same place. [17:00] and happy alpha3 testing! [17:00] #endmeeting [17:00] Meeting finished at 11:00. }}}