=== tgm4883` is now known as tgm4883 [04:34] Hi! [04:35] smartboyhw, \o [04:35] * smartboyhw waves at SergioMeneses [04:36] smartboyhw, working? [04:37] No, I don't have a job [04:37] dude, I'm 14 [04:39] smartboyhw, :O you're young [04:40] sergiomeneses: But I'm already a core member of Ubuntu Studio! [04:40] smartboyhw, great! congratz [04:41] and you're member of testing team too, rigth? [04:41] yep [04:41] Id see you early (#ubuntu-meeting) [04:42] !? [04:42] I don't understand [04:44] smartboyhw, sorry, my bad [04:44] smartboyhw, I saw you earlier at #ubuntu-meeting [04:45] OK, I'm in the QA Meeting as chair [04:45] So? [04:46] smartboyhw, I dont remember the topic xD sorry but I think so [04:46] ! [04:56] smartboyhw, I have to go, see you later [04:56] nice to see you [04:57] Bye!@ === yofel_ is now known as yofel === _salem is now known as salem_ [14:00] balloons, good morning [14:00] Hi balloons [14:01] balloons, when you can we need to talk about raid1 and the relative testcase [14:02] njin: balloons is usually online on 14:10 - 14:30:) [14:03] smarboyhw, I can see that is online, but probably cannot pay attention, so i left a message [14:04] OK, I am waiting for balloons to chat about 12.04.1 testing [14:05] I catch the ticket befor .-P [14:05] I know, you first [14:05] lol [14:05] lol! [14:05] Hi balloons! [14:06] smartboyhw, I changed it up on you.. I was around since 12:30 [14:06] Oh! [14:07] So, today is the freeze of 12.04.1 [14:07] We need a hell lot of testing to do, right? [14:10] smartboyhw, next week, next week [14:10] njin, fire away [14:10] ok [14:11] 1- we have to explain how to do the raid1 setiing up ? [14:11] in real hw and vm [14:12] actually I'm not able to do a raid in kvm, i can do only in VB [14:12] and in real hw [14:13] we also need to modify the old testcase that is wrong because actully it don't autosync anymore to avoid loss of data 8 just explainned to me xnox) [14:14] njin: you can do mdadm raid in kvm & real hw. [14:14] njin: in hw you want to do ISMS (intel matrix storage), which is available in the certification labs [14:14] which is hardware raid [14:14] njin, you should talk to SpamapS on #ubuntu-server before modifying this test case [14:14] (also known as fakeraid, cause it's not full raid controllers) [14:15] i did try to get access to the certification lab machine, but no replies so far. [14:15] ok guy, too many problems i go on with other stuff let this for last, thanks everybody [14:16] njin, heh, yes not a bad idea [14:16] xnox, can you give me a little help in this ? [14:17] njin: what help do you want? first of all which test case are you talking about? [14:17] to point the autosync situation [14:18] njin, are you working on this? http://testcases.qa.ubuntu.com/Install/ServerRAID1 [14:18] http://testcases.qa.ubuntu.com/Install/ServerRAID1 point 17 [14:19] Server basic RAID1 install, point 17.. I see it [14:19] balloons: njin: but in point 16.L it already says that you need to manually add the drives and explains why (pointer to the bug) [14:22] so njin I see the test.. I'm not sure what your confused about.. Test 17 seems to just be testing the automatic sync. Upon power-on, it should notice the drives are not the same, and sync them back up again [14:22] ok xnox and should only happen on precise or later (see bug 943397). [14:22] Launchpad bug 943397 in mdadm (Ubuntu) "cannot re-add a briefly disconnected disk to a RAID1 set" [High,Invalid] https://launchpad.net/bugs/943397 [14:23] njin: i will be testing this as part of 12.04.1 rc images which will happen today/tomorrow and i will be going step by step through instructions. [14:23] last time i did it was all correct in the test cases [14:23] but i will report any discrepancies. [14:24] I'm testing just now in quantal and it don't automatically sync [14:24] well that is PITA ain't it [14:24] njin: does it start syncing after adding it? [14:24] njin: or better [14:24] nah there is no better [14:25] sure just added it sync [14:25] it works great [14:25] just this little particular of the autosync of the reconnected device [14:30] njin, so help me understand.. You remove the disk and boot it up [14:30] everything is fine on one drive [14:30] then, step 17 you put the disk back in an boot it up, but it doesn't re-sync.. UNLESS you add the disk back to the array manually. Is this right? [14:44] balloons, yes, sorry I was out for a moment [14:45] njin, ahh ok. got it [14:45] well, I don't think that affects the testcase then.. sounds like what's stated is what should be happening [14:45] so I'd say you found a bug instead [14:50] ballons, is a new feature, https://bugs.launchpad.net/ubuntu/+source/mdadm/+bug/943397 [14:50] Ubuntu bug 943397 in mdadm (Ubuntu) "cannot re-add a briefly disconnected disk to a RAID1 set" [High,Invalid] [14:50] starting from precise [14:52] njin, ahh.. i missed that bug [14:53] well, i go on with openssh server [14:53] njin, awesome [15:44] balloons: Still here? [15:44] yes [15:44] Actually, I'm a bit afraid astraljava won't make it to chair on 29th [15:47] smartboyhw, we can ping him and as [15:47] *ask [15:48] astraljava ping [15:55] gema: jibel: do you have a minute to talk about adding extra jenkins tests? [15:56] xnox, I do, I'd like gema to participate too since they'll be deployed in utah [15:57] I am [15:57] all eyes [15:57] xnox, ok go ahead [15:58] Ok. So for the following systems (precise, quantal) x (ubuntu desktop, ubuntu server) x (i386, amd64) I would like to add a following test [15:58] install, but make the presseed file change the default filesystem type from ext4 (current default) to btrfs [15:59] this test should depend on successful completion of the typical/default test, similar to how the raid/lvm/crypto are [15:59] -- [15:59] second test [15:59] -- [15:59] for systems (ubuntu desktop, ubuntu server) x (i386,amd64) I would like the following test [16:00] -- [16:00] install precise with btrfs filesystem -> upgrade to quantal -> check that it reboots fine [16:00] -- [16:00] this one should depend on the two btrfs tests (one in precise, one in quantal) for the respective combo of (flavour, arch) [16:00] -- [16:00] what do you think? [16:00] .. [16:01] jibel: gema ^^^ [16:01] xnox: I think you just duplicated the number of jobs we currently have in jenkins with those "two" test cases [16:01] xnox, crypto = encrypted filesystem ? [16:01] :D [16:01] jibel: yes. [16:02] * xnox giggles [16:02] it is ok if we limit this to e.g. desktop only and on one arch only [16:02] or not run it too often (e.g. once a week, when everything is green) [16:02] xnox: by when do you need this? [16:02] as this is not a default installation, but it possibly might be the *next* default. [16:03] gema: precise -> quantal btrfs upgrade ideally before we release 12.04.1.... or i will be doing it manually here with 12.04.1 candidate images [16:04] in general btrfs automated testing is a workitem for quantal which was assigned to me =/ before i got loads of ubiquity work =) [16:05] do the test sound reasonable to you? [16:05] do we have capacity? or will we have capacity soon for these? [16:05] where can I record these tasks? (and implement preseeds / do initial runs) [16:05] xnox: we don't have the capacity at all, but I am trying to figure out how long it would take to do [16:05] and if we could teach your or someone else how to do it and just verify that [16:06] with the current setup [16:06] is it juju / cloud-init friendly at all? [16:06] we wouldn't be doing it with utah [16:06] ok [16:07] it would be best to do on baremetal to exercise grub + btrfs [16:07] xnox: right now we don't have manouvre time [16:08] it'll be faster to test it manually [16:08] xnox: but if that is what is going to become the default install soon [16:08] we need to start working on it for the future [16:09] btrfs comes up every uds.... and every uds we don't know how reliable it is [16:10] fedora post poned it again for a cycle [16:10] xnox, can we test reliably btrfs on VMs ? [16:11] yes, it will be at least something. [16:11] baremetal is a dream more or less [16:13] xnox: I think we need a BP for next cycle to tackle this [16:14] from the QA perspective [16:14] we will start our adding coverage to ubuntu with adding coverage to btrsf [16:14] next cycle [16:14] ok [16:14] i will create one [16:14] next [16:15] no worries, I will [16:15] ubiquity has added support for lvm & encrypted-lvm [16:15] it will live with QA and we will deliver it [16:15] do we have capacity to add for: [16:15] we will need you there to tell us what tests are relevant [16:15] ok [16:15] ..? [16:17] ubiquity has support for lvm & encrypted-lvm. [16:17] Do we have capacity to add: [16:17] - (i386,amd64) x ubuntu desktop x quantal x (lvm, ecrypted-lvm) [16:17] install test, with dependency on regular install? [16:19] to run what tests? [16:19] xnox: we have 0 capacity right now, I would need to drop something we are doing to pick this, and we cannot get there for next week for sure [16:20] xnox: so if something is important enough for quantal milestones and we need to add it, we will, but we will have to choose what to drop [16:20] gema: that install finishes & it reboots, similar to the current quantal-iso-testing-dashboard-tests [16:20] I understand. I will speak with slangasek as well about it. [16:20] xnox: which job are you talking about in particular? [16:20] ok. [16:21] our lvm and encrypted test cases are rather thin [16:21] These: https://jenkins.qa.ubuntu.com/view/Quantal/view/ISO%20Testing%20Dashboard/ [16:21] which job in particular [16:21] ubiquity lvm can replace alternative lvm: https://jenkins.qa.ubuntu.com/view/Quantal/view/ISO%20Testing%20Dashboard/job/quantal-alternate-i386_lvm/ [16:22] and https://jenkins.qa.ubuntu.com/view/Quantal/view/ISO%20Testing%20Dashboard/job/quantal-alternate-amd64_lvm/ [16:22] because the plan is to drop alternative cd [16:22] and the alternative-lvm code path is exercised with server-lvm [16:22] ok [16:22] so you want to change those two jobs to desktop? [16:22] does ubiquity support that now? [16:23] yes, it does. I implemented it =) [16:23] xnox: can we preseed it as well? [16:23] for capacity reasons we could change: https://jenkins.qa.ubuntu.com/view/Quantal/view/ISO%20Testing%20Dashboard/job/quantal-desktop-amd64_encryptedhome/ to be encrypted home + LUKS encryption as well [16:24] gema: no you cannot, I will need to do some tweks to make it preseedable. [16:24] gema: where is your preseed file for server-crypt, such that I can make that preseeding experience similar / the same. [16:24] let me find you the page [16:25] gema: the ubiquity lvm & encrypted-lvm are critical for quantal, so I think it makes sense to find room for it (e.g. replace the alternative-lvm job & replace desktop-encryptedhome job) [16:25] ... or piggy-bag on top of desktop-encryptedhome job. [16:26] xnox: https://code.launchpad.net/~ubuntu-server-iso-testing-dev <- server-tests-quantal [16:28] xnox: the jobs we are currently running for iso testing are based on those repositories [16:28] ok [16:28] thanks [16:28] np [16:28] xnox: let me know your thoughts whenever you've looked into it [16:29] gema: so lp:~ubuntu-server-iso-testing-dev/+junk/alternate-tests-quantal is for the alternate runner? [16:29] for quantal [16:29] yes [16:29] cool. [16:29] ok. I will poke them a little bit, make sure that ubiquity can preseed it's encrypted installation. [16:30] xnox: ack, then we will have the discussion with the powers that be on when to put in production what [16:30] and then we can discuss testing proposed jobs [16:30] yes [16:30] and then we can decide to flip some of them off, in favor of enabling ubiquity-encrypted tests. [16:30] ok. [16:31] great. [16:31] stgraber: that task from 12.04.1 meeting to discuss fs testing with qa people is done. Intead I now have a bunch of new tasks to actually implement it ;-) [16:32] xnox: ok :) [16:33] xnox: the capacity problems I was talking about was people to be able to make the changes to the test code and put those jobs up there [16:34] xnox: not the servers not coping [16:34] xnox: just to clarify x) [16:34] xnox: the lab is not at full capacity yet ! [16:35] gema: oh, so if I write the tests & babysit them there is capacity to *execute* them? [16:35] xnox: indeed [16:35] I am up for that! =) [16:35] if you have a jenkins job working we can add it to the bunch no problem [16:35] you'll need to send a merge request to the right branch (which can be messy as you can see) and get it reviewed and merged [16:36] and you can set up the jenkins job [16:36] indeed [16:36] do you have a test instance or a how to to get it up and running the way you do it? [16:36] xnox: do you have access to the QA lab? [16:36] gema: i think I do have vpn, but i don't have access to any machines on the other end (as far as I know) [16:36] xnox: you may not know it, but you do [16:37] can you help me with that? in private? =) [16:43] sure [16:43] where are you? [18:55] balloons: I should be able to chair on 29th, I have it on calendar. :) [18:55] astraljava, awesome ;) === jalcine is now known as Jacky === salem_ is now known as _salem [20:40] balloons, pratically remaining the 3 MAAS and the install preseed and install JeOS-KVM to complete the mandatory, then 4 run once and 1 optional [20:40] see you tomorrow, I go sleep Bye [20:41] ah the raid too [20:53] balloons: did you get called into a meeting?