Thursday, 2014-03-27

StevenKtchaypo: That may break toci00:00
tchaypoyeah, and it may break existing users to00:00
StevenKActually00:01
tchaypoeg, I've always run the scripts from ~/src/tripleo-incubator so that's where my existing scripts are00:01
tchaypothe next time I run after this change, they won't be found00:01
tchaypoI could check ./ first and fall back to $TRIPLEO_ROOT/triple-incubator only if they're not in the current directory?00:01
openstackgerritSteve Kowalik proposed a change to openstack-infra/tripleo-ci: No longer source undercloudrc  https://review.openstack.org/8325100:02
lifelessthats what I'm about to ask for in my review of it00:02
*** giulivo has quit IRC00:02
StevenKtoci uses $TRIPLEO_ROOT/triple-incubator anyway00:02
StevenKlifeless: You forgot to remove the source when you moved the overcloud to the seed. :-)00:03
lifelessStevenK: no, that was deliberate00:03
lifelessStevenK: we need to exercise the undercloud slightly00:04
lifelessStevenK: just not really formed up00:04
StevenKlifeless: Exercise how, since I'd be happy to munge that patch into that form00:05
tchaypoactually it's probably better to put the backwards-compatibility into the setup-*-passwords scripts - let them move the file across to $TRIPLEO_ROOT if they can find it at the old location. Then  the devtest_*cloud scripts will call setup-*-passwords and be guaranteed to find the file at the right locaiton afterwards00:05
StevenKlifeless: Was patch set 15 of no-more-backwards-compat another rebase?00:06
openstackgerritSteve Baker proposed a change to openstack/os-collect-config: When extracting deployments, don't delete original  https://review.openstack.org/8324800:06
lifelessStevenK: boot something on it with heat00:06
lifelessStevenK: without the full overcloud overhead00:07
lifelessStevenK: one example would be to deploy undercloud.qcow2 in the undercloud00:08
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325000:11
*** lblanchard has joined #tripleo00:12
*** noslzzp has joined #tripleo00:18
tchaypoIf anyone wants to take a look at my latest comment on https://bugs.launchpad.net/tripleo/+bug/1297324 there's a decent chance you'll spot something I missed and get to laugh at me00:18
uvirtbotLaunchpad bug 1297324 in tripleo "iptables element scripts don't work on Ubuntu" [Low,Incomplete]00:18
*** lblanchard has quit IRC00:18
*** matsuhashi has joined #tripleo00:19
*** lblanchard has joined #tripleo00:19
*** lblanchard has quit IRC00:20
lifelesstchaypo: nope, focusing on the actual problem is good :)00:23
*** lucas-afk has quit IRC00:25
*** epim has quit IRC00:26
*** lblanchard has joined #tripleo00:36
*** CaptTofu_ has quit IRC00:37
*** lblanchard has quit IRC00:37
lifelessdevananda: https://review.openstack.org/#/c/74626/00:40
devanandalooks00:41
lifelessdevananda: wanting a fact check on my interpretation of the situation00:42
lifelesstranslated programming interfaces scare the hebes out of me00:43
devanandalifeless: AIUI, only the message strings are translated00:44
devananda_not_ the exception names00:44
devanandaeg, look at https://www.transifex.com/projects/p/ironic/viewstrings/#pt_BR/ironic-translations/1887595000:45
lifelessah ok then!. but ... you filed a bug about clients not needing to match on lots of different things00:45
lifelesswhich I just reviewed and -1'd a patch for00:45
lifelessso I'm a little confused :)00:45
devanandahttps://github.com/openstack/ironic/blob/master/ironic/common/exception.py#L14900:46
devanandanvm00:46
devanandaso00:46
devanandaactually, compare that list of exceptions to00:46
devanandahttps://github.com/openstack/python-ironicclient/blob/master/ironicclient/exc.py00:47
devanandaactually, really nvm00:48
devanandai'm not sure where the conversion is(n't) happening00:48
lifelesslol00:48
devanandahttps://github.com/openstack/python-ironicclient/blob/master/ironicclient/common/http.py#L16700:50
lifelessjogo: https://review.openstack.org/#/c/74600/00:50
*** rpodolyaka1 has joined #tripleo00:50
*** mestery has quit IRC00:51
devanandaand https://github.com/openstack/python-ironicclient/blob/master/ironicclient/exc.py#L15000:52
devanandaso... there *is* a mapping00:52
devanandaand it's probably incomplete00:52
devanandainstead of matching __name__, things should be added t othat mapping00:52
devanandabut i hadn't dug into it and that was a quick fix00:52
*** rpodolyaka1 has quit IRC00:55
*** blamar has joined #tripleo00:57
lifelessdevananda: so, open question, should we run an overcloud test job on top of Ironic ?01:02
lifelessright now I just configured seed+undercloud which tells us01:02
lifelessa) we can build images01:02
lifelessb) that heat deploy of Ironic works (or doesn't)01:02
*** eghobo has quit IRC01:02
lifelessc) that we can deploy to baremetal with it (or the udnercloud deploy would barf)01:03
lifelessWe don't know if the deployed undercloud really works (and an overcloud test wouldn't tell us that as we run those against the seed)01:03
lifelessbut it would test multiple node deploys for ironic vs just the one that is the undercloud01:03
devanandalifeless: quick answer then i need to run01:04
devanandalifeless: ideally i'd run temepst on that undercloud01:04
*** sgrasley_ has joined #tripleo01:05
devanandalifeless: but in tripleo's case, the equivalent is "deploy the overcloud"01:05
devanandatesting what's IN the overcloud has no benefit to ironic aFAICT01:05
devanandaall tat ^ presumes you'd be delploying said overcloud on bare metal01:05
lifelessdevananda: right, thats why I skipped the overcloud teset01:06
devanandaso you're looking at a total of 4 machines for the etest01:06
lifelessdevananda: we don't do any on-real-metal jobs yet, they're all deployed to kvm vms pretending to be bm'01:06
devanandaah, right01:06
devanandaso same thing applies, actually, to VMs01:06
lifelessright :)01:06
devanandatseting multiple operations in parallel01:06
devananda:)01:06
lifelessso tempest in the jenkins slave against the deployed undercloud makes sense to me01:07
lifelesswe can register 3 nodes with it to deploy stuff to01:07
devanandalifeless: caveat that tempest does not have support today for a limited set of nova tests01:07
devanandaand ironic does not (and probably will not ever) support the full nova feature matrix01:07
lifelessdevananda: you can give it fairly arbitrary test lists01:07
lifelessdevananda: see the tempest element we have01:08
devanandayes. adam_g is working on this for gate support for the virtual-ironic job01:08
devanandain principle, the same list should work in both virtual-ironic and tripleo-ci-ironic01:08
devanandaok - gotta run :)01:08
Lotus907efianyone know how to make a GPG public key available so when building images from an internal Ubuntu repo with unique keys the system will quit complaining about "the public key is not available"01:09
Lotus907efiI am trying to run devtest_overcloud.sh and load packages from a private internal repo01:10
Lotus907efiit loads some packages but errors out on others01:10
lifelessLotus907efi: Private repo or private mirror ?01:10
Lotus907efiand they are all supposed to be signed with the same keys so I am corn-fused01:11
tchaypoapt-key is your friend01:11
tchaypohttp://askubuntu.com/questions/291035/how-to-add-a-gpg-key-to-the-apt-sources-keyring01:11
lifelessLotus907efi: e.g. is it customised content, or is it just a mirror  of Ubuntu ?01:11
tchaypobut if it's just a mirror I'm worried about why the key isn't already in the key ring01:11
lifelessif its just a mirror, the key should be in the image being build already01:11
Lotus907efiit is a private repo that jeb maintains for the cloud infrastructure systems01:11
lifeless^5 tchaypo01:11
Lotus907efiso not really a mirror01:12
lifelessLotus907efi: ok, so you need to get the key from jeb, and you'll need to write a new element, or an enhancement to the dpkg element, to support adding keys01:12
Lotus907efiI have the key01:12
*** noslzzp has quit IRC01:12
*** sgrasley has quit IRC01:12
*** rpodolyaka has quit IRC01:12
*** matsuhashi has quit IRC01:12
*** nati_ueno has quit IRC01:12
*** Matt2 has quit IRC01:12
*** d0ugal has quit IRC01:12
*** zigo has quit IRC01:12
*** StevenK has quit IRC01:12
*** jrist has quit IRC01:12
*** john-n-seattle has quit IRC01:12
Lotus907efihmmm,01:12
lifelesstime to write a patch then :)01:13
lifelessas tchaypo says, apt-get is the tool to use01:13
tchaypos/apt-get/apt-key/01:13
tchaypoand just to pre-empt the usual01:13
Lotus907efihmmm, so I have to dig down into the elements that come with diskimage-builder?01:13
tchaypos/devstack/devtest/01:13
lifelessyou'll need to use the extra-data.d hook to copy it in - see diskimage-builder/README.md for lots of information on how these things work01:13
lifelessLotus907efi: you're adding a usecase we haven't had, so yes :)01:13
Lotus907efiah, wonderful01:13
lifelessLotus907efi: should be 10-15 lines01:14
Lotus907efioh01:14
Lotus907efiok01:14
lifelessand two scripts01:14
lifelessone to copy the key in01:14
lifelessand one to apply it in pre-install.d01:14
Lotus907efihmm, I am still puzzled why some of the packages were accepted and others were not01:15
Lotus907efithey are all supposed to be signed with the same key01:15
lifelessapt archive validation doesn't look at per package signatures01:16
lifelesswhat happens is the repo's top level metadata is signed01:16
lifelessand from there down the tree its transitively hashed01:16
Lotus907efiah, ok01:17
*** rpodolyaka has joined #tripleo01:20
Lotus907efisounds like a job for tomorrow01:21
Lotus907efisince it is 7:20PM here01:21
lifelessnight :)01:23
lifelessSpamapS: btw shiny for you https://review.openstack.org/#/c/83188/01:40
jogolifeless: so we are still using the old libvirt as a minumum 0.9.8 (unless that changed under me)01:40
*** fandi has quit IRC01:41
*** eghobo has joined #tripleo01:41
jogolifeless:  re: https://review.openstack.org/#/c/74600/01:41
jogowhich makes me inclined to say -1 to that patch01:41
lifelessjogo: why does that prevent changing the default ?01:41
*** Lotus907efi has quit IRC01:41
jogolifeless: because we say this works for x and then our confid won't. But I guess that is what the DocImpact is for01:44
lifelessjogo: it passes CI01:44
lifelessjogo: so either we don't test block migration (in which ase we don't actually say it works at all) or we test it and it works on 0.9.8.01:44
*** Lotus907efi has joined #tripleo01:45
*** nosnos has joined #tripleo01:45
*** blamar has quit IRC01:45
*** eghobo has quit IRC01:45
jogolifeless: we don't test block migration01:46
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Update documentation.  https://review.openstack.org/8274101:46
lifelessjogo: so we don't actually know if the current values work either.01:46
lifelessjogo: quod erat demonstratum01:46
jogolifeless: well people tell us they work or don't01:46
lifelessjogo: right, which I am - they don't work [well] today :)01:46
jogowhich version is needed for this01:47
lifelessthere are a raft of blog posts describing the change in my patch01:47
lifelessI don't know, noone who has blogged has bothered to specify a minimum from what I recall01:47
lifelessok, I have to run, lunch and removing dustsheets etc.01:47
openstackgerritSteve Kowalik proposed a change to openstack/tripleo-incubator: Add baremetal bridge support to devtest_testenv  https://review.openstack.org/8326701:47
*** lifeless has quit IRC01:48
*** matsuhashi has joined #tripleo01:49
*** nati_ueno has joined #tripleo01:49
*** Matt2 has joined #tripleo01:49
*** d0ugal has joined #tripleo01:49
*** zigo has joined #tripleo01:49
*** StevenK has joined #tripleo01:49
*** jrist has joined #tripleo01:49
*** john-n-seattle has joined #tripleo01:49
openstackgerritSteve Kowalik proposed a change to openstack/tripleo-incubator: Add baremetal bridge support to devtest_testenv  https://review.openstack.org/8326701:51
StevenKYay Freenode01:52
StevenKMar 25, 2014 12:00PM SYDNEY, AU Cleared Customs02:03
StevenKDHL, why do you taunt me so? :-(02:03
*** fandi has joined #tripleo02:07
*** CaptTofu has joined #tripleo02:15
openstackgerritDan Prince proposed a change to openstack/diskimage-builder: Make stable-interface-names its own element.  https://review.openstack.org/8327102:18
openstackgerritDan Prince proposed a change to openstack/tripleo-incubator: Revert "Fedora uses em1 for undercloud/overcloud"  https://review.openstack.org/8327502:24
openstackgerritDan Prince proposed a change to openstack/diskimage-builder: Make stable-interface-names its own element.  https://review.openstack.org/8327102:26
*** vkozhukalov has joined #tripleo02:30
*** lifeless has joined #tripleo02:30
lifelessand I'm back02:33
*** lifeless has quit IRC02:35
*** matsuhashi has quit IRC02:37
*** matsuhashi has joined #tripleo02:38
*** sseago has quit IRC02:46
*** matsuhashi has quit IRC02:47
*** marun has joined #tripleo02:50
*** killer_prince has quit IRC02:50
*** lazy_prince has joined #tripleo02:51
*** lazy_prince is now known as killer_prince02:51
StevenKtchaypo: I replied on a review with some clues as to where the openstack tools live.02:51
*** lifeless has joined #tripleo02:53
*** nosnos has quit IRC03:04
*** sseago has joined #tripleo03:06
*** ramishra has joined #tripleo03:11
*** cwolferh has quit IRC03:11
*** ruagair has quit IRC03:26
*** cwolferh has joined #tripleo03:27
*** ruagair has joined #tripleo03:28
*** fandikurnia has joined #tripleo03:32
*** ramishra_ has joined #tripleo03:35
*** blamar has joined #tripleo03:35
*** blamar has quit IRC03:36
*** ramishra has quit IRC03:37
notqbuild out the undercloud with a laptop, then ha, then scale out. that's so neat. heh.03:40
lifelessnotq: thanks ;)03:41
notqi'm catching up on everything crash coursing, a ton of neat ideas. :)03:43
*** nosnos has joined #tripleo03:48
*** matsuhashi has joined #tripleo03:48
*** yamahata has joined #tripleo03:51
*** fandikurnia has quit IRC03:53
*** fandi has left #tripleo03:53
*** fandikurnia has joined #tripleo03:54
*** ramishra_ has quit IRC03:57
*** ramishra has joined #tripleo03:57
openstackgerritSteve Kowalik proposed a change to openstack-infra/tripleo-ci: ensure-test-env now uses devtest_testenv  https://review.openstack.org/8328503:58
*** untriaged-bot has joined #tripleo04:00
untriaged-botUntriaged bugs so far:04:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129048804:00
uvirtbotLaunchpad bug 1290488 in tripleo "Baremetal: Invalid credentials" [Undecided,Incomplete]04:00
*** untriaged-bot has quit IRC04:00
*** yamahata has quit IRC04:05
lifelessnotq: well, welcome :)04:06
lifelessnotq: are you planning on development, or using openstack (or both?)04:06
lifelessStevenK: 79051 has a really weird error on it04:07
lifelessStevenK: as in WTF04:07
notqlifeless, i've been a sysadmin for a long time, not much of a developer although i can read code.04:09
notqread through the bug list and git changes to try to understand better.04:09
notqi'd be happy to develop provided i can provide value, however that isn't my focus.04:10
StevenKlifeless: Oh?04:10
notqlooking at potentially doing a large bare metal install04:11
openstackgerritSteve Kowalik proposed a change to openstack/tripleo-incubator: Store seed details in the JSON  https://review.openstack.org/7905104:17
tchaypoStevenK: Oh, i know where the tools live, my poinit was more that I don't think an assumption that the user has them installed locally is in keeping with our docs04:20
*** eghobo has joined #tripleo04:20
StevenKtchaypo: But the script was source'ing devtest_variables before running glance04:21
tchaypoho hum.04:21
tchaypoTheres some small tidgy chance I was overlooking that fine detail.04:21
* tchaypo trusts that sarcasm is evident04:22
lifelessStevenK: go look at it :)04:23
lifelessnotq: cool04:23
*** ruagair has quit IRC04:26
*** ruagair has joined #tripleo04:27
lifelessdevananda: re I9e910bbb82b2ab456ed15af4dee2b67491d6e24b - I thought there was an exact match scheduler for nova-bm ?04:28
lifelessdevananda: did it get removed?04:29
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325004:36
tchaypoStevenK: I've fixed the quoting but kept the separate lines for if/then to be consistent with style within the file04:37
tchaypobut if you feel strongly about it I'm happy to tweak the whole file (ie, the single other occurrence) to be smashed04:37
StevenK% git grep -l '^then$'04:40
StevenKscripts/setup-overcloud-passwords04:40
StevenKscripts/setup-undercloud-passwords04:40
StevenKscripts/write-tripleorc04:40
StevenKClearly, they are special snowflakes04:41
* StevenK takes an axe to write-tripleorc04:41
StevenKBut not in the way I want to, sadly.04:41
* tchaypo will fix the two passwords scripts04:44
*** ramishra has quit IRC04:44
*** matsuhashi has quit IRC04:45
openstackgerritSteve Kowalik proposed a change to openstack/tripleo-incubator: Use if; then style in write-tripleorc  https://review.openstack.org/8328904:45
greghaynestchaypo: Why is it preferred to write the passwords file in the tripleo-incubator repo rather than in just $TRIPLEO_ROOT?04:46
greghaynesSeems like mucking up the repo is less than ideal if not necessary04:46
tchaypoplease explain why writting seedrc/undercloudrc/overcloudrc there is okay, but the passwords not okay04:47
greghaynesI think thats also less than idea :)04:47
greghayneser, ideal04:47
tchaypowhich is to say that I agree with you, but we have precedent in that we write the rc files there; and it seems to me preferable to foul up a single known location than to have turds scattered hither and yon04:47
greghaynesugh04:47
greghaynesthats ugh as in I agree and dislike the result04:48
tchaypoas lifeless has said, what we really want is to have a heat json file that has all the things, that heat can read and we can interrogate via jq04:48
greghaynesyes, so that json file doesnt live in a repo :)04:48
greghaynesit gets written to $TRIPLEO_ROOT IIRC04:49
StevenKIt does04:49
tchaypoyes. $TRIPLEO_ROOT seems like the better place04:49
tchaypowait, do the rc files get written there?04:49
tchaypoplease tell me I was wrong and they do04:49
tchaypoit04:49
greghaynesno, they are in the incubator repo04:49
tchaypoit'll make me so happy04:49
tchaypobah04:49
*** matsuhas_ has joined #tripleo04:49
StevenKHah04:49
tchaypobut $TE_DATAFILE is in $TRIPLEO_ROOT, right?04:49
StevenKClearly, tchaypo can't be happy.04:49
StevenKtchaypo: It is04:49
greghaynesok, I just wanted to sanity check that there wasnt a strong reason to put in repo - ill comment in review so we can hash it out there04:50
tchaypoI think that $TE_DATAFILE is as important a precedent as the location of the rc files. How about I tweak my change to use $TRIPLEO_ROOT/ ?04:50
StevenKThe images get written to $TRIPLEO_ROOT, too04:50
StevenKtripleorc too04:50
greghaynestchaypo: I prefer that, yes04:50
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325004:51
tchaypoignore that one04:51
greghaynesaaah you push changes faster than I can comment :p04:51
lifelessgreghaynes: tchaypo: we don't write the rc files at all04:53
lifelessthey are static04:53
greghayneseven better04:53
lifelessif you're referring to the per cloud rc fiels04:53
lifelesswe have queries in them04:54
lifelesse.g. to testenv.json04:54
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325004:54
greghaynestchaypo: ty ty04:54
tchaypoIt's strange how at-home some throughtful review makes me feel04:57
tchaypoI didn't realise how much I missed it until I started getting it again.04:57
StevenKAtlassian doesn't do that?04:57
lifelessspeaking of04:58
lifelessthey're trolling nzers for recruits04:58
lifelessI need to send them a polite 'we weren't right for each in 2009, whats changed' mail :P04:58
lifelesss/mail/reply/04:59
tchaypoI'm sure atlassian does that in development teams, but I wasn't on a dev team04:59
lifelessdevops!04:59
StevenKlifeless: Surely then they reply with "Everything! Now talk to us."04:59
tchaypoalthough even there - my understanding, based on things I picked up during my on-boarding, was that the general practice is for post-commit code review - often many weeks post-commit05:00
tchaypoI'm sure that varied from team to team and from time to time05:00
StevenKToo late.05:00
tchaypoin fact, I know for a fact that some teams did pre-commit reviews05:00
*** Rakesh5 has joined #tripleo05:00
StevenKSure, let's close the barn after the horse is in the next state.05:00
lifelessits certainly embracing velocity05:00
lifelessbut its odd for an org that makes 'agile' products to be so disconnected from the culture they're selling to05:01
lifelessIMNSHO05:01
tchaypomy feeling is that it's generaly a false economy05:01
*** marun has quit IRC05:01
tchaypoyou might gain speed, but you spend more time circling back over the same ground, so your overall velocity tends to be lower05:01
lifelessfor me its a communications thing05:02
lifelessits far to easy for folk (like me :)) to go completely off reservation if we aren't talking with other folk and convincing them about direction (or being convinced)05:02
tchaypothere there are emergencies where getting a fast fix out, but generally I'd prefer to get it solid the first time05:03
tchaypowith the caveat that I've seen people (including me) mistake "solid" for "perfect" and end up not committing anything at all, which can be even worse than committing the wrong thing05:03
*** rpodolyaka1 has joined #tripleo05:04
tchaypolifeless: you're 3 steps ahead of us at the best of times, if we help to keep you on course it's only because you're dragging us along in your wake05:04
tchaypolike sea-anchors05:05
lifelesstchaypo: flattery will get you nowhere05:05
lifeless:)\05:05
tchayponow i shall relocate myself up to the library and start attempting to grok devtest_undercloud.sh05:07
tchaypoactually, before I do that05:07
tchaypoWould it make sense to update our docs on devtest.html to suggest that people "mkdir -p ~/.cache/tripleo; cd ~/.cache/tripleo; git clone tripleo-incubator"?05:08
tchaypoat the moment I find myself "cd ~/src/tripleo-incubator; source scripts/devtest_variables.sh", but everything I do after that runs out of ~/.cache/tripleo/tripleo-incubator05:08
tchaypowhich is confusing when I start editing scripts inside ~/src and then wondering why they don't get executed05:08
StevenKOne of the scripts already populates it05:08
StevenKI tend to apply diffs using git apply to the cached copy05:09
greghaynestchaypo: Left you one more minor nit :)05:09
tchaypoyea, but step 2 of http://docs.openstack.org/developer/tripleo-incubator/devtest.html doesn't work05:10
StevenKgreghaynes: I am disappointed you didn't comment on my eval use05:10
greghaynesoh no, where!05:10
tchaypobecause devtest_variables added $TRIPLEO_ROOT to the path, but there's nothing there to run05:10
tchaypoit works the *second* time, but the user has to Just Know(TM) to run it from their checkout Just This One Time05:10
StevenKgreghaynes: It's already landed, scripts/register-nodes05:10
greghaynesoh my05:11
greghaynesI have to say, yours is a lot worse than mine :p05:11
lifelesstchaypo: so what I do is05:12
lifelessmkdir tripleo; export TRIPLEO_ROOT=~/tripleo05:12
lifelessbecause hidden in .cache sucks05:12
lifelessI really dislike that we do that at all05:12
StevenKgreghaynes: I don't like it either, but having six almost identical lines offended more.05:12
*** matsuhas_ has quit IRC05:12
greghaynesyep. Welcome to the fun house05:12
StevenKWeeeeeel, it is a cache of sorts05:13
lifelessStevenK: the elements and heat templates are direct inputs05:13
lifelessStevenK: dib should be installable these days05:13
lifelessStevenK: client-tools is a workaround for distros (best case) up to 6 months behoind05:14
lifelessStevenK: though I take your point. But - do you enjoy editing stuff in .cache ?05:14
lifelessStevenK: where it's not concurrency safe05:14
tchaypolifeless: and if a user does that *checks* it looks like write-tripleorc will eventually save that somewhere...05:14
lifelessSo we have an rc file devtest sources05:15
lifelessI'd kind of like to have a mandatory rc file to feed it05:15
lifelesswhich could be two lines or something05:15
lifelessand would then mean you can run two tripleo runs at once05:15
*** cwolferh_ has joined #tripleo05:16
StevenKlifeless: So, I sometimes fall into the trap of editing in ~/openstack/openstack/tripleo-incubator and then wondering why my changes didn't work05:16
lifeless'Your membership in ubuntumembers is about to expire05:16
lifeless'05:16
tchaypohttp://docs.openstack.org/developer/tripleo-incubator/devtest_variables.html kinda refers to creating a moral equivalent of ~/.devtestrc05:16
lifelessStevenK: right, thats the part of the fail05:16
tchaypobut on the other hand, the docs we present to the user mention ~/tripleo, which obscures what the script actually does if you run it05:17
clarkblifeless ++ libraries exposing options lulwat05:17
StevenKlifeless: *Every* time I got that e-mail when I was working on LP, I would create a branch to fix the stupid e-mail and then die due to the horribleness.05:17
lifelessStevenK: you don't want reminders?05:17
StevenKNo, the e-mail lies.05:17
lifelessok thats brilliant - Membership renewed until 2015-04-01.05:18
lifelessI get to renew on apr 1st05:18
StevenK"We will tell you when you have expired." which is a lie. "We will spam you every day until you give in and press the button or leave the team in disgust." is closer.05:18
*** cwolferh has quit IRC05:19
tchaypoon a different note - the docs in devtest.sh include "git checkout <stable-branch>", which has always confused me, as I can't find any stable branches05:20
StevenKThey're all stable05:21
StevenK*cough*05:21
greghaynesthey are all equally (un)stable05:21
tchaypookay, rephrase. I can't find any branches, full stop05:21
lifelesstchaypo: I believe thats aspirational05:21
lifelessclarkb: you are welcome ;)05:21
StevenKlifeless: I liked your idea, generate it during sdist.05:23
tchaypogenerate what during sdist?05:23
StevenKSample config files05:23
StevenKSee d19ecd1b5b8e192c896ac1b2711465568222db53 in nova for the fun05:24
StevenKOh god. I love that the current storm cell that's about to hit Sydney streches to far north Queensland05:26
tchaypololwut05:26
StevenKLook at the national radar, it's pretty much one cell that covers the east coast05:27
tchaypohttp://www.bom.gov.au/products/national_radar_sat.loop.shtml05:27
tchaypolooks like a tropical low to me, based on the satellite pics05:27
tchaypobut http://www.bom.gov.au/australia/charts/4day_col.shtml says it isn't05:28
lifelesshttp://www.jillesvangurp.com/2013/02/13/puppet/05:29
*** matsuhashi has joined #tripleo05:29
greghaynesoh, puppet05:31
greghaynesSpeaking of local places trolling for recruits ;)05:32
tchaypogive nigel a cuddle from me05:32
greghayneslol05:33
clarkbgreghaynes: heh, after my interview many a year ago I never heard back ever. Somehow I escaped05:33
greghaynesclarkb: yea, somehow I avoided assimilating to that whole crew05:34
greghaynescant say the same for a large percentage of our class :p05:35
tchaypo "sometimes A really comes before B if I specify it in that order in one file" - iirc that's a language feature as of 3.3 or so05:35
tchaypobut I don't know how well documented that is05:35
clarkbtchaypo: it is and I think its for 3.0 and on05:35
tchayposaying that puppet is more verbose than bash for trivial examples misses the point. If you only ahvea  trivial example you don't need puppet05:35
tchaypobut - a huge +1 to jilles' first point about how abysmal puppet is as a language05:37
tchaypoit smells like a tool written by sysadmins with no dev experience bigger than a moderate-size bash script05:38
greghayneshaha, hilarious because watching the people who were getting hired there that is basically what put me off from pursuing them :)05:39
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Add some clarity to the first-time user experience  https://review.openstack.org/8329405:40
tchayposo there you go, I done a thing. Please give it your numbers.05:40
tchaypoor even comments if you like :)05:40
* greghaynes rolls 3 sided die05:40
tchayporeading jilles' about-me page makes me homesick for .nl, which is odd because I was only there for ~12 days05:41
openstackgerritGregory Haynes proposed a change to openstack/tripleo-incubator: Add OVERCLOUD_NOTCOMPUTESCALE support  https://review.openstack.org/8329605:53
lifeless.oO05:54
* lifeless clicks05:54
greghaynesits not that exciting05:54
greghaynesI just got tired of rebasing05:55
tchaypogreghaynes: https://review.openstack.org/#/c/83296/1/scripts/devtest_variables.sh,unified05:59
tchaypowhy is your new line #nodocs?05:59
tchaypoalso, that should be #nodocs not # nodocs06:00
greghaynesBecause non 1 values dont currently work, so figure not documenting it was the way to go06:00
greghaynesI mean, they *could* work with the proper heat template and image elements06:01
greghaynesbut we dont have those06:01
tchaypoI'd rather have an explicit comment to say that it doesn't work06:01
tchaypomaybe with a link to a blueprint/etherpad showing the work in progress, if such a thing exists #idealist06:01
* greghaynes had flashbacks as my last company was called idealist and used twitter a lot06:02
greghayneshrm, so the reason I didnt do that is that theres nothing in tripleo-incubator preventing it from working06:02
greghaynesbut docs arent a bad idea06:02
greghaynesAlso, im hoping people chime in wit "please change the notcompute name now before we really get commited to it"06:04
lifeless.oO mellanox have hardware offload for overlay networks now06:04
lifelessshiny06:04
*** killer_prince is now known as lazy_prince06:07
*** vkozhukalov has quit IRC06:10
*** jtomasek has joined #tripleo06:14
*** jtomasek has quit IRC06:14
*** rpodolyaka1 has quit IRC06:21
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Add some clarity to the first-time user experience  https://review.openstack.org/8329406:21
StevenKCONTROLLERSCALE ?06:24
greghayneshuh. It appears the cause of all my pain this evening is that my 32gb ram box has been OOMing with my changes :(06:24
lifelesswell06:24
lifelessgreghaynes: \o/06:24
lifelessso the name is bad06:24
lifelessbut06:24
tchaypoWEBSCALE06:25
lifelessspamaps has a massive overhaul of the overcloud06:25
lifelessso06:25
tchaypoor in this case NOTWEBSCALE perhaps?06:25
lifelesslets not mess with trivial06:25
lifelesstchaypo: MONGOSCALE?06:25
lifelesstchaypo: FLASHSCALE?06:25
StevenKIf the name is going to change, then let's use the new name?06:25
lifelessStevenK: We have precisely one person hacking on heat support right now; lets not cause bottlenecks more work.06:25
tchaypoclarkb: thanks for the git-review thing06:25
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Add some clarity to the first-time user experience  https://review.openstack.org/8329406:25
*** marun has joined #tripleo06:29
tchaypobtw, how many hours battery life do all y'all get with your slices?06:29
greghaynes>1206:29
StevenKI'd tell you if I had one.06:29
* StevenK sobs06:29
tchaypois your slice the bb09, or do you have something else?06:29
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Make it possible to pass real hardware in.  https://review.openstack.org/8318806:29
lifelessgreghaynes: what machine?06:29
lifelessgreghaynes: model I mean06:29
greghaynesNot sure, its the one that said my model on the smartbuy...06:30
greghayneslifeless: your old one06:30
greghaynes9470m?06:30
lifelessgreghaynes: yeah, the 9470m - was wayyyy better at batter than this06:30
lifelessI'm suspecting the video card06:30
lifelessor possibly backlight06:30
lifelessreminds me I need to try the backlight=vendor thing06:30
tchaypoI'm on the 8570w; the BB09 only gives me another ~3 hours battery life06:31
tchaypofor a total of 6-7 hours06:31
SpamapSlifeless: nvidia chipset is just not going to have the same PM capabilities.06:31
lifelessSpamapS: this has both06:31
SpamapSlifeless: if it is Haswell, it should have integrated intel graphics .. you might be able to disable the nvidia.06:31
lifelessSpamapS: I'd like to shoot the radion in the head06:31
lifelessSpamapS: exactly06:31
SpamapSlifeless: perhaps bumblebee can do it in software?06:32
lifelessbumblebee?06:32
SpamapSI never got it working on my one dual mode box (old ivybridge mac book pro)06:32
lifeless$ bumblebee06:32
lifelessbumblebee: command not found06:32
lifelessoh optimus support06:32
lifelesshmm06:32
SpamapSlifeless: https://github.com/Bumblebee-Project/bbswitch06:33
greghaynesafter some tweaking I was able to get haswell pm down to be crazy low06:33
lifelessso I have06:33
lifeless00:02.0 VGA compatible controller: Intel Corporation Haswell-ULT Integrated Graphics Controller (rev 0b)06:33
greghaynespowertop is pretty cool06:33
lifeless03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Mars [Radeon HD 8730M]06:33
lifeless^ is that optimus still ?06:33
lifelessgreghaynes: it is isn't it ;)06:33
tchaypospeaking of the 8570w - http://poseidon.epfl.ch/english/hp-offers-the-last-8570w-at-50-off.html06:33
*** ohadlevy has quit IRC06:33
lifelesstchaypo: ahahahaha so you're keeping yours ?06:34
lifelesswow these cpus have 16 different speeds06:34
lifelessthats crazy06:34
greghaynesyep, the haswell clocks down to something like a p306:35
greghaynes:p06:35
lifelessI'd be happy with burning, ok, trickle, off.06:35
lifelessthank you pulseaudio...             100.0%        PCI Device: Advanced Micro Devices, Inc. [AMD/ATI] Cape Verde/Pitcairn HDMI Audio [Radeon HD 7700/7800 Series]06:36
greghaynesI remember looking at /proc/cpuinfo once and thinking wow, I think some of my microcontrollers have a higher clockspeed06:36
*** killer_prince has joined #tripleo06:36
lifeless430 wakeups/second06:37
lifelessusermodehelper + firefox FTW06:37
tchaypolifeless: it was preferable, at the time, to continuing to not have a machine06:37
tchaypocpu is decent and it's expandable to 32Gb, so it should last me for a while06:38
lifelesstchaypo: roger wilco06:39
tchaypodownside, it's the size and weight of 3 house bricks side-by-side06:39
*** yamahata has joined #tripleo06:39
*** ohadlevy has joined #tripleo06:40
*** ohadlevy is now known as Guest6162106:40
*** marun is now known as marun_afk06:46
lifelessok, break time06:49
*** ramishra has joined #tripleo07:05
*** Guest61621 is now known as ohadlevy_07:11
*** bkero has quit IRC07:11
*** cwolferh_ has quit IRC07:15
lifelessSpamapS: can you review https://review.openstack.org/#/c/81959/ with an aim to +2 not +A ?07:16
lifelessSpamapS: so that once the dependency is through (https://review.openstack.org/#/c/81627/) I can recheck and +A it ?07:17
*** bkero has joined #tripleo07:17
*** rpodolyaka1 has joined #tripleo07:21
*** ramishra has quit IRC07:22
*** ramishra has joined #tripleo07:22
openstackgerritA change was merged to openstack/diskimage-builder: rename udev.rules to dhcp-all-interfaces-udev  https://review.openstack.org/8217307:24
*** rpodolyaka1 has quit IRC07:25
*** ramishra has quit IRC07:27
*** sirushti is now known as shortstop07:27
*** eghobo has quit IRC07:28
openstackgerritA change was merged to openstack/tripleo-image-elements: Fixup Ironic support.  https://review.openstack.org/8162707:32
*** rpodolyaka1 has joined #tripleo07:32
*** sabah has joined #tripleo07:34
*** rpodolyaka1 has quit IRC07:34
*** jprovazn has joined #tripleo07:36
*** vkozhukalov has joined #tripleo07:38
lifeless\o/07:40
*** mrunge has joined #tripleo07:41
*** mrunge is now known as mrunge_brb07:41
jprovazngreghaynes, Hi07:42
greghaynesjprovazn: Hey07:44
jprovazngreghaynes, https://review.openstack.org/#/c/79884/6/elements/mariadb/element-deps - the thing is that mariadb-dev adds the needed mariadb.repo from where packages are being installed07:45
greghaynesjprovazn: Seems odd, why doesnt the mariadb element add the repo?07:46
jprovazngreghaynes, the same repo needs to be added if you use only mariadb-dev (w/o full server stuff)07:47
lifelessmoment of truth, retry no bug on an *actual ironic run*07:47
openstackgerritMatthew Macdonald-Wallace proposed a change to openstack/tripleo-image-elements: Install the check_mk Agent  https://review.openstack.org/8148507:47
*** jistr has joined #tripleo07:47
jprovazngreghaynes, sure, some "common" element would be cleaner - but this would be dependency hell07:48
greghaynesjprovazn: ah, so the goal is to allow people to use -dev without mariadb element07:49
greghaynesmakes a lot more sense07:49
*** ramishra has joined #tripleo07:49
* greghaynes removes -107:50
greghaynesAlso +1 on not making a bunch of extra dependency complexity :)07:51
jprovazngreghaynes, I thought I put this into some prev reply earlier, but yes - mariadb-server will probably not be used for compute images but at the same time you need mariadb-dev to build python stuff (pip install)07:51
jprovazngreghaynes, cool, thanks07:51
*** rdopieralski has joined #tripleo07:53
*** rdopieralski has quit IRC07:58
*** jcoufal has joined #tripleo07:58
*** rdopieralski has joined #tripleo07:59
*** greghaynes has quit IRC08:02
*** athomas has joined #tripleo08:02
*** athomas has quit IRC08:05
*** greghaynes has joined #tripleo08:06
openstackgerritA change was merged to openstack/tuskar-ui: Fixed a few typos in runtest.sh  https://review.openstack.org/8312708:15
*** ifarkas has joined #tripleo08:15
*** bauzas has joined #tripleo08:16
openstackgerritA change was merged to openstack/tuskar-ui: Fix checkboxes on the nodes page  https://review.openstack.org/8256808:17
*** noslzzp has joined #tripleo08:24
openstackgerritRadomir Dopieralski proposed a change to openstack/tuskar-ui: Make @handle_errors support any kind of function  https://review.openstack.org/7815308:27
chmouelreading thru https://review.openstack.org/#/c/81959/20/scripts/register-nodes I was wondering if running devstack's bash8 as a gate would make sense for tripleo incubator bash scripts08:27
*** andreaf has joined #tripleo08:28
*** sabah has quit IRC08:29
lifelesschmouel: I'm not sure anything makes sense for incubator:)08:31
chmouellifeless: fair enough :) so no gates on incubator?08:32
*** matsuhashi has quit IRC08:33
openstackgerritA change was merged to openstack/tripleo-image-elements: Update elements to use os-svc-install -i.  https://review.openstack.org/7696408:34
openstackgerritA change was merged to openstack/tripleo-image-elements: Wire in OPENSTACK_VENV_DIR default if set  https://review.openstack.org/7696508:34
lifelesschmouel: its a proving ground, not a product08:34
lifelesschmouel: there are gates, but on functionality, we don't really want to overthink things there08:34
*** rpodolyaka1 has joined #tripleo08:34
lifelesschmouel: like - register-nodes is moving to os-cloud-config and probably becoming python08:35
lifelesschmouel: its not that we don't care, just about things that matter :)08:35
lifelessdang08:39
lifelesshttps://jenkins06.openstack.org/job/check-tripleo-ironic-seed-precise/88/08:39
*** rpodolyaka1 has quit IRC08:39
lifeless2014-03-27 08:15:46.676 | + user-config08:39
lifeless2014-03-27 08:15:47.153 | + echo 'Waiting for Nova Compute to be available'08:39
lifeless2014-03-27 08:15:47.153 | Waiting for Nova Compute to be available08:39
lifeless2014-03-27 08:15:47.153 | + wait_for 30 10 nova service-list --binary nova-compute '2>/dev/null' '|' grep 'enabled.*\ up\ '08:39
lifeless2014-03-27 08:21:00.903 | Timing out after 300 seconds:08:39
lifelessso nova compute didn't start08:39
lifelessand yet I just ran the same build locally08:41
lifelesshmm08:42
lifeless(successfully, end to end ironic)08:42
killer_princelifeless: regarding a1beca7b44a788dc627a210550ab2d96714f391c, on L41, there is a check for /usr/lib/grub/*-efi08:42
killer_princeI am running a ubuntu VM with EFI, and when i install grub on chrooted Fedora image, i see that08:43
*** nosnos has quit IRC08:43
killer_princethere is no *-efi at the location. and so i see that there is no target being set..08:43
*** ramishra has quit IRC08:43
killer_princecould it be the reason for 1231946 (Fedora images built on a UEFI host are configured to boot UEFI.)08:44
*** ramishra has joined #tripleo08:44
lifelesscould be08:45
rpodolyakamorning all08:45
killer_princei guess i need to update that...08:46
killer_princelifeless: on a side note, as per yesterday, were you able to re-review I8fbccc13a2c61a5191ef9ea5d2a8302a3e43b00008:47
*** mrunge_brb is now known as mrunge08:48
lifelessno, we had to fumigate the house today08:48
lifelesswill try after the call with galway08:48
*** ramishra has quit IRC08:49
killer_princeok..08:49
openstackgerritThomas Goirand proposed a change to openstack/tuskar-ui: Use "python -m coverage" rather than use "coverage" explicitly  https://review.openstack.org/8076208:53
*** hashar has joined #tripleo08:55
*** gcha has joined #tripleo08:55
*** giulivo has joined #tripleo08:58
SpamapSlifeless: headed to bed shortly.. https://review.openstack.org/#/c/81959/ <-- has real tabs :-/08:58
lifelessSpamapS: didn't I fix that?09:00
lifelessSpamapS: if thats the only thing, can you say its +2-with-that-fixed?09:00
SpamapSlifeless: Yes09:01
lifelessSpamapS: retabbed and pushing09:01
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Add Ironic support to incubator.  https://review.openstack.org/8195909:01
lsmola_tabs are fun09:01
lifeless'fun'09:01
greghaynesNow with 100% real tabs09:02
SpamapSlifeless: +2'd and now 'tis sleepy time09:02
lifelessSpamapS: nightynight09:02
lsmola_hehe09:02
lifelessgreghaynes: nice troll :)09:02
SpamapSgreghaynes: and all natural whitespace09:02
lsmola_SpamapS: good night09:02
lifelesstrying to reproduce the check failure09:02
*** yassine has joined #tripleo09:11
*** andreaf_ has joined #tripleo09:13
*** ohadlevy_ is now known as ohadlevy09:14
*** ohadlevy has quit IRC09:14
*** ohadlevy has joined #tripleo09:14
*** ruagair has quit IRC09:16
*** giulivo has quit IRC09:16
*** sabah has joined #tripleo09:20
*** derekh has joined #tripleo09:24
lifelessok this is nuts making09:31
lifelessI have the patch in question here working, its failing in CI09:32
lifelessgnarr09:32
*** e0ne has joined #tripleo09:33
openstackgerritRyan Moore proposed a change to openstack/tripleo-image-elements: Allow setting snapshots image format and directory  https://review.openstack.org/8312209:33
*** andreaf_ has quit IRC09:34
*** rpodolyaka1 has joined #tripleo09:35
openstackgerritgerry-drudy proposed a change to openstack/tripleo-image-elements: Set include_service_catalog to False  https://review.openstack.org/8332309:36
openstackgerritColeman Corrigan proposed a change to openstack/tripleo-incubator: Add set-pip-vars script to simplify pip manifests  https://review.openstack.org/8319409:36
*** tburnes has quit IRC09:36
openstackgerritJiri Stransky proposed a change to openstack/os-cloud-config: PKI key/cert generation for use with Keystone  https://review.openstack.org/8332409:37
*** tburnes has joined #tripleo09:37
*** jistr has quit IRC09:38
*** rpodolyaka1 has quit IRC09:39
*** matsuhashi has joined #tripleo09:45
*** rdopieralski has quit IRC09:48
*** rdopieralski has joined #tripleo09:50
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Add Ironic support to incubator.  https://review.openstack.org/8195909:55
*** mclaren has quit IRC09:55
lifelessok that should do it09:55
*** jtomasek has joined #tripleo09:57
*** untriaged-bot has joined #tripleo10:00
untriaged-botUntriaged bugs so far:10:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129048810:00
uvirtbotLaunchpad bug 1290488 in tripleo "Baremetal: Invalid credentials" [Undecided,Incomplete]10:00
*** untriaged-bot has quit IRC10:00
*** e0ne_ has joined #tripleo10:01
*** jp_at_hp has joined #tripleo10:02
*** jistr has joined #tripleo10:04
*** martyntaylor has joined #tripleo10:04
*** e0ne has quit IRC10:05
smulcahyif I'm using NODE_ARCH=amd64 do I need to change any of the other NODE_'s from default? or should it work out of the box10:13
*** pbelanyi has joined #tripleo10:16
lifelesssmulcahy: devtest_testenv covers that; I believe no, though we used to allow more memory for amd6410:17
ifarkasdtantsur, you can run most openstack services without keystone.10:17
ifarkasdtantsur, or you can use other auth service in theory10:18
lifelessifarkas:  most ?10:18
ifarkasdtantsur, that's why they don't register themselves with keystone automatically10:18
ifarkaslifeless, not?10:18
lifelessifarkas: AFAIK only swift, right ?10:18
lifelessifarkas: they don't register themselves automatically because they can't know their visible location10:18
ifarkaslifeless, you can set something like no_auth=true in most of them :-)10:18
dtantsurifarkas, it's not the reason to not enable this as an option. lifeless's point makes more sense, though10:19
lifelessifarkas: not automatically anyhow; and if you're going to configure it, configuring it once in keystone is easy.10:19
lifelessifarkas: vs once per api instance10:19
ifarkaslifeless, ahh, right10:19
dtantsurifarkas, for Ironic it seems like it's nove driver won't work w/o enrolling API to keystone10:20
dtantsureven if you set endpoint in config10:20
lifelessdtantsur: enrolling ?10:20
lifelessdtantsur: oh making the endpoint known in keystone?10:20
dtantsurlifeless, aka registering. I don't know the correct term10:21
lifelessregistering I think10:21
lifelessanyhow10:21
lifelessdtantsur: is that a problem for you ?10:21
dtantsurlifeless, no, we just have a general discussion :) once I wrote similar service discovery and it had automatical enrolling/registering, but the service still needed to know it's public API beforehand10:22
smulcahyNODE_CPU=${NODE_CPU:-1} NODE_MEM=${NODE_MEM:-2048} NODE_DISK=${NODE_DISK:-30} NODE_ARCH=${NODE_ARCH:-i386} #nodocs10:22
smulcahy##    For 64bit it is better to create VMs with more memory and storage because of10:22
smulcahy##    increased memory footprint::10:22
smulcahy##10:22
smulcahy##         NODE_CPU=1 NODE_MEM=2048 NODE_DISK=30 NODE_ARCH=amd6410:22
smulcahy##10:22
smulcahyis that a bug?10:22
openstackgerritDerek Higgins proposed a change to openstack/diskimage-builder: Ensure interface is only configure once  https://review.openstack.org/8333610:22
smulcahythe first line seems to be the same as the example for amd64? Or am I missing something?10:23
openstackgerritDerek Higgins proposed a change to openstack/diskimage-builder: Ensure interface is only configured once  https://review.openstack.org/8333610:23
smulcahyor should the second NODE_MEM be more than 2048?10:23
lifelesssmulcahy: no, like I said we used to have i386 leaner10:24
lifelesssmulcahy: and found it was thrashing10:24
lifelesssmulcahy: so its arguably a bug but perhaps not harmful ?10:24
lifelesswowooowooowowow10:25
lifeless2014-03-27 10:24:33.460 | Finished: SUCCESS10:25
smulcahyright, thats a paste from devtest_testenv though - so it sounds like that comment is a bug - not harmful but confusing. At first glance it seems to be saying you should bump the settings from the defaults for amd6410:25
StevenKsmulcahy: I've talked it over, and lifeless wanted hard data before I changed the line, so I did something more productive then test images.10:25
smulcahylifeless: tell me that shout of joy wasn't for a successful run of TripleO devtest ...10:25
lifelesssmulcahy: I though it was an ironic seed run completing10:26
smulcahyStevenK: which line are you considering changing? the first line or the suggested amd64 line?10:26
lifelesssmulcahy: then I realised I'd been watching the wrong jenkins console10:26
StevenKsmulcahy: The suggested amd64 line.10:26
smulcahyStevenK: ok, and what are you thinking of bumping it to?10:26
StevenKsmulcahy: 4096 RAM, everything else is probably fine.10:27
StevenKsmulcahy: Our CI cloud uses 4096 amd64 instances, as a data point.10:28
GheRiveromorning TripleO!10:28
smulcahyStevenK: kewl, thank you - will try those myself. Any benefit to bumping CPUs too if your server has plenty of cores?10:29
StevenKsmulcahy: Not certain about that, sorry.10:29
lifelessgnar dammit10:31
lifeless2014-03-27 10:28:29.602 | No valid host was found. Reason: No conductor service registered which supports driver null.10:31
lifeless2014-03-27 10:28:29.603 | + get_state_from_hosts10:31
lifelessleast its a new error10:31
*** martyntaylor has quit IRC10:31
lifelesswe need a wait state on ironic-conductor being up or something10:32
StevenKlifeless: Like the nova bm canary ?10:32
lifelessStevenK: like the 'waiting for nova compute to be ready10:32
lifelessStevenK: I may be misinterpreting the failure10:33
lifelessStevenK: all the pre-reqs for https://review.openstack.org/#/c/81959/ have landed10:33
lifelessStevenK: and it works for me10:33
lifelessStevenK: but is failing in new and interesting ways in CI10:34
StevenKHeh10:34
lifelessStevenK: more resource contention / slower CPUs - dunno the cause, but yay race conditions10:34
lifelessanyhow10:34
*** lazy_prince has quit IRC10:34
lifelessfeel free to poke (e.g. add -x to get more data from register nodes in https://jenkins03.openstack.org/job/check-tripleo-ironic-seed-precise/78/console10:34
lifelessfor now, gnight all!10:35
lifelessGheRivero: some more suggestions in the bringup perf bug10:35
StevenKlifeless: Hm, not a bad idea for your patch10:35
GheRiverolifeless: thx. taking a look now10:36
GheRiverolifeless: gnight!10:36
*** nosnos has joined #tripleo10:44
openstackgerritColeman Corrigan proposed a change to openstack/tripleo-incubator: Add set-pip-vars script to simplify pip manifests  https://review.openstack.org/8319410:50
*** martyntaylor has joined #tripleo10:51
dtantsurifarkas, not wow. This instack thing doesn't seems to build any deploy images (at least nothing in glance image-list). How is it supposed to work? Or am I missing something?10:51
openstackgerritLadislav Smola proposed a change to openstack/tuskar-ui: Do not send blank parameters  https://review.openstack.org/8334010:55
openstackgerritLadislav Smola proposed a change to openstack/tuskar-ui: Do not send blank parameters  https://review.openstack.org/8334010:57
ifarkasdtantsur, you can build your own images with diskimage-builder: http://docs.openstack.org/developer/tripleo-incubator/devtest_overcloud.html10:57
dtantsurifarkas, sure, that's what I'm doing10:57
dtantsurI'm just wondering, shouldn't it be part of instack itself (even for Nova baremetal)10:58
dtantsurslagle, ^^10:58
*** nati_ueno has quit IRC11:04
ifarkasdtantsur, I don't think the instack workflow involves any image building, but I might be wrong11:04
*** fandikurnia has quit IRC11:05
dtantsurifarkas, to deploy any baremetals you need deploy images, right?11:05
slagledtantsur: the deploy images are downloadable11:07
*** matsuhashi has quit IRC11:07
*** yamahata has quit IRC11:07
openstackgerritOm Kumar proposed a change to openstack/diskimage-builder: Fix Grub configurations for Fedora images built on a UEFI host.  https://review.openstack.org/8334211:08
dtantsurslagle, oh, I didn't know it. Where can I find them?11:08
slagledtantsur: instack-prepare-for-overcloud will download them11:08
slagledtantsur: or look in that script itself for the download links11:09
slagledtantsur: the download links are internal to RH only11:09
dtantsurslagle, I see. But will these images work with Ironic instead of Nova baremetal (when using diskimage-builder we supply it with deploy-ironic instead of deploy, afaik)11:10
slagledtantsur: Oh. No.11:10
slagledtantsur: you would have to build deploy images for ironic11:10
dtantsurslagle, ok. Thanks anyway11:10
slagledidn't realize you were trying ironic. see devtest for how to build deploy images11:11
slaglethere are some outstanding reviews that add ironic support11:11
slagleyou should be able to piece that together to see how to build the ironic deploy images11:11
dtantsurslagle, yeah, I already know how to do it. The problem is 2G of memory in instack VM, btw. But I think users are not expected to build their images, so it may be a reasonable default11:16
*** ifarkas has quit IRC11:18
*** morganfainberg is now known as morganfainberg_Z11:19
*** fandikurnia has joined #tripleo11:23
*** nati_ueno has joined #tripleo11:28
slagledtantsur: you can always give it more memory11:28
slagledtantsur: the virt setup stuff is lacking in details right now, sorry about that11:28
slagledtantsur: we decided to focus on making baremetal work well first this time :)11:29
dtantsurslagle, well, that makes sense :)11:29
*** sabah has quit IRC11:29
*** matsuhashi has joined #tripleo11:32
openstackgerritGhe Rivero proposed a change to openstack/tripleo-incubator: Use a cirros image as the end user image  https://review.openstack.org/8334711:34
*** akrivoka has joined #tripleo11:34
*** rpodolyaka1 has joined #tripleo11:35
*** hashar is now known as hasharEat11:39
*** rpodolyaka1 has quit IRC11:39
*** ifarkas has joined #tripleo11:45
*** matsuhashi has quit IRC11:45
*** lynxman has quit IRC11:48
*** lynxman has joined #tripleo11:48
openstackgerritAlexis Lee proposed a change to openstack/tripleo-image-elements: Allow configuration of extended logging  https://review.openstack.org/8335311:54
openstackgerritAlexis Lee proposed a change to openstack/tripleo-image-elements: Add logstash elements  https://review.openstack.org/8335411:56
openstackgerritDerek Higgins proposed a change to openstack/tripleo-incubator: Switch default nic driver to virtio  https://review.openstack.org/7383711:57
lsmola_derekh: hello11:58
derekhlsmola_: hi ya11:58
lsmola_derekh: seems like i have the same issue with devtest on fedora 20 http://paste.openstack.org/show/74419/11:58
*** athomas has joined #tripleo12:00
derekhlsmola_: yup, looks like it dprince mentioned last night that these should solve the issue, I'm going to try them out now https://review.openstack.org/#/q/Ia71e1d32b93db0c4c844a6dc1ebcd04ab0c13f05,n,z12:00
openstackgerritMichael Kerrin proposed a change to openstack/tripleo-image-elements: Make enabled_apis option in nova.conf configurable  https://review.openstack.org/8335512:00
lsmola_derekh: ok great, going to try now, thank you12:01
*** weshay has joined #tripleo12:04
*** matsuhashi has joined #tripleo12:05
*** sballe has quit IRC12:09
openstackgerritLadislav Smola proposed a change to openstack/tuskar-ui: Making overcloud init work  https://review.openstack.org/8334012:12
openstackgerritLadislav Smola proposed a change to openstack/tuskar-ui: Making overcloud init work  https://review.openstack.org/8334012:14
*** fandikurnia has quit IRC12:18
openstackgerritA change was merged to openstack/tripleo-incubator: Use ip route replace in devtest_seed  https://review.openstack.org/8324912:24
*** mestery has joined #tripleo12:25
*** sballe has joined #tripleo12:25
*** mestery has quit IRC12:26
*** mestery has joined #tripleo12:26
*** e0ne_ has quit IRC12:28
*** lblanchard has joined #tripleo12:29
*** jdob has joined #tripleo12:30
openstackgerritJon-Paul Sullivan proposed a change to openstack/tripleo-incubator: Add the select-cloud script  https://review.openstack.org/8280712:31
openstackgerritJon-Paul Sullivan proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325012:31
*** martyntaylor has quit IRC12:33
openstackgerritJon-Paul Sullivan proposed a change to openstack/tripleo-incubator: Add the select-cloud script  https://review.openstack.org/8280712:34
*** hasharEat is now known as hashar12:35
*** rpodolyaka has quit IRC12:36
Gonerihi guys, can a code give a +2 here: https://review.openstack.org/#/c/82315/ That's a doc only changes with mostly syntax ReST fixes.12:37
Goneris/code/core/12:37
slagleGoneri: that still needs to pass CI12:41
slagleeven though it's only docs changes, the docs are generated from the scripts12:41
openstackgerritJon-Paul Sullivan proposed a change to openstack/diskimage-builder: Write a dpkg manifest to list installed packages  https://review.openstack.org/8317112:41
slagleGoneri: anyway, i've +2'd it. and kicked off another CI run for it12:42
Gonerislagle: oh, indeed for the CI.12:43
Gonerislagle: thanks :)12:43
*** matsuhashi has quit IRC12:45
*** matsuhashi has joined #tripleo12:45
*** martyntaylor has joined #tripleo12:49
*** killer_prince has quit IRC12:49
*** matsuhashi has quit IRC12:50
*** ramishra has joined #tripleo12:58
*** e0ne has joined #tripleo12:59
*** ifarkas has quit IRC13:02
*** e0ne has quit IRC13:04
*** e0ne has joined #tripleo13:07
*** andreaf_ has joined #tripleo13:07
*** hewbrocca has joined #tripleo13:08
*** noslzzp has quit IRC13:11
*** dprince has joined #tripleo13:12
*** matty_dubs|gone is now known as matty_dubs13:15
*** matsuhashi has joined #tripleo13:18
*** julim has joined #tripleo13:18
*** markmc has joined #tripleo13:19
openstackgerritDerek Higgins proposed a change to openstack/tripleo-incubator: Install kpartx on Ubuntu  https://review.openstack.org/8337213:20
openstackgerritA change was merged to openstack/tripleo-image-elements: Adds mysql-common element  https://review.openstack.org/7988513:22
openstackgerritA change was merged to openstack/tripleo-image-elements: Adds mysql-dev and mariadb-dev elements  https://review.openstack.org/7988613:23
*** matsuhashi has quit IRC13:25
*** matsuhashi has joined #tripleo13:26
*** jistr has quit IRC13:26
*** matsuhashi has quit IRC13:26
*** matsuhashi has joined #tripleo13:26
*** spzala has joined #tripleo13:27
Ngmorning13:27
openstackgerritRyan Moore proposed a change to openstack/tripleo-image-elements: Allow setting snapshots image format and directory  https://review.openstack.org/8312213:29
openstackgerritA change was merged to openstack/diskimage-builder: Adds package mapping for mariadb packages  https://review.openstack.org/8319913:30
*** blamar has joined #tripleo13:32
*** geerdest has joined #tripleo13:33
*** yamahata has joined #tripleo13:36
*** andreaf_ has quit IRC13:39
*** noslzzp has joined #tripleo13:42
*** andreaf_ has joined #tripleo13:42
*** ramishra has quit IRC13:43
openstackgerritgerry-drudy proposed a change to openstack/tripleo-image-elements: Set include_service_catalog to False  https://review.openstack.org/8332313:43
*** jistr has joined #tripleo13:44
openstackgerritjan grant proposed a change to openstack/tripleo-image-elements: Make mysql element work on debian  https://review.openstack.org/8338113:45
*** lucasagomes has joined #tripleo13:46
openstackgerritjan grant proposed a change to openstack/tripleo-image-elements: Ensure the (block) loop device is available.  https://review.openstack.org/8338313:48
*** andreaf_ has quit IRC13:48
openstackgerritColeman Corrigan proposed a change to openstack/tripleo-incubator: Add set-pip-vars script to simplify pip manifests  https://review.openstack.org/8319413:53
*** mestery has quit IRC13:55
openstackgerritgerry-drudy proposed a change to openstack/tripleo-image-elements: Set include_service_catalog to false  https://review.openstack.org/8332314:00
*** jdob has quit IRC14:01
*** zigo has quit IRC14:01
*** jdob has joined #tripleo14:02
*** zigo has joined #tripleo14:03
*** mtaylor has joined #tripleo14:04
*** blamar has quit IRC14:06
*** blamar has joined #tripleo14:06
*** mordred has quit IRC14:08
*** mtaylor has quit IRC14:08
*** mtaylor has joined #tripleo14:08
*** mtaylor is now known as mordred14:08
*** jehb has joined #tripleo14:15
openstackgerritjan grant proposed a change to openstack/diskimage-builder: Parameterise PXE kernel and initrd selection  https://review.openstack.org/8339014:16
*** blamar has quit IRC14:17
*** blamar has joined #tripleo14:17
*** blamar has quit IRC14:21
*** blamar has joined #tripleo14:21
*** jcoufal has quit IRC14:23
*** jcoufal has joined #tripleo14:24
*** lazy_prince has joined #tripleo14:32
*** lazy_prince is now known as killer_prince14:32
*** hashar has quit IRC14:34
*** hashar has joined #tripleo14:35
*** Rakesh5 has quit IRC14:38
openstackgerritJon-Paul Sullivan proposed a change to openstack/diskimage-builder: Write a dpkg manifest to list installed packages  https://review.openstack.org/8317114:41
*** andreaf has quit IRC14:45
*** andreaf_ has joined #tripleo14:45
*** ramishra has joined #tripleo14:46
lsmola_derekh: ok, my devtest is completed14:49
lsmola_derekh: this remains to be merged https://review.openstack.org/#/c/83271/14:49
openstackgerritA change was merged to openstack/tripleo-incubator: Revert "Fedora uses em1 for undercloud/overcloud"  https://review.openstack.org/8327514:50
openstackgerritDan Prince proposed a change to openstack/diskimage-builder: dhcp-all-interfaces: correct ifquery return stmt  https://review.openstack.org/8340814:51
openstackgerritLoganathan Parthipan proposed a change to openstack/tripleo-incubator: Make libvirt volume pool configurable  https://review.openstack.org/8340914:53
derekhdprince: is that dhcp-all-interfaces fix fixing the same thing as this attempt ? https://review.openstack.org/#/c/83336/14:54
derekhlsmola_: looks like mine hit a unrelated problem, rerunning14:55
dprincederekh: yes, no offense but my fix is what we should do14:55
derekhdprince: tis fine :-)14:55
dprincederekh: Sorry, I didn't realize we were both chasing this sir14:56
derekhdprince: I hit the problem this morning when testing precise with virtio...14:56
dprincederekh: ifquery (on Ubuntu) is meant to tell us whether it exists and should happen before we reach the 'enable_interface' function14:56
dprincederekh: well, I hit this problem because I'm messing with all the things.14:57
derekhdprince: no prob, to be honest I didn't even look a t much beyond the 'enable_interface' function14:57
dprincederekh: that I broke :(14:57
openstackgerritA change was merged to openstack/tuskar: Adding swift role back  https://review.openstack.org/8249914:58
openstackgerritA change was merged to openstack/tripleo-image-elements: Allow os-svc-daemon to specify dir name with -d option  https://review.openstack.org/7803314:58
openstackgerritGhe Rivero proposed a change to openstack/tripleo-incubator: Use a cirros image as the end user image  https://review.openstack.org/8334715:00
*** mestery has joined #tripleo15:01
openstackgerritGhe Rivero proposed a change to openstack/tripleo-incubator: Allow to use a cirros image as the end user image  https://review.openstack.org/8334715:02
*** julim has quit IRC15:02
*** julim has joined #tripleo15:04
openstackgerritJames Slagle proposed a change to openstack/tripleo-heat-templates: Add BlockStorage0Config Resource  https://review.openstack.org/8341615:04
openstackgerritJames Slagle proposed a change to openstack/tripleo-heat-templates: Add enable_tunnelling to swift storage metadata  https://review.openstack.org/8341715:04
*** spzala_ has joined #tripleo15:05
*** ramishra has quit IRC15:06
marios_what's good for getting --trash-my-machine completing these days? fedora 20 expected to work?15:06
*** spzala has quit IRC15:06
*** spzala_ is now known as spzala15:06
openstackgerritRyan Moore proposed a change to openstack/tripleo-image-elements: Allow setting live migration configuration options  https://review.openstack.org/8341815:07
jdobmarios_: i tried about 2 weeks ago and was getting timeouts during... key generation?15:07
jdobseems like forever ago15:07
dprincemarios_: mostly, but grab this first https://review.openstack.org/#/c/83271/15:07
marios_jdob: dprince: thanks15:07
*** ramishra has joined #tripleo15:10
openstackgerritDerek Higgins proposed a change to openstack-infra/tripleo-ci: Remove unneeded setup from toci_gate_test.sh  https://review.openstack.org/8179615:12
*** ramishra has quit IRC15:14
openstackgerritAlexis Lee proposed a change to openstack/tripleo-heat-templates: Add logstash HOT  https://review.openstack.org/8342215:14
*** e0ne has quit IRC15:15
openstackgerritRyan Moore proposed a change to openstack/tripleo-image-elements: Allow setting live migration configuration options  https://review.openstack.org/8341815:16
*** e0ne has joined #tripleo15:18
*** matsuhashi has quit IRC15:18
*** nosnos has quit IRC15:19
openstackgerritRadomir Dopieralski proposed a change to openstack/tuskar-ui: Make @handle_errors support any kind of function  https://review.openstack.org/7815315:21
openstackgerritAlexis Lee proposed a change to openstack/tripleo-image-elements: Allow configuration of extended logging  https://review.openstack.org/8335315:22
openstackgerritLadislav Smola proposed a change to openstack/tuskar-ui: Making overcloud init work  https://review.openstack.org/8334015:22
openstackgerritA change was merged to openstack/tuskar-ui: Use "python -m coverage" rather than use "coverage" explicitly  https://review.openstack.org/8076215:23
openstackgerritAlexis Lee proposed a change to openstack/tripleo-image-elements: Add logstash elements  https://review.openstack.org/8335415:25
*** spzala_ has joined #tripleo15:25
*** matty_dubs is now known as matty_dubs|czech15:26
*** spzala has quit IRC15:28
*** spzala_ is now known as spzala15:28
*** mrunge has quit IRC15:33
*** hashar has quit IRC15:36
openstackgerritA change was merged to openstack/os-cloud-config: Initial Keystone users, roles, domains, projects  https://review.openstack.org/7814815:40
*** hashar has joined #tripleo15:44
*** bcrochet has quit IRC15:45
*** killer_prince is now known as lazy_prince15:45
*** bcrochet has joined #tripleo15:47
*** ewindisch has quit IRC15:47
*** ewindisch has joined #tripleo15:50
*** lxsli has joined #tripleo15:50
openstackgerritRyan Moore proposed a change to openstack/tripleo-image-elements: Allow setting of compute configuration options  https://review.openstack.org/8343815:50
*** marun_afk has quit IRC15:53
*** rdopieralski has quit IRC15:58
*** untriaged-bot has joined #tripleo16:00
untriaged-botUntriaged bugs so far:16:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129846616:00
uvirtbotLaunchpad bug 1298466 in tripleo "devtest cleanup-env doesn't remove seed image" [Undecided,New]16:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129846516:00
uvirtbotLaunchpad bug 1298465 in tripleo "Baremetal recreation may fool the heat stack." [Undecided,Confirmed]16:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129048816:00
uvirtbotLaunchpad bug 1290488 in tripleo "Baremetal: Invalid credentials" [Undecided,Incomplete]16:00
*** untriaged-bot has quit IRC16:00
*** andreaf has joined #tripleo16:01
*** eghobo has joined #tripleo16:03
*** andreaf has quit IRC16:04
*** andreaf_ has quit IRC16:04
*** andreaf has joined #tripleo16:04
*** epim has joined #tripleo16:09
*** noslzzp has quit IRC16:12
devanandahi! anyone know if/howbadly http://git.openstack.org/cgit/openstack/tripleo-incubator/tree/images is out of date?16:13
openstackgerritAna Krivokapic proposed a change to openstack/tuskar-ui: Fix exception logging  https://review.openstack.org/8344816:14
openstackgerritA change was merged to openstack/diskimage-builder: Make stable-interface-names its own element.  https://review.openstack.org/8327116:14
derekhdprince: thanks ^16:14
devanandahmm. nvm. looks like they haven't changed since my last update a year ago16:14
*** epim_ has joined #tripleo16:14
dprincederekh: cool, hopefully that puts things back like they were... working again16:15
*** cwolferh_ has joined #tripleo16:15
derekhdprince: tried it with and without virtio, in both cases I have eth0, overcloud failed to fully come up, but It was unrelated16:16
*** epim has quit IRC16:16
*** epim_ is now known as epim16:16
dprincederekh: Yeah. Lets just not use it for vms. Too much trouble.16:16
devanandajcoufal: hi! I've been meaning to chat with ya'll about tuskar/ironic interaction16:17
devanandajcoufal: saw your email just now :) have a minute to get me up to speed?16:17
dprincederekh: I may propose that we stop using the baremetal element in devtest... because well, it should be for 'baremetal' things.16:18
*** gcha has quit IRC16:18
dprincederekh: we'll need to swap the kernel, ramdisk extraction into another element first though16:18
devanandadprince: what element configures nova to use the baremetal/ironic driver?16:19
dprincedevananda: ironic I think16:19
dprincedevananda: no, the Nova element can do it. We added a compute_driver setting recently.16:19
dprincedevananda: https://review.openstack.org/#/c/82637/16:20
dprincedevananda: nah, wrong link16:20
jcoufaldevananda: sure thing16:20
jcoufaldevananda: do you have any specific questions? :)16:21
dprincedevananda: https://review.openstack.org/#/c/82677/16:21
devanandadprince: it looks like nova-baremetal element is creating the necessary system req's (eg, install tftp and create /tftpboot dir, etc)16:22
devanandajcoufal: maybe .. let's start with, how do you see tuskar and ironic interacting?16:23
dprincedevananda: SO the 'baremetal' element above is something could become that thing, but I don't think that is what it is today.16:23
devanandadprince: what is different about a deployment /actually/ on baremetal vs one that uses nova-bm // ironic, but with virtual machines?16:24
*** nati_ueno has quit IRC16:24
derekhdprince: ya sounds like the baremetal elements name could jut be renamed or something but I'm not sure we should go too far down the road of having a different set of elements for devtest when compared to baremetal16:25
devanandadprince: AFAICT, it's just runtime configuration. For nova-bm, it's a CONF setting. For ironic, it's just a change via the API to node.driver16:25
dprincedevananda: in my experience lots of things. Vms aren't good at many aspects. Recently, the network interface naming (via biosdevname) was what was causing pain.16:25
dprincedevananda: and I've also, hit several power management issues as well which the virtual driver doesn't allow us to test16:26
dprincedevananda: and race conditions in the ramdisk/deployement too16:26
devanandadprince: ahh, gotcha. so an element to separate out the build differences for physical hardware16:26
dprincedevananda: that what is was meant to be yes, but IMO it can't be used for that yet because it is actually required by devtest...16:27
devanandadprince: i would call that element just "baremetal" then16:27
devanandanot "nova-baremetal"16:27
dprincedevananda: that is what it is called :)16:27
devanandai don't see a "baremetal" element???16:27
dprincedevananda: it lives in DIB16:28
devanandaoooh16:28
devanandadprince: nvm! :)16:28
devanandajcoufal: so, does tuskar have (or plan to have) a UI tab for CRUD of ironic nodes?16:29
*** athomas has quit IRC16:30
jcoufaldevananda: sorry, I got a bit distracted16:30
jcoufaldevananda: so first of all let me split Tuskar and the Management UI16:31
jcoufaldevananda: Tuskar itself is separate service, which is modelling heat templates and helping to deploy/manage the deployment16:32
jcoufaldevananda: it has nothing to do with nodes16:32
jcoufaldevananda: the whole openstack management story is represented by UI16:33
jcoufaldevananda: where the UI should consume ironic and manage the whole CRUD of ironic nodes16:33
jcoufaldevananda: so the answer is yes - we would love to expose in the UI a tab for node management16:33
jcoufaldevananda: we actually do have that but with nova-baremental16:34
jcoufal*baremetal16:34
*** epim has quit IRC16:34
devanandajcoufal: gotcha. so. is that UI tab for node management then a horizon plugin or something else?16:34
jcoufaldevananda: currently the management UI is Horizon plugin, but it disables project and admin tabs. Leaves just 'infrastructure' - which is whole about undercloud service including ironic16:35
devanandajcoufal: hmm, I may need to understand taht better at some point. there are use-cases for Ironic outside of TripleO16:36
devanandajcoufal: so a UI plugin for Ironic needs to take those into account as well16:36
jcoufaldevananda: so if we discuss the use cases, we can figure out where the Ironic tab belongs the best16:37
jcoufalnow we have whole separate dashboard called infrastructure, where is everything related with the underlying infrastructure16:37
devanandajcoufal: sure. also a bit of context, I think for graduation, we will need to have a horizon plugin specifically for ironic16:38
openstackgerritAndrea Rosa proposed a change to openstack/tripleo-heat-templates: Define nova enabled_apis as Heat-configurable  https://review.openstack.org/8348016:38
*** bauzas has quit IRC16:38
jcoufaldevananda: what functionality should it have?16:38
devanandajcoufal: for example, dedicated hosting service providers (eg, Rackspace) may use Ironic themselves to manage hardware nodes, and expose access to their users only through nova16:39
openstackgerritJenkins proposed a change to openstack/os-collect-config: Updated from global requirements  https://review.openstack.org/8348316:39
openstackgerritAndrea Rosa proposed a change to openstack/tripleo-heat-templates: Define nova enabled_apis as Heat-configurable  https://review.openstack.org/8348016:39
devanandajcoufal: so the operators / admins would need the ability to add/update/monitor hardware state and perform certain tasks directly, which would not be exposed to non-privileged users16:40
*** marun has joined #tripleo16:40
devanandajcoufal: eg, set the "maintenance" mode flag on a node, toggle serial console, update MAC address, etc16:40
*** zigo has quit IRC16:41
jcoufaldevananda: great. I wouldn't like to duplicate this functionality on two places. Infrastructure tab is for managing your underlying infrastructure. If we have node management there, it can be used for your use-cases as well.16:42
devanandajcoufal: ++16:42
jcoufaldevananda: http://people.redhat.com/~jcoufal/openstack/tripleo/2014-02-05_tripleo-ui-icehouse.pdf - see the first few slides - there is node management which we have atm16:42
devanandajcoufal: this is all functionality that tripleo infrastructure operators will want to use, too :)16:42
devanandajcoufal: UI looks good, should fit on top of Ironic easily16:44
jcoufaldevananda: we should add node detail views as well, they are missing in wireframes, but partly are already implemented16:44
devanandayep16:44
jcoufaldevananda: yeah, I designed the UI so it fits for Ironic, so it should be easy to switch from nova-baremetal to ironic one it is ready to use16:45
devanandajcoufal: it's ready IMO :)16:45
jcoufaldevananda: as far as I remember there were some issues :-P but maybe they are fixed already, I will check with the guys16:46
jcoufaldevananda: would be good if we can discuss what functionality you need to expose in the UI16:46
devanandajcoufal: i'm assuming it's possible to use this UI w/o Tuskar service -- how would one do that?16:46
jcoufaldevananda: What you need to do is to enable plugin in the Horizon - then you have it available for you16:48
devanandaahh, great!16:48
devanandai must admit, I haven't used Horizon in probably more than a year ...16:49
*** e0ne has quit IRC16:49
jcoufaldevananda: how big priority is the UI for Ironic?16:50
jcoufalis it graduation requirement?16:50
jcoufalyou mentioned something in this sense...16:50
devanandayep16:51
devanandajcoufal: so ironic is not graduating in icehouse, mostly due to the need to integrate with everything before graduation16:52
jcoufaland the graduation is expected sometimes early Juno, correct?16:52
devanandajcoufal: graduation only happens at the end of a cycle. But. the TC watches the project's progress for the whoel cycle16:52
devanandajcoufal: so, once Juno opens, the sooner that ironic has a horizon plugin, the better16:53
jcoufaldevananda: so what I think what we can do is to setup etherpad and write down requirements for the UI with some user-stories, so we have better understanding of what is missing and how to design that16:54
*** pbelanyi has quit IRC16:55
jcoufaldevananda: I wanted to suggest that at the Summit we can discuss more details, but I guess that it would be good to start sooner16:55
*** pbelanyi has joined #tripleo16:56
SpamapS2014-03-27 10:28:28.877 | Deleted chassis 79c47733-d752-457e-bb65-e3c4be968f7916:59
SpamapS2014-03-27 10:28:29.602 | No valid host was found. Reason: No conductor service registered which supports driver null.16:59
devanandajcoufal: summit ++16:59
SpamapSwhile testing https://review.openstack.org/#/c/81959/2217:00
SpamapSdevananda: ^17:00
jcoufaldevananda: would you mind to start the etherpad with basic requirements on top of which we can build?17:01
SpamapSDriverNotFound: FakeSeaMicroDriver17:01
devanandaSpamapS: yep. you should be able to ignore that error17:02
devanandaSpamapS: humm. why is it exiting?17:04
devanandathat is not a critical error17:04
*** jehb has quit IRC17:04
devanandaSpamapS: eg, in devstack-gate: http://logs.openstack.org/93/82993/2/check/check-tempest-dsvm-virtual-ironic/fd1207b/logs/screen-ir-cond.txt.gz#_2014-03-27_06_56_56_13317:05
*** matty_dubs|czech is now known as matty_dubs|lunch17:07
openstackgerritJonathan Brownell proposed a change to openstack/diskimage-builder: Fix "(None)" seed hostname on Debian  https://review.openstack.org/7730117:08
openstackgerritA change was merged to openstack/tripleo-incubator: Use if; then style in write-tripleorc  https://review.openstack.org/8328917:10
bnemecslagle: I wrote up my results with pypi-mirror: http://blog.nemebean.com/content/using-pypi-mirror-devtest17:12
openstackgerritGonĂ©ri Le Bouder proposed a change to openstack/diskimage-builder: fix grub2 installation on Debian Wheezy  https://review.openstack.org/8350617:13
openstackgerritColeman Corrigan proposed a change to openstack/tripleo-incubator: Add set-pip-vars script to simplify pip manifests  https://review.openstack.org/8319417:14
*** marun has quit IRC17:15
*** nati_ueno has joined #tripleo17:15
openstackgerritA change was merged to openstack/tuskar: Return template attributes  https://review.openstack.org/7755517:16
*** UtahDave has joined #tripleo17:16
*** vkozhukalov has quit IRC17:16
*** UtahDave has left #tripleo17:17
*** jehb has joined #tripleo17:17
*** e0ne has joined #tripleo17:18
slaglebnemec: cool!17:19
dprinceSpamapS: So I'm trying to get https://review.openstack.org/#/c/69918/ working on Ubuntu. The reason it appears to fail is that the OVS bridge isn't assuming the MAC address of the first device that gets added to it.17:25
dprinceSpamapS: The weird thing is I tested the same thing on baremetal and it worked there...17:26
dprinceSpamapS: any ideas?17:26
SpamapSdprince: o_O .. that is pretty important behavior if I have any understanding of how this is supposed to work. :-P17:26
*** morganfainberg_Z is now known as morganfainberg17:26
SpamapSdprince: as a "hey let's just try this" .. perhaps try trusty?17:27
dprinceSpamapS: It is. I'm surprised and perplexed. On Fedora this just works... and great.17:27
SpamapShttps://launchpad.net/ubuntu/+source/openvswitch <--17:27
SpamapStrusty == ovs 2.017:27
dprinceSpamapS: sigh. This is a super wheel spin man17:28
SpamapSdprince: we'll need to be on trusty ASAP anyway17:28
SpamapSdprince: saucy is dead in ~3 months from now17:28
dprinceSpamapS: what do I do to switch to this 'trusty', set DIB release?17:29
SpamapSDIB_RELEASE=trusty yeah17:30
dprinceSpamapS: and you actually use this already?17:30
* dprince wants not another wild goose chase17:30
SpamapSdprince: have seen one build of seed work.17:30
SpamapSdprince: did not push beyond there.. was trying to do something similar (see if something was fixed.. and, it was!)17:31
SpamapSdprince: it's not a wild goose chase, it is bisecting17:31
SpamapSdprince: whats the version of ovs in fedora userland and kernel?17:32
dprinceSpamapS: 2.0.1 I think17:32
SpamapSso further evidence... :)17:32
*** e0ne has quit IRC17:32
*** e0ne has joined #tripleo17:33
SpamapSdprince: if it is in fact fixed in trusty, then we are close enough to trusty's release, I'd say you just say "fixed in trusty" and wait for us to bump the default about a week after trusty releases and CI passes with that. :)17:33
SpamapShttps://wiki.ubuntu.com/TrustyTahr/ReleaseSchedule17:33
SpamapSApril 1717:33
dprinceSpamapS: well, the thing that really kicks me is that this all worked on real machines. Its Ubuntu VMs (CI or whatever) that give me trouble with the MAC not being assigned it seems17:34
SpamapSdprince: hm ... wait now.. hmm17:34
SpamapSdprince: these are the bridges on the "host" ?17:34
SpamapSas in, trying to bridge brbm's together?17:34
dprinceSpamapS: no. In the VM17:34
SpamapSThat's truly weird then.17:35
SpamapSSame kernel I presume?17:35
dprinceSpamapS: yeah, I even tried switching to virtio to (which is hugely faster BTW)17:35
SpamapSyeah I can't remember why we did e100017:35
SpamapSthere was some horrible crash I think back in the 12.10 days17:35
*** martyntaylor has quit IRC17:36
*** yassine has quit IRC17:36
SpamapSdprince: seems like it wouldn't be affected by being in a VM versus on hardware. That is a purely software thing to assimilate the MAC17:37
SpamapSdprince: one thing .. maybe it has to do with whether the interface is fully up or not?17:38
SpamapSor udev timing..17:38
SpamapSbecause on hardware link often takes a lot longer...17:38
SpamapSmight be that net-device-up and thus ovs takeover happened way later and something else had already happened17:38
* SpamapS always blames a race17:38
*** zigo has joined #tripleo17:39
SpamapSdprince: ok, I'm mid-bring-up of an overcloud.. once it finishes, I'll try it out and see if I can debug this with my more-interested-in-ubuntu-than-dprince hat on ;)17:39
dprinceSpamapS: you should also land this fix soon too https://review.openstack.org/#/c/83408/17:42
dprinceSpamapS: cause you're gonna hit that first. Plus it'll make you hate me less17:43
SpamapSOh I doubt that17:45
SpamapS;)17:45
SpamapSdprince: i've seen that btw.. tens of interfaces in /etc/network/interfaces ... good to know I wasn't crazy17:45
* SpamapS notes that this script does not use -e .. and as usual... mistakes were made.. errors were missed ;)17:46
dprinceSpamapS: yeah, well when I saw it I thought the world was ending. Then I did git blame and was like, aw... man. I did this?17:46
*** e0ne has quit IRC17:47
SpamapSdprince: I love it when I find my own errors. I figure that way I can avoid the judgemental patches from others and judge myself instead :)17:47
*** hashar has quit IRC17:48
SpamapSGheRivero, slagle , lifeless , Ng https://review.openstack.org/#/c/83408/ needs a +A17:49
* bnemec notes that soon there will be a way to catch such mistakes: https://review.openstack.org/#/c/81637/ ;-)17:49
openstackgerritJay Dobies proposed a change to openstack/python-tuskarclient: Changed role count assignment to use role name  https://review.openstack.org/8005217:53
SpamapSbnemec: love it! http://logs.openstack.org/37/81637/6/check/gate-diskimage-builder-pep8/9223907/console.html is full of TODO's now ;)17:53
*** newell has joined #tripleo17:55
bnemecSpamapS: Exactly.  I've been holding off on fixing those until I had a decision on whether to also enforce -u or -x, but I'm thinking -u may break a bunch of scripts and -x, while useful, probably isn't something to enforce universally.17:57
bnemec-e may break some things too, but only if they were already broken. :-)17:57
*** e0ne has joined #tripleo18:00
lifelessmorningish18:00
*** e0ne has quit IRC18:01
*** lucasagomes has quit IRC18:01
openstackgerritA change was merged to openstack/diskimage-builder: dhcp-all-interfaces: correct ifquery return stmt  https://review.openstack.org/8340818:01
*** matty_dubs|lunch is now known as matty_dubs18:03
derekhlifeless: morning, can we unblock the virtio netwirk drive please ? https://review.openstack.org/#/c/73837/18:04
derekhlifeless: ran a bunch of tests details on commit message18:04
*** bauzas has joined #tripleo18:04
lifelessGheRivero: slagle SpamapS https://review.openstack.org/#/c/73837/18:07
lifelessderekh: awesome thank you!18:07
*** jistr has quit IRC18:07
lifelessa for ironic18:08
lifeless        PM_TYPE=$(jq -r ".[\"pm_type\"]" <<< $NODE_JSON)18:08
lifelesswe need a new testenv deploymen18:08
lifelessspecifically using devtest_testenv18:09
derekhlifeless: no prob, thanks18:09
SpamapSlifeless: is that regarding the fail in https://review.openstack.org/#/c/81959/2218:09
lifelessderekh: hey - so to land https://review.openstack.org/#/c/81959/18:09
SpamapSNobodyCam: ^^18:09
lifelessSpamapS: yes18:09
SpamapSok will stop investigating18:09
lifelessthis line18:09
lifeless        NODE_ID=$(ironic node-create -d $PM_TYPE | awk '$2=="uuid" { print $4 }')18:09
lifelessis the one reporting a driver value of null18:09
lifelessbecause its looking up a missing key18:10
lifelessthat gets set by  the thing creating the testenv18:10
lifelessand we've added a bunch of keys as part of StevenK's refactoring18:10
lifelessI mean, let me double check18:10
lifelessit may be I had a thinko and added it in the same patch18:10
*** jehb has quit IRC18:12
lifelessah yeah I did18:12
lifelessso let me split this real fast18:12
lifelessderekh: I have an urgent patch for the testenvs18:12
lifelessbut it will need https://review.openstack.org/#/c/83285/18:12
lifelessah, I see thats blocked, ok, will tackle toci directly18:13
lifelessand i"m wrong18:14
lifelesscreate-nodes is all I need18:14
NobodyCamlifeless: just fyi I do have pm_type in my nodes def from the latest run18:14
lifelesswill split out in a sec18:14
lifelessNobodyCam: yes, but not in CI18:14
NobodyCamahh18:14
NobodyCam:)18:14
lifelessNobodyCam: see tripleo-ci/elements/testenv-worker/bin/ensure-test-env18:14
SpamapSNo distributions at all found for iso8601>=0.1.9 (from -r requirements.txt (line 2))18:16
SpamapSargh18:16
derekhlifeless: so if there anything I need todo?18:17
*** blamar has quit IRC18:17
derekhlifeless: actually I gotta run a catch a train but will log on in about an hour to help review stuff through18:18
*** blamar has joined #tripleo18:18
*** derekh has quit IRC18:19
*** jehb has joined #tripleo18:25
openstackgerritA change was merged to openstack/tripleo-incubator: Switch default nic driver to virtio  https://review.openstack.org/7383718:27
openstackgerritNicholas Randon proposed a change to openstack/tripleo-incubator: Extend devtest to allow real hardware deploys.  https://review.openstack.org/8353118:29
*** panda has quit IRC18:30
*** fandi has joined #tripleo18:31
*** panda has joined #tripleo18:32
tteggelany core that can look at https://review.openstack.org/#/c/82074/ needs a +A :)18:32
lifelessbah missed derekh :)18:36
tteggelSpamapS: thanks18:37
SpamapStteggel: np18:37
openstackgerritA change was merged to openstack/diskimage-builder: Uses policy-rc.d to prevent dpkg starting daemons  https://review.openstack.org/8207418:38
*** rpodolyaka has joined #tripleo18:39
*** marun has joined #tripleo18:40
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Add pm_type to the nodes list.  https://review.openstack.org/8353218:40
openstackgerritlifeless proposed a change to openstack/tripleo-incubator: Add Ironic support to incubator.  https://review.openstack.org/8195918:40
lifelessSpamapS: slagle: https://review.openstack.org/83532 - doesn't run under CI , but we need that deployed on the CI machines to land https://review.openstack.org/8195918:41
*** akrivoka has quit IRC18:41
*** jprovazn has quit IRC18:42
*** epim has joined #tripleo18:43
*** epim has quit IRC18:44
*** epim has joined #tripleo18:44
NobodyCamNo distributions at all found for oslo.config>=1.2.0 grrr18:45
*** john-n-seattle has left #tripleo18:46
*** spzala has quit IRC18:47
Lotus907efilifeless: if I write this new element to diskimage-builder to copy the GPG key in for a private repo would this be of interest to others? I.E. should I write it with the intention of submitting it upstream?18:48
lifelessLotus907efi: yes, in fact I was suggesting you can just add it as a feature to the existing dpkg element18:48
SpamapSStevenK: https://review.openstack.org/#/c/83238/1 <-- your -1 is erroneous.18:48
lifelessLotus907efi: but you can do it separately if you'd like18:48
Lotus907efithat would be preferable to writing a new element?18:49
lifelessLotus907efi: 6/1 1/2 dozen of the other18:49
lifelessNobodyCam: from CI or locally?18:49
NobodyCamthat was local18:49
lifelessNobodyCam: do you have a pypi-mirror ?18:50
lifelessSpamapS: https://review.openstack.org/#/c/82741/ looks ready to land18:50
NobodyCamlifeless: not atm.. I can start on on the brick.18:50
lifelessNobodyCam: https://review.openstack.org/#/c/82741/ has a bunch of making things fast/reliable advice18:50
*** rpodolyaka has joined #tripleo18:52
* NobodyCam reads :) TY lifeless 18:52
dprinceSpamapS: just tried it again on a real machine and my patch (as-is) works great.18:53
dprinceSpamapS: So the issue is that in on Ubuntu, when using a VM the OVS bridge doesn't assume the MAC of the first NIC for dhcp18:54
dprinceSpamapS: I gotta say, I'm now inclined to split this into a two patch series because we really do want this in. If fixes many of the underpinnings with our underlying bridge config (namely its DHCP config)18:56
SpamapSlifeless: https://review.openstack.org/#/c/82741/ caught 1 error, and have a few optional suggestions18:57
SpamapSdprince: Give me 2 hours. I'm just now ready to spin up that patch and test it on Ubuntu vms here locally.18:57
SpamapSdprince: I'm hoping there is something non-obvious that we can tweak to get it working.18:57
SpamapSlifeless: ^^ note dprince's struggles with ovs bridges not assuming the right MAC18:58
lifelessSpamapS: dprince: which patch?18:58
lifelessis it the one that was using the e/n/i integration for ovs which is heinous ?18:58
dprinceSpamapS: cool. Just wanted you to have the extra assurance that I've done this multiple times now and it works... (just not in Ubuntu VMs yet :(18:58
lifeless(not the patch, the integration)18:59
SpamapSdprince: thanks. I am bolstered by your success. :)18:59
lifelessShrews: btw try tuning all the tunables in powertop, that made quite a diff for me19:01
*** markmc has quit IRC19:04
*** jp_at_hp has quit IRC19:04
Shrewslifeless: hrm, installing...19:06
*** epim has quit IRC19:06
*** hewbrocca has quit IRC19:07
*** bauzas has quit IRC19:07
*** jdob has quit IRC19:09
*** spzala has joined #tripleo19:09
*** jdob has joined #tripleo19:09
*** bauzas has joined #tripleo19:11
lifelessmarios_: Ng: GheRivero: https://review.openstack.org/#/c/83532/ <- critical path for Ironic (after that lands I need to rebuild the testenvs)19:11
* Ng looks19:12
lifelessNg: note that it doesn't get CI tested, because too-hard-so-far so it can't affect job pass/fail19:12
*** spzala has quit IRC19:12
openstackgerritA change was merged to openstack/tripleo-incubator: Add pm_type to the nodes list.  https://review.openstack.org/8353219:15
lifeless\o/19:17
* lifeless goes to redeploy the CI environments19:17
*** lifeless changes topic to "CI down (testenv redeploy) | tripleo-cd running preserve-ephemeral WIP patches and https://review.openstack.org/#/c/62042/ | Using OpenStack to deploy OpenStack;meetings Tuesday 1900 UTC in #openstack-meeting-alt"19:18
*** rpodolyaka has quit IRC19:18
*** rdopieralski has joined #tripleo19:18
* SpamapS lunches while new overcloud w/ Dan's changes tries to deploy.19:20
* lifeless stabs the undercloud network, sheese.19:22
*** rdopieralski has quit IRC19:23
*** GheRivero has quit IRC19:28
dprinceSpamapS: So this isn't an Ubuntu thing. It is a VM thing. I just tested on a Fedora undercloud VM and the same thing happens there too19:29
dprinceSpamapS: So are real baremetal all is well. But when using devtest with this patch OVS bridges aren't assuming the correct MAC addresses (of the first NIC)19:30
*** vkozhukalov has joined #tripleo19:31
SpamapSdprince: this goes back to my wondering about races19:31
SpamapSdprince: because vms have instant link19:31
dprinceSpamapS: could very well be something in that realm, anyways proceed but know that it is a generic VM issue it seems19:33
SpamapSdprince: that's quite helpful.19:36
SpamapSdprince: as then you can tinker in parallel without casting doubting glances towards "those Ubuntu guys" ;)19:36
openstackgerritClint "SpamapS" Byrum proposed a change to openstack/tripleo-heat-templates: WIP: Switch overcloud to software-config  https://review.openstack.org/8166619:37
lifelessSpamapS: dprince: what was the patch # ?19:38
openstackgerritJames Slagle proposed a change to openstack/tripleo-image-elements: Install tgt package for cinder package install  https://review.openstack.org/8354419:38
SpamapSlifeless: https://review.openstack.org/#/c/69918/19:38
SpamapS"that one" ;)19:38
*** jcoufal has quit IRC19:38
*** ohadlevy has quit IRC19:39
*** jcoufal has joined #tripleo19:39
devanandalifeless: i have not gotten around to the scheduler yet19:41
devanandalifeless: been fighting with stevedore and translations not working19:41
*** rpodolyaka has joined #tripleo19:41
NobodyCamhumm undercloud booted this time. but ran right into Falied to resize filesystem.19:44
*** epim has joined #tripleo19:44
lifelessdprince: which reminds me, did you mananage to pull down that reproducer image?19:45
dprincelifeless: not yet19:46
lifelesskk19:46
*** GheRivero has joined #tripleo19:46
*** lblanchard has quit IRC20:00
*** fandi has quit IRC20:01
*** rpodolyaka has quit IRC20:03
*** akrivoka has joined #tripleo20:11
Lotus907efiquestion: the diskimage-builder/README.md file says the extra-data.d inputs consists of $TMP_HOOKS_PATH20:16
Lotus907efibut I am looking at elements/apt-conf/extra-data.d/99-override-default-apt-conf and it uses $TMP_MOUNT_PATH20:16
Lotus907efiIs that a valid path available for other elements?20:16
Lotus907efialso if I am building an ubuntu image is the dpkg element included by default or do I have to specifically include it in DIB_COMMON_ELEMENTS ?20:17
lifelessdpkg is included as a dependency of ubuntu20:17
Lotus907efiok20:17
Lotus907efican I use $TMP_MOUNT_PATH in my extra-data.d script in the dpkg element?20:18
NgLotus907efi: that probably will work. Originally we were trying to use $TMP_HOOKS_PATH in the extra-data.d scripts, and then have an install.d script put things in the right places20:21
lifelessLotus907efi: TMP_HOOKS_PATH is where you can copy the key to.20:21
Lotus907efiI do not see $TMP_MOUNT_PATH mentioned in either diskimage-builder/README.md or in elements/apt-conf/README.md20:22
Lotus907efiok20:22
lifelessLotus907efi: you'll need to use an install.d script to load the key.20:22
Lotus907efiI thought I needed to use a pre-install.d script?20:22
Lotus907efi(which I have written)20:23
lifelesspre-install.d yes20:23
Lotus907efibut I can move it to elements/dpkg/install.d if that is more appropriate20:23
lifelesssorry :)20:23
Lotus907efiok, cool20:23
lifelessno, you'll want it to happen before the first apt-get update20:23
Lotus907efiright20:23
Lotus907efiso should I submit a bug or a blueprint for this?20:25
lifelessjust the code patch should be fine20:26
*** julim has quit IRC20:29
*** rpodolyaka has joined #tripleo20:30
Lotus907efiah ha! it worked (sort of)20:48
Lotus907efijeb's repo seems to have some broken dependencies20:48
Lotus907efibut it did seem to import the GPG key inthe the apt keyring ok20:49
Lotus907efibut it did seem to import the GPG key into the the apt keyring ok20:49
Lotus907efino GPG errors this time in my log file20:49
*** dprince has quit IRC20:49
*** jdob has quit IRC20:55
*** cwolferh_ has quit IRC20:58
*** cwolferh_ has joined #tripleo20:58
*** vkozhukalov has quit IRC20:59
lifelessLotus907efi: great21:01
*** jehb has quit IRC21:02
*** bauzas has quit IRC21:03
*** derekh has joined #tripleo21:03
derekhlifeless: I see your redeploying testenv's, probably too late for this but have you seen the update to the deploy script? https://review.openstack.org/#/c/81686/ , I'll update the other Rack later and try it out21:10
lifelessderekh: hi21:10
lifelessderekh: I'm trying but having it fail :(21:10
lifelessderekh: and no ports showing up in overcloud neutron21:11
lifelessderekh: looks nice21:12
derekhlifeless: usually means the 60-setup-testenvs script failed, taking a look, still on train so may loose connecting21:12
lifelessderekh: nova was failing at deploying some of the machines too21:15
lifelessderekh: I'm considering a mellanox driver reload again21:15
derekhlifeless: k21:16
derekhhmm, can't ssh in, might just need to wait a bit longer21:19
lifelessto ?21:19
lifelessenv0 ?21:20
lifelessahha 148 is up21:20
derekhlifeless: env0 5 or 621:20
derekhlifeless: first problem device names have changed again s/em3/eth2/21:22
lifelessSTABBBBBBBBB21:22
lifelessdeleting21:23
lifelessderekh: its odd that that doesn't trigger a failure21:23
derekhlifeless: a failure of the stack create?21:24
lifelessof os-collect-config21:24
lifeless--------------------- END PROFILING ---------------------21:24
lifeless[2014-03-27 21:27:30,071] (os-refresh-config) [INFO] Completed phase migration21:24
lifelessit got all the way through21:24
lifelessno error21:24
lifelessif it errored heat would know via the waitcondition21:24
derekhlifeless: well, I didn't look at the logs but I'm pretty sure I know what happens21:24
derekhlifeless: it failed on the first pass21:25
derekhlifeless: and succedded on the second21:25
derekhlifeless: because of ...21:25
lifelessderekh: oh, because of your don't-do-this-twice thing (rather than being idempotent)21:25
derekhlifeless: yup, thats it, I've started a local branch to fix it, but got caught up on createing a dummy nic for the overcloud IP21:26
derekhlifeless: will see if I can get it sorted tomorrow21:26
lifelessderekh: ok21:26
*** jcoufal has quit IRC21:27
lifelessrighto time for a hot drink21:27
*** derekh has quit IRC21:30
*** rpodolyaka has quit IRC21:39
adam_ghttp://paste.openstack.org/show/74500/21:42
adam_g^ i can't seem to build a seed with a functioning python environment. anything obvious i may have borked?21:42
SpamapSadam_g: need more context21:43
SpamapSadam_g: that's happening inside the seed?21:43
adam_gSpamapS, yes21:44
SpamapSadam_g: weirdness21:44
SpamapSadam_g: I just built a seed 3 hours ago.. working fine21:44
adam_gSpamapS, i noticed config wasns't being applied, OVS bridge wasns't coming up and then relized my entire python environment is fubard21:44
SpamapSadam_g: building on saucy?21:45
adam_gSpamapS, building saucy i386 images on a trusty amd6421:45
SpamapSadam_g: hm, I haven't tried that combo yet21:45
adam_gis trusty a known working target for images?21:45
SpamapSno21:46
SpamapSbut it needs to be21:46
SpamapSI want to switch as soon after release as possible.21:46
openstackgerritBen Nemec proposed a change to openstack/tripleo-image-elements: Add cinder-lio element  https://review.openstack.org/7846321:46
lifelessok we have testenvs coming up21:46
SpamapSadam_g: it is possible we have new broken cloud images?21:47
SpamapSadam_g: hm, no mine are from yesterday21:47
adam_gi'm using a cached set of images from a couple weeks ago, let me see if pulling a newer one works21:47
SpamapSadam_g: do you have /usr/lib/python2.7/plat-i386-linux-gnu/_sysconfigdata_nd.py ?21:49
openstackgerritSteven DuChene proposed a change to openstack/diskimage-builder: Add ability to upload a new GPG into the apt keyring for the image  https://review.openstack.org/8359621:49
SpamapSadam_g: thats where that stuff should come from21:49
adam_gSpamapS, yes21:49
Lotus907efilifeless: what happens now with https://review.openstack.org/83596 ???21:51
SpamapSadam_g: # python -c 'import sys; print(hasattr(sys, "gettotalrefcount"))'21:51
SpamapSFalse21:51
SpamapSadam_g: what do you get?21:51
NgLotus907efi: folk start reviewing it and it mails you :)21:52
*** epim has quit IRC21:52
Lotus907eficool21:52
Lotus907efithis is a first for me21:52
SpamapSLotus907efi: but don't buy any viagra from Gerrit... stuff doesn't work.. ;)21:52
adam_gSpamapS, False21:52
Lotus907efihehehehe21:52
SpamapSadam_g: ok that is _super_ weird then21:52
SpamapSadam_g: yeah, try new image :-P21:53
adam_gSpamapS, yeah, let me tear down and try with a fresh image21:53
*** matty_dubs is now known as matty_dubs|gone21:53
*** eguz has joined #tripleo21:58
*** untriaged-bot has joined #tripleo22:00
untriaged-botUntriaged bugs so far:22:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129846622:00
uvirtbotLaunchpad bug 1298466 in tripleo "devtest cleanup-env doesn't remove seed image" [Undecided,Confirmed]22:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129846522:00
uvirtbotLaunchpad bug 1298465 in tripleo "Baremetal recreation may fool the heat stack." [Undecided,Confirmed]22:00
untriaged-bothttps://bugs.launchpad.net/tripleo/+bug/129048822:00
uvirtbotLaunchpad bug 1290488 in tripleo "Baremetal: Invalid credentials" [Undecided,Incomplete]22:00
*** untriaged-bot has quit IRC22:00
SpamapShah uh oh22:00
SpamapSwe're in trouble now22:00
Nghave we manually paused landing stuff while CI is being updated? (just looking at some stuff that had already had its CI runs)22:02
Nghmm, actually, I need to sleep, so I'll just do that instead22:02
SpamapSNg: go22:02
*** eguz has quit IRC22:02
*** eghobo has quit IRC22:03
*** eghobo has joined #tripleo22:03
*** yamahata has quit IRC22:06
*** rbrady1 has joined #tripleo22:08
*** derekh has joined #tripleo22:08
*** rbrady has quit IRC22:08
lifelessderekh: wehave testenvs now, I think22:09
derekhsweet22:09
*** dprince has joined #tripleo22:09
*** gothicmindfood has joined #tripleo22:10
*** gothicmindfood has left #tripleo22:10
dprinceSpamapS: had an idea while I was running, I think it might be the mac addresses we assign via create nodes22:10
dprinceSpamapS: they aren't low enough perhaps to trigger the bridge to assume their address22:11
dprinceSpamapS: I've got to go but wanted to pass that along...22:11
*** dprince has quit IRC22:13
*** akrivoka has quit IRC22:27
StevenKSpamapS: Yeah, I saw your message late last night, and forgot to change the review at the time.22:27
SpamapS StevenK ACK22:28
*** slagle has quit IRC22:34
*** slagle has joined #tripleo22:36
adam_gSpamapS, hmm. new image seems to help wrt python, but now http://paste.ubuntu.com/7165382/22:36
lifelessoh look, bugs to invalidate ;)22:36
* adam_g tries matching arch22:36
* tchaypo gains tmux-foo22:37
*** jtomasek has quit IRC22:37
SpamapSadam_g: _weird_ .. now I'm wondering if something from the build system is stumbling into the built images22:37
tchayposo many comments on my changes from yesterday. Do we win a prize for getting the most comments?22:37
lifelessderekh: can the RH nodes handle a 500GB flavor ? 2 CPU, 8GB of ram ?22:37
lifelessNg: nuts, I missed you ?22:38
* lifeless wanted more dns love22:38
derekhDo we intend to use the 500GB ?22:39
lifelessderekh: 150GB of ubuntu, a few GB of pypi-mirror, however much for F2022:39
lifelessderekh: a few base cloud images22:39
lifelessderekh: so - one VM of this nature22:39
lifelessderekh: per region22:39
derekhlifeless: I would imagine it would be ok, we got22:41
derekh/dev/md2        853G   41G  769G   6% /22:41
lifelesscool22:41
derekhthe CI nodes wouldn't use up much of the disk space there allocated22:41
lifelessI'll need to convince clarkb or pleia2 to help me with the puppet stuff22:41
lifelessI'm going to prototype using the cd-undercloud's pypi and ubuntu mirrors22:41
pleia2I can help22:42
lifelesscool!22:42
lifelessderekh: do you have the url to the faster CI stuff handy?22:42
derekhlifeless: the faster ci stuff ?22:43
lifelessderekh: we brainstormed in sunnyvale22:43
lifelessderekh: about mirrors etc etc22:43
derekhahh ok, one sec22:43
lifelessin an therpad22:43
lifelessStevenK: https://review.openstack.org/#/c/79051 is that ready in your view?22:43
derekhlifeless: https://etherpad.openstack.org/p/tripleo-parallel-test22:44
lifelesspleia2: ^22:44
*** tchaypo is now known as jammyjam22:44
lifelesspleia2: option 7 in there, filling it out now22:44
jammyjamcasual nick friday!22:45
SpamapSnew trick learned22:52
SpamapS# os-collect-config --force --one --command /opt/stack/os-config-refresh/post-configure.d/99-refresh-completed22:52
SpamapSrun one orc script with full access to the tree :)22:52
jammyjamgreghaynes has objected to me making the setup-*-passwords scripts suddenly depend on $TRIPLEO_ROOT and wants me to instead make -f a mandatory parameter. I think that's a better design, but worried about it breaking backwards compatibility22:54
greghaynesI grepped around and that script is only called one place AFAICT22:56
greghaynesso figured it was a pretty low hanging tech debt prevention22:57
lifelessgreghaynes: hey so hows ha? can I help ?22:57
lifelesspleia2: does that make sense? all -infra work AFAICT other than a cd-admin making a flavor for it22:58
jammyjamgreghaynes: not used in toci anywhere?22:58
pleia2lifeless: looks good, hoping the pypi-mirror stuff in regular -infra can be just reused on our end and the repo mirrors are most of our work22:59
greghaynesprobably - my devtest is exploding when running the notcomputescale > 1 and the dirty heat template with copied notcompute0 -> notcompute1 resource22:59
greghaynesspecifically nova scheduler isnt happy and I think I was also running out of disk space22:59
greghaynesonce this build fails ill link some specific23:00
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325023:00
greghaynesjammyjam: good point, not sure. If it is then I think that just makes the tech debt more of an issue :)23:00
greghaynesjammyjam: but if you do change it and it breaks things / ends up being any amount of effort ill +1 using the variable23:01
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Standardise location of environment password files.  https://review.openstack.org/8325023:02
jammyjamif it's going to break, it should break on this change, right?23:02
StevenKjammyjam: Casual nick Friday breaks my brain.23:02
openstackgerritClint "SpamapS" Byrum proposed a change to openstack/tripleo-image-elements: Add support for "signals" to os-refresh-config  https://review.openstack.org/8361423:02
pleia2StevenK: me too23:02
SpamapS^^ that will unblock move to software-config23:03
SpamapSI think ;)23:03
StevenKLooks like Lenovo are recalling batteries that were made four years ago -- including those sold with the X20123:03
greghaynesjammyjam: dont think so, the arg is still optional there...23:04
greghaynesmybe thats not a bad middle ground23:04
greghaynes*shrug*23:04
greghaynesalso, woo laggy internet makes irc fun23:04
jammyjamgreghaynes: not sure what you mean by "still optional there"23:05
greghaynesah, what I had in mind is making -f a requried option to setup-*-passwords and not referencing the env var $TE_DATAFILE at all in the code23:05
jammyjamI reference $TE_DATAFILE?23:07
greghayneser, sorry, $TRIPLEO_ROOT23:07
StevenK$TRIPLEO_ROOT is exported by _variables, so it doesn't buy you much to not use it?23:09
jammyjamsure, but it's not used anywhere else in the script23:11
greghaynesI get that when its convenient, like in devtest_* but seems like if we can make stuff an actual script which doesnt depend on env vars easily and instead on arguments we should23:11
jammyjamin fact, afaict, none of the common vars are used in the script at all23:11
jammyjamso I agree with greghaynes that we probably shouldn't be slipping backwards23:11
*** eghobo has quit IRC23:15
openstackgerritJames Polley proposed a change to openstack/tripleo-incubator: Add some clarity to the first-time user experience  https://review.openstack.org/8329423:24
greghaynes\O/ this build is not resulting in a bunch of ERRORD overcloud vms23:25
greghaynesI think runninng out of disk space somehow caused the nova scheduler memory check to fail...23:26
jammyjamhah23:26
greghaynesI think runninng out of disk space somehow caused the nova scheduler memory check to fail...23:27
greghaynesoops23:27
greghaynesIts amazing, im literally across the street from the datacenter where my irc box is, yet my packets are going about 1k miles to get there23:28
lifelesspleia2: yes, reuse pypi-mirror, will just want to change it to draw from pypi.o.o as its own upstream23:30
clarkbgreghaynes: I bet my packets would only travel down i523:30
clarkbSIX is pretty amazing23:30
greghaynesclarkb: Where is your irc box?23:31
clarkbgreghaynes: my irc box is in ORD23:31
greghaynesORD?23:31
clarkbyup rackspace23:32
greghaynesoh, psh23:32
jammyjamgreghaynes: part of me wants to investigate this weechat thing you're using. Part of me thinks that clearly my mosh+irssi setup is better.23:32
greghaynesI was hoping you were at PSU, my DC is amazingly close to them so I get <1ms pings to people there ;)23:33
pleia2lifeless: yeah, I'm hoping it'll just be as easy as updating the pypirc.erb template in puppet, we'll see23:33
greghaynesjammyjam: People who customize their irssi a lot tend to hate the weechat, if you use defaults weechat is nicer imo :)23:33
jammyjamI customised irssi a bit because the defaults were so goddamn horrible23:39
*** rpodolyaka has joined #tripleo23:39
jammyjamI hate the fact that just to get a decent little channel list floating on the left I need to use tmux to split a pane and run a seperate thing in the left pane23:40
greghayneshaha, yep23:40
clarkbweechat is pretty awesome23:41
clarkband doesn't preclude the use of mosh23:41
greghaynesand god forbid you were running screen, youd have to patch it to get the vertical split ;)23:41
clarkbsmart filters are the best thing ever23:41
jammyjamnah, debianised screen has included that patch ever since like 15 years after it had been around when it was clear the devs didn't care any more23:42
jammyjamit's only screen on OS X that I regularly had to use that didn't have the split23:42
jammyjamwhich is why i started using iterm, then item's tmux support, and then tmux everwhere.23:42
jammyjamjust this morning I learned how to use list-windows and select-layout to save/restore a layout in tmux23:42
jammyjamhah. looking at weechat screenshots and i recognise nicks23:43
*** rpodolyaka has quit IRC23:44
greghayneslol23:44
lifelessnuts23:44
lifeless81959 passed ironic seed. \o/23:45
jammyjamhttps://review.openstack.org/#/c/82807/ says that its dependency is "outdated"; I've been reading gerrit docs but I can't figure out if that's a "i can't merge this" error, or just a "hey, that thing you like, it changed a bit" notification23:45
lifelessbut failed at undercloud23:45
lifelesshttps://review.openstack.org/#/c/81959/23:45
greghaynesI think that means it needs to be rebased23:45
greghaynesthe dependency23:45
lifelessSpamapS: ^ think we can land this and iterate on the undercloud job ?23:45
lifelesse.g. lets make ironic undercloud non-voting still23:46
lifelessassuming overcloud-precise passes23:46
lifelessNobodyCam: so CI system is fixed up23:47
lifelessdevananda: ^23:47
lifelessthe failures now are genuine23:47
NobodyCam:) w00t Thank you lifeless :)23:47
*** shardy has quit IRC23:51
*** shardy has joined #tripleo23:51
Lotus907efilifeless: about a month or so ago you gave me a fairly complete step-by-step directions for two different ways to boot a baremetal node23:51
Lotus907efiI am trying to figure out to duplicate that again but use an image from our internal precise apt repo23:52
Lotus907efiI can build an image from that repo using vmbuilder23:52
Lotus907efiand I end up with a qcow2 image23:53
*** mestery has quit IRC23:53
Lotus907efibut I am trying to figure out how to do this somehow within tripleo environment23:53
SpamapSlifeless: will look in a bit.. mid-stream on some long-running-tests-to-confirm fixes :)23:53
*** wfoster has quit IRC23:54
* jammyjam is now in weechat as well as irssi23:55
*** stevehuang has quit IRC23:55
clarkbjammyjam: turn on smart filters23:56
clarkbjammyjam: single best weechat feature23:56
*** wfoster has joined #tripleo23:56
lifelessSpamapS: ack23:57
lifelessLotus907efi: youneed to extract the ramdisk and kernel and place them in the same dir as the qcow2 file and then use load-image, which is in the incubator scripts dir23:58
*** derekh has quit IRC23:58
Lotus907efihmmm, ok23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!