Friday, 2016-02-26

*** chlong_ has quit IRC00:03
*** ayoung has quit IRC00:04
*** penick has quit IRC00:05
*** eil397 has joined #tripleo00:32
*** Goneri has joined #tripleo00:49
*** lblanchard has joined #tripleo00:52
*** ayoung has joined #tripleo00:52
*** cwolferh has quit IRC00:55
*** eil397 has quit IRC00:57
*** saneax is now known as saneax_AFK01:13
*** _{Tite}_ has joined #tripleo01:14
_{Tite}_hi there ... anybody with a minute to help me jumpstart the incubator ? can't seem to get the seed's nova to work01:15
*** cwolferh has joined #tripleo01:16
*** dmacpher-afk has quit IRC01:21
*** shivrao has quit IRC01:22
*** shivrao has joined #tripleo01:37
openstackgerritOpenStack Proposal Bot proposed openstack/python-tripleoclient: Updated from global requirements
openstackgerritOpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements
*** shivrao has quit IRC01:46
*** chlong_ has joined #tripleo01:49
*** Goneri has quit IRC01:50
*** _{Tite}_ has quit IRC01:57
*** dmacpher has joined #tripleo02:08
*** tiswanso has joined #tripleo02:15
*** tiswanso has quit IRC02:16
*** tiswanso has joined #tripleo02:17
*** yamahata has quit IRC02:39
openstackgerritMerged openstack/tripleo-heat-templates: Use pymysql database driver for OpenStack DBs
*** rhallisey has quit IRC02:44
*** cwolferh has quit IRC03:05
*** ianw has quit IRC03:16
*** ianw has joined #tripleo03:19
*** lblanchard has quit IRC03:21
*** links has joined #tripleo03:26
*** panda has quit IRC03:37
*** panda has joined #tripleo03:38
*** masco has joined #tripleo03:53
*** Marga_ has quit IRC03:57
*** Marga_ has joined #tripleo03:58
*** tzumainn has quit IRC04:01
*** mandre has quit IRC04:02
*** Marga_ has quit IRC04:03
*** mandre has joined #tripleo04:07
*** masco has quit IRC04:11
*** Marga_ has joined #tripleo04:19
*** Marga_ has quit IRC04:21
*** Marga_ has joined #tripleo04:21
*** masco has joined #tripleo04:23
*** tiswanso has quit IRC04:48
*** dustins has quit IRC04:56
*** saneax_AFK is now known as saneax05:15
*** anande has joined #tripleo05:37
*** anande has quit IRC05:40
*** shinobu__ has quit IRC05:46
*** Marga_ has quit IRC06:00
*** Marga_ has joined #tripleo06:17
*** links has quit IRC06:26
*** rlandy has quit IRC06:42
*** aufi has joined #tripleo06:52
*** rcernin has joined #tripleo07:05
*** dmacpher has quit IRC07:09
*** chlong_ has quit IRC07:26
*** panda has quit IRC07:38
*** panda has joined #tripleo07:38
*** jprovazn has joined #tripleo07:49
*** athomas has joined #tripleo07:50
openstackgerritDougal Matthews proposed openstack/tripleo-common: Use PyYAML's safe_dump to avoid outputting Python specific values
*** jtomasek has joined #tripleo07:52
*** mkovacik has quit IRC07:58
*** ohamada has joined #tripleo07:59
*** jtomasek_ has joined #tripleo07:59
*** fgimenez has joined #tripleo08:03
*** fgimenez has quit IRC08:03
*** fgimenez has joined #tripleo08:03
*** rdopiera has joined #tripleo08:09
*** dshulyak has joined #tripleo08:20
*** mikelk has joined #tripleo08:22
openstackgerritIshant Tyagi proposed openstack/os-collect-config: Add insecure option to the cfn collector
*** ifarkas has joined #tripleo08:29
*** masco has quit IRC08:36
*** xinwu has quit IRC08:38
*** hjensas has joined #tripleo08:41
*** mkovacik has joined #tripleo08:44
*** jtomasek_ has quit IRC08:48
*** masco has joined #tripleo08:50
*** lucas-afk is now known as lucasagomes09:02
*** mcornea has joined #tripleo09:06
*** jcoufal has joined #tripleo09:10
*** rcernin has quit IRC09:14
*** bvandenh has joined #tripleo09:16
*** masco has quit IRC09:17
*** gfidente has joined #tripleo09:22
*** jistr has joined #tripleo09:27
*** rcernin has joined #tripleo09:28
*** electrofelix has joined #tripleo09:30
*** masco has joined #tripleo09:33
jistrgfidente, marios: good morning guys :) we have some upgrade patches to land -- 3 to stable/liberty starting here and one to master here (gfidente already reviewed that one)09:34
*** ishant has joined #tripleo09:35
jistralso marios' compute patch is ready code-wise, but red CI, looking at that now
ishantHi folks, need review for os-collect-config patch
*** nico_auv has joined #tripleo09:36
mariosjistr: going to revisit my review momentarily to update for your comments from last night (+ on the tripleo-common patch)09:37
jistrmarios: ok. the CI failure is unrelated -- unreachable centos mirror09:38
mariosjistr: fail on ... yeah i was just poking09:38
jistrand on tripleo-common it's some delorean issue, not sure if that's intermittent or not09:38
*** bvandenh has quit IRC09:46
jistrmarios: wdyt about the script naming issue -- i was thinking it might be easiest if Heat uploaded the correct per-role script to the machines, but it on the machines it would always be named the same regardless of role, so that the tripleo-common script wouldn't have to do some sort of detection what script to run (not sure right now how to best/reliably implement09:48
jistrsuch detection). Just a thought :)09:48
*** bvandenh has joined #tripleo09:50
mariosjistr: yeah will include09:51
openstackgerritMerged openstack/tripleo-heat-templates: Add resources for major upgrade in Pacemaker scenario
jistrgfidente: thanks! ^09:54
jistrgfidente: the other two just need a +2, then we can recheck (somehow jenkins didn't run the unit tests on it) and when unit test results come in, i can +A myself09:55
*** mgould has joined #tripleo09:56
openstackgerritMerged openstack/tripleo-heat-templates: Add a sample network-environment.yaml file to environments
*** paramite has joined #tripleo10:00
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: Write the compute upgrade script for tripleo major upgrade workflow
*** social has joined #tripleo10:00
*** r-mibu has quit IRC10:07
*** r-mibu has joined #tripleo10:08
mariosjistr: i think having the same target filename is a good idea @ going to update both10:08
jistrmarios: cool. Just commented on the other review with the same thing, but +2'd anyway. Will +2 the new patch set too :)10:08
mariosjistr: (we can still deliver *per node* using the conroller upgrade workflow). k thanks10:09
jistrmarios: yeah, deliver per node, but we can name the file the same so that tripleo-common doesn't have to care10:10
mariosjistr: right10:10
mariosjistr: little thing like this can make the script much simpler (having to cater for alternate file targets for each node type)10:10
jistrmarios: can you just +2 the backports here  -- they are both clean10:11
jistrmarios: i will then make sure that unit tests run on those, and take care of merging myself. I'd like to tell QE today "use stable/liberty" instead of "use stable/liberty and cherry-pick this and that"10:12
mariosjistr: looking10:12
*** derekh has joined #tripleo10:13
jistrmarios: perfect, thanks. I'm chasing down Jenkins now :D10:15
mariosjistr: cherrypick clean and clean ci run which is rare on stable these days10:15
jistrheh yes10:15
mariosjistr: so since cherrypicks just landed10:15
jistrmarios: didn't land yet, there are missing unit tests, jenkins won't merge. I'm trying to get it to "verified +1" via recheck. Will add another +A right after the unit tests arrive, since we do have the long-CI green.10:16
mariosjistr: ah sry i saw the green jenkins run10:17
marios(well not containers but expected)10:17
openstackgerritMarios Andreou proposed openstack/tripleo-heat-templates: Write the compute upgrade script for tripleo major upgrade workflow
openstackgerritMerged openstack/tripleo-heat-templates: Introduce update/upgrade workflow
*** athomas has quit IRC10:20
openstackgerritMerged openstack/tripleo-heat-templates: Add UpgradeLevelNovaCompute parameter
jistrmarios, gfidente: thanks!!10:22
gfidenteoh man10:22
openstackgerritGiulio Fidente proposed openstack/tripleo-heat-templates: Add meta notify=true to rabbitmq resource
gfidentebandini ^^10:23
jistrmarios: +2'd tht part, i think that's final :)10:23
*** athomas has joined #tripleo10:24
bandinigfidente: oh wowsers the one in master passed CI?10:25
*** masco has quit IRC10:25
bandinithat + it's Friday..I know exactly which bottle I will have10:25
jistr+2'd the backport of notify=true10:26
* jistr +2 all the things10:26
bandinijistr: I am mostly done with the constraint changes, deploy is fine and from a quick look constraints seem fine. Am trying to do more testing. Ok if I submit on Monday?10:27
openstackgerritMarios Andreou proposed openstack/tripleo-common: Adds a bash script that operators will use during upgrades
*** ishant has quit IRC10:27
mariosjistr: thx10:28
jistrbandini: sure thing, whenever it's ready. Thanks for looking into that!10:28
openstackgerritMerged openstack/tripleo-heat-templates: Add meta notify=true to rabbitmq resource
*** dtantsur|afk is now known as dtantsur10:30
openstackgerritMichele Baldessari proposed openstack/tripleo-heat-templates: Remove the delay resource and its constraints
bandinijistr: this ^ new review is just something I noticed while working on the constraints10:37
bandinijistr: basically right now delay-clone is just a useless leaf (
jistrbandini: cool. could you include that in the migration bash script later too? Essentially, for every change we do to HA in puppet, we need a corresponding migration for existing deployments somewhere (either update or upgrade scripts, depending whether the change is meant for backporting (serious bugs) or not (convenience changes, new features))10:39
bandinijistr: ah righto10:40
bandinijistr: let me abandon that change then. I first need to setup a test env to test these upgrade and I will add that as well10:40
jistrbandini: you can set it to "Workflow -1" which means work in progress10:41
mariosjistr: fyi (see comment v3, customizability of the upgrade script was a main concern for you)10:41
bandinijistr: ah good to know. done10:41
*** tosky has joined #tripleo10:43
*** masco has joined #tripleo10:44
gfidentebandini, thanks for fwiw :)10:45
mariosgfidente: thanks for looking I think I addressed your comments10:45
mariosgfidente: appreciate a repass when you can10:45
gfidentemarios, I didn't get we needed it in a known location10:46
gfidenteisn't the first creating it with ENDOFCAT and the second with scp?10:46
mariosgfidente: we just invented it (seriously makes sense we think(10:46
mariosgfidente: see discussion in the other review too10:47
bandinigfidente: :)10:50
mariosgfidente: also thanks10:51
openstackgerritMarios Andreou proposed openstack/tripleo-common: Adds a bash script that operators will use during upgrades
gfidentemarios, so it runs from the controller10:53
gfidentebut need undercloud stackrc?10:53
mariosgfidente: there are two scripts10:53
gfidenteok I think I see it now10:54
gfidentethe one for the controllers is delivered by heat, the one for the non-controllers is scp-ed?10:54
*** saneax is now known as saneax_AFK10:55
mariosgfidente: we distinguish between controllers and non-controllers for the upgrade workflow10:56
mariosgfidente: controllers are upgraded via heat with an upgrade environment file that delivers an upgrade wokflow (remove from cluster pinning etc)10:56
mariosgfidente: during that upgrade workflow we deliever to non-controllers an upgrade script to /root/tripleo_upgrade_node.sh10:56
mariosgfidente: that will later be invoked by the operator, after the controllers are upgraded, from the undercloud10:57
mariosgfidente: the second script is what the operators will be using to invoke the /root/ on the non-controllers.10:57
mariosgfidente: hope it helps10:57
mariosgfidente: so the review at is what delivers the file to /root/tripleo_upgrade_node.sh10:58
mariosand the review at is the bash script the operator runs from the undercloud which will invoke that /root/tripleo_upgrade_node.sh10:59
gfidenteso the first review delivers a file on the computes11:00
openstackgerritGabriele Cerami proposed openstack/tripleo-heat-templates: mysqld config: set innodb_file_per_table to ON
openstackgerritMarios Andreou proposed openstack/tripleo-common: Adds a bash script that operators will use during upgrades
mariosgfidente: yeah for now just computes. we can add delivery there to the other nodes too. but the location will be the same. /root/tripleo_upgrade_node.sh11:00
mariosgfidente: and the same undercloud script to run that on your non-controller11:01
gfidenteoh the scp in deliver_script is not run by default unless operator passes --script11:02
mariosgfidente: right11:02
gfidenteand it's meant to override the default >11:02
mariosgfidente: it it an override11:02
mariosgfidente: seem comments11:02
mariosgfidente: see comments on the review please11:03
mariosgfidente: on set 3 i explain how --script and $UPGRADE_SCRIP works11:03
openstackgerritGabriele Cerami proposed openstack/instack-undercloud: mysqld config: set innodb_file_per_table to ON
gfidentemarios, jistr ok thanks for explaining11:05
gfidentestill I wonder if having those in /root is a good idea11:05
*** ifarkas_ has joined #tripleo11:08
mariosgfidente: yeah i share that concern, mainly I think amplified by us making it confiugurable (so you can upload what you like to be executed there)11:09
*** ifarkas has quit IRC11:11
mariosgfidente: but i think we need to try it out to prove the workflow (we are still working things out). we can remove that configurability for the file later if we keep this. jistr wdyt11:11
gfidenteI also wonder if it makes sense for deliver to skip controllers11:12
mariosgfidente: would you agree that not allowing the op to overwrite /root/ would help?11:12
gfidentewe never know11:12
mariosgfidente: it never goes to controllers11:12
gfidenteyes but we don't know if it never will I mean11:12
mariosgfidente: yeah right. well we already have/make certain expectations on the names of nodes11:12
*** masco has quit IRC11:12
mariosgfidente: so i can expect the convention to be overcloud-controller-0/1/2/3 etc11:13
mariosgfidente: otherwise, no unless we register mac addresses  we can't know11:13
gfidenteyes so  I am thinking should really skip controllers?11:13
mariosgfidente: yes it should11:14
mariosbecause it is .. ahem11:14
gfidenteyeah but as long as we don't provision some upgrade script11:14
mariosgfidente: what are you thinking (some scenario?)11:14
gfidenteon the controller, it'll just skip the controller11:14
gfidenteI'm thinking we might want to run a script on the controllers11:14
mariosgfidente: yeah ok. but this isn't about providing a way for the op to send a nice script to his nodes11:15
gfidenteyes it is!11:15
mariosgfidente: this is a very specific thing for the upgrades11:15
mariosgfidente: and as part of that workflow, this will only be relevant to non-controllers.11:16
mariosgfidente: the script is already at 120lines doing a very specific thing. i don't what it to be any longer/complex11:16
gfidenteyeah so not skipping controllers would make it shorter no?11:16
*** leanderthal|afk is now known as leanderthal11:17
mariosgfidente: yes but then you would want to override the target location because what use would a nice mechanism for sending arbitrary script to your overcloud nodes would be complete without that etc11:17
gfidenteI like that it can overwrite and provision arbitrary things11:19
gfidenteit's what we needed for 55-heat-config11:19
gfidenteso yes I was thinking about it as a more generic hack11:19
gfidenteI think it might somewhat conflict with the deployartifacts thing11:20
gfidentebut feel more user-friendly11:20
mariosgfidente: so this review is NOT about implementing or providing anythign other than the ability for the operator to invoke the upgrade script at /root/tripleo_upgrade_node.sh11:20
mariosgfidente: not sure how elese to say it11:21
marioson *non controllers*11:21
mariosgfidente:  a simple thing like the ability to set arbitrary permissions on the file you're transferring will add complexity here.11:22
mariosgfidente: if once we've decided to use this for the upgrades, you think it can evolve into a more general thing, without breaking what it is intended for originally, then i think it can happen later imo11:23
*** masco has joined #tripleo11:25
*** trown|outtypewww is now known as trown11:27
*** panda has quit IRC11:37
*** panda has joined #tripleo11:38
*** dshulyak has quit IRC11:51
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Add mirror server
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: [NO MERGY] Test a fake periodic job
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Used the mirror server for the centos image and git repos
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Upload the instack qcow image after periodic jobs
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Use instack-qcow2 from cache if available and appropriate
openstackgerritMerged openstack-infra/tripleo-ci: Use consistent delorean symlink for periodic job
jistrmarios: +2'd the tripleo-common part and weighed in on the controller vs. non-controller thing. If we need to amend it in the future because of that, we can. ATM i don't see a use case for controllers there, we're attempting to do everything we need via Heat.12:20
*** thrash|g0ne is now known as thrash12:22
derekhlow hanging review fruit
mariosjistr: thanks12:24
mariosderekh: looking12:25
openstackgerritMerged openstack-infra/tripleo-ci: Redirect undercloud install output to file
EmilienMderekh: how is tripleo CI going? no bug with puppet-neutron?12:27
EmilienMjust checking because we changed some default parameters lately and just to make sure it's allright here12:28
derekhEmilienM: none that I am aware of12:28
derekhEmilienM: thanks for checking ;-)12:28
*** lblanchard has joined #tripleo12:39
*** dshulyak has joined #tripleo12:40
openstackgerritEmilien Macchi proposed openstack/instack-undercloud: Use pymysql database driver for OpenStack DBs
EmilienMslagle: I changed the Depends-On, but I think it's useless since we still use OPM in stable/liberty, right?12:45
*** julim has joined #tripleo12:45
slagleEmilienM: probably, yes.12:48
slagleEmilienM: can we just get the needed backport into opm?12:49
EmilienMslagle: working on it.12:50
openstackgerritMerged openstack/tripleo-common: Use PyYAML's safe_dump to avoid outputting Python specific values
*** rhallisey has joined #tripleo12:53
*** masco has quit IRC12:55
*** gfidente has quit IRC12:56
*** dprince has joined #tripleo12:57
*** rlandy has joined #tripleo13:01
*** masco has joined #tripleo13:08
*** lucasagomes is now known as lucas-hungry13:09
*** jayg|g0n3 is now known as jayg13:16
dtantsurfolks, are there known problems with the ha job today?13:22
trownI think it only has problems on days ending in y13:24
trowndtantsur: but not looking great
trownI do see a few passes from today though13:25
trownso might just be the usual (HA is hard with limited resources)13:26
dtantsurso it's worth trying I guess :)13:26
derekhtrown: them 4 fails in a row are my borked patch series, that is ci doing its job correctly ;-)13:26
trownderekh: ya, had to scroll down to see some green13:26
*** aufi has quit IRC13:28
*** pradk has joined #tripleo13:31
jistrmarios: fyi i triggered a recheck on
*** rlandy has quit IRC13:38
dprincederekh: regarding your comment here I agree.13:42
dprincederekh: I think network isolation is probably broken on trunk13:42
*** pradk has joined #tripleo13:43
derekhdprince: yup13:44
*** aufi has joined #tripleo13:45
*** dmacpher has joined #tripleo13:52
*** rlandy has joined #tripleo13:52
*** masco has quit IRC13:54
*** Goneri has joined #tripleo13:55
hewbroccaderekh, dprince hey, qq for you guys14:09
hewbrocca bug 1312342 in openstack-nova "Deployment configuration for "cell of one"" [High,New] - Assigned to eglynn14:09
hewbroccaSo... Nova in Mitaka is going to require some new database configuration or something14:10
dprincehewbrocca: we are already doing it14:10
hewbroccayou guys rock14:10
hewbrocca ?14:11
dprincehewbrocca: and your commit too which uses the puppet-nova change14:11
*** tzumainn has joined #tripleo14:18
*** tiswanso has joined #tripleo14:18
*** lucas-hungry is now known as lucasagomes14:23
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Convert the container job to a noop
rhalliseyderekh, probably for the best until I can fix it.. having issues with OVS atm14:30
rhalliseyderekh, noop for the time being and I'll try and get it fixed soon14:31
derekhrhallisey: yup, when your working on it you should be able to test it with a patch to remove the exit 1 and a depends on14:32
rhalliseyderekh, ya I'll let you know14:33
derekhrhallisey: thanks14:33
*** hjensas has quit IRC14:37
*** dustins has joined #tripleo14:43
pradkcan anyone here help me understand why ci is failing with these errors on Aodh patch?
pradkseems like httpd throwing these ? "Could not evaluate: Cannot allocate memory - fork(2) "14:46
hewbroccapradk: a good guess would be that CI is out of memory14:50
hewbroccajust a guess...14:50
derekhpradk: does Aodh have a large memory footprint ?14:52
pradkhewbrocca, yea that would have been my guess, but interestingly i dont see other patches fail with the same.. so wondering if there is something unique here, we're loading aodh api under wsgi i dont know if thats causing issues14:52
pradkderekh, nothing unusual afaik14:53
derekhCan somebody merge this  or this and we can see how much RAM the various process are using14:54
derekhpradk: If either of those get merged ^^ do  a recheck and we will have a little more info to go on14:54
pradkderekh, cool, will do14:55
openstackLaunchpad bug 1550380 in tripleo "network isolation broken when SoftwareConfigTransport == POLL_TEMP_URL" [Critical,New] - Assigned to Dan Prince (dan-prince)14:58
derekhdprince: ack14:59
dprincederekh: I haven't been using network isolation for a month or so (I was developing Mistral....). When I just tried switching back this morning it wasn't working for me either :/15:00
*** mkovacik has quit IRC15:03
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Set Heat transport to CFN when using netiso
openstackgerritDan Prince proposed openstack-infra/tripleo-ci: Use netiso in the ha job
*** paramite is now known as paramite|afk15:20
jistrmarios: could you review the Nova unpinning when you have a minute please?
mariosjistr: sure. had a look already yesterday, didn't realise i forgot to vote15:29
jistrand one of yours is ready to land too, just needs one more +2
*** dprince has quit IRC15:35
mariosjistr: i'll pimp it on the call in a bit thanks15:36
mariosjistr: your review lgtm i think we can land it now15:36
jistrmarios: cool, thanks15:36
mariosjistr: i put a clarification there, mostly for gfidente question15:36
*** panda has quit IRC15:37
hewbroccaderekh: hey, quick question for you15:38
hewbroccaderekh: you have any idea how difficult it will be for us to get IPv6 coverage into TripleO CI?15:38
jistrmarios: replied on the review15:38
*** panda has joined #tripleo15:38
*** jprovazn has quit IRC15:39
derekhhewbrocca: no clue, but assuming all of the ipv6 traffic is happening between the undercloud to the overcloud I don't see any immediate roadblocks15:39
derekhhewbrocca: besides that I'm not sure, I've never tried ipv6 at all15:40
hewbroccaWell, we have to configure the public APIs on the overcloud using IPv615:40
hewbroccaand then test that works15:40
openstackgerritMerged openstack/tripleo-heat-templates: Nova RPC unpinning
derekhhewbrocca: yup, calls to the overcloud API's would be from the undercloud so that should be possible15:41
*** adarazs has joined #tripleo15:43
hewbroccaAll right well Attila is on the way to get it together15:43
adarazsderekh: hi, what do you think about this?
hewbroccaderekh: any help you can give would of course be much appreciated15:44
adarazsderekh: because we have downstream IPv6 jobs, so that would be relatively easy to push to upstream and make a gate out of it, but it won't be a super fast gate :/15:44
derekhhewbrocca: ack15:44
trown-1 to khaleesi upstream just for ipv615:44
openstackgerritMerged openstack/tripleo-image-elements: Deprecated tox -downloadcache option removed
derekhadarazs: are we hoping to add a new job for this or adjust one of our existing jobs ?15:45
adarazsderekh: well, that's my question. I have some upstream jobs that are silent (i.e. they are not reporting, but I'm running them) like this: so I could add a khaleesi gate to upstream that gates any repo, and I have the method to run IPv6 jobs.15:47
trownadarazs: there is no way that centosci can handle every tripleo commit15:48
adarazstrown: it might. but we're just talking about ipv6 gating foe now.15:48
trownadarazs: ya, but how would you determine what changes to run against if not all of them?15:49
trownthis needs a spec if we are introducing third-party CI for features15:49
derekhadarazs: would it be possible to have that existing job report as 3rd party ci ?15:50
adarazsderekh: yes, I just have it set to silent, but it could report already. I have everything set up for it, I just didn't pull the trigger.15:51
adarazsderekh: but that job is not ipv6, just a simple khaleesi run with minimal settings.15:51
hewbroccaahh but15:52
*** ifarkas_ has quit IRC15:52
hewbroccaif you made it an ipv6 job... ?15:52
derekhSo we got 3 options here,15:52
adarazshewbrocca: I could, that's what I'm saying, it's probably not too difficult.15:52
derekh1. make you job a 3rd party ci and start reporting, we need to then pay attention to it and treet it as gating15:52
adarazsor just add an new one that's ipv615:52
derekh2. add a new job and take the hit, at peak times accepting that we may hit capacity limits from time to time15:53
derekh3. add ipv6 to an existing job, for example include it in the upgrades job that is being created, it would then be a upgrades/ipv6 job15:54
trownmy vote is for 315:54
trown1 will not actually work... unless we are only gating instack-undercloud... that is all that the job that is silent currently runs against15:55
adarazsokay, so 3) is something I'm least familiar with, so I might need more guidance/time for it to happen.15:55
trownwe WILL explode centosci if we gate on every tripleo commit15:55
*** penick has joined #tripleo15:55
adarazsderekh: also we should figure out what repos we want gating, this for now is only instack-undercloud, but I can create more with different repos if we don't need to push the packages to the overcloud for proper testing. I don't have that mechanism implemented, they only get installed on the undercloud.15:55
derekhadarazs: 3) would involve adding ipv6 on top of this patch to test upgrades
trownadarazs: I dont think that part should be a problem... all tripleo packages are undercloud15:56
trownit gets a bit weird with depends-on though15:56
adarazsderekh: why should we add ipv6 to the upgrade job? just to not add another job and save resources?15:57
derekhadarazs: yup15:58
*** yamahata has joined #tripleo16:00
hewbroccaderekh, adarazs so... how hard is that going to be?16:00
derekhadarazs: I mean we could add a new job, but I reckon we'll cause things to queue up a peak times16:00
adarazsderekh: okay. so how much work do you think adding ipv6 there would be? do I see it correctly that it might be only a matter of including the proper env file for ipv6?16:00
*** aufi has quit IRC16:00
derekhadarazs hewbrocca I don't know how hard it is to actully enable ipv6 /me hasn't ever tried it16:01
derekhI'm trying to help figure out where it can go16:01
derekhthen we can figure out how16:01
hewbroccaYou might have one other issue16:01
* hewbrocca thinks16:02
* adarazs tried it and it wasn't all that bad at least in khaleesi, but I'm not familiar with the tripleo-ci scripts, so I might not see the hidden problems.16:02
hewbroccathe TripleO CI test is, deploy a heat stack and ping the resulting VM16:02
hewbroccayeah, nm, that's fine16:02
hewbroccareally just knowing that the cloud comes up is sufficient for now, I would think16:02
derekhadarazs: so what needs to be added to the various openstack commands for it to work? and we can figure out what needs to be updated and where16:03
hewbroccaI mean it would be *nice* if we could ping a VM listening on IPv616:03
adarazsderekh: yeah, what I needed mostly was this for khaleesi:
*** paramite|afk is now known as paramite16:04
adarazshewbrocca: for that we will have to make sure the floating IP networks are properly created as IPv6, that will take a little more fiddling probably.16:05
hewbroccano need to worry about that right now16:05
derekhadarazs: If that is then it should be just a matter of adding the parameters to OVERCLOUD_DEPLOY_ARG here
adarazsderekh: plus we need to feed it this:
derekhadarazs: I mean I'm sure we'll hit other problems but thats the start16:07
adarazsderekh: yep, that's what I figured.16:07
*** rpothier has joined #tripleo16:07
adarazsderekh: okay, so I will try to work on this monday next week and see where it gets me.16:07
adarazsderekh: thanks for your insights. :) (+trown's)16:08
derekhadarazs: ok, you can start by adding that to the nonha job and seeing if you can get the patch to pass that can happen in parallel to figuring out where exactly to put it16:08
derekhadarazs: no prob16:09
trownya good idea to start with nonha job, since that one rarely fails without good reason16:09
hewbroccaadarazs: keep in mind thatyep16:11
adarazsokay, I'll add a summary on the card about this. ( )16:11
hewbroccaadarazs: keep in mind that TripleO CI doesn't run network isolation yet16:13
hewbroccaadarazs: so whatever env file you pass, you need to take out the multi networks bit16:14
adarazshewbrocca: hmm, I don't think we can do it without, the IPv6 env file is a single_nic_vlan network isolation one...16:15
adarazshewbrocca: but using vlans is okay?16:15
hewbroccaderekh: ^^^ I dunno16:16
derekhhewbrocca: dprince has pushed a patch to hopefully fix netiso and gfidente has a path to start testing it
hewbroccaoh, no shit16:17
hewbroccaderekh: I thought you guys had to actually reconfigure the rack to do that16:18
hewbroccaso, that's very good16:19
hewbroccaadarazs: OK, what you need to do16:19
derekhhewbrocca: so did we but then people started talking about it and reckoned it should work, I'm getting vague here I know /me is lost when it some to some of this stuff having not looked into most of it16:20
hewbroccaadarazs: when dprince gets back online, connect with him on adding the IPv6 env to the upgrade job16:21
hewbroccaThat means... we need to land net-iso, and then we can land IPv616:21
adarazshewbrocca: ookay. :)16:22
openstackgerritMerged openstack/diskimage-builder: Install IPA in a virtual environment
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Add mirror server
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Used the mirror server for the centos image and git repos
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Use instack-qcow2 from cache if available and appropriate
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Upload the instack qcow image after periodic jobs
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: [NO MERGY] Test a fake periodic job
*** rcernin has quit IRC16:31
*** penick_ has joined #tripleo16:37
*** penick has quit IRC16:40
*** penick_ is now known as penick16:40
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: Function library for major upgrades
*** mgould has quit IRC16:42
jistrbandini: ^ created a stub place where bash functions for the pcmk migrations can go. We'd then call those bash functions from our upgrade scripts as needed. Perhaps removing the delay can be a separate function from doing the WSGI migration. The functions probably need to be idempotent (we've already done idempotent constraint changes as part of minor updates, so i'm hoping idempotency shouldn't be too hard to achieve, but in case it is then16:44
jistrwe can sync up and try to figure out a different battle plan)16:44
jistrbandini: anyway, have a nice weekend!16:46
*** rdopiera has quit IRC16:50
bandinijistr: yes let's sync up on monday. Hopefully I manage to test one update run tonight, otherwise it will be monday16:51
* bandini bookmarks16:51
*** Marga_ has quit IRC16:51
jistrbandini: yeah no stress, enjoy the weekend, tty on monday16:52
*** Marga_ has joined #tripleo16:52
*** mikelk has quit IRC16:53
bandinijistr: aye aye, you too!16:53
*** trown is now known as trown|lunch16:56
*** Marga_ has quit IRC16:57
*** Marga_ has joined #tripleo16:57
*** mgould has joined #tripleo16:58
*** Marga_ has quit IRC17:02
*** electrofelix has quit IRC17:04
*** hewbrocca has quit IRC17:06
*** cwolferh has joined #tripleo17:06
*** xinwu has joined #tripleo17:07
*** dprince has joined #tripleo17:07
*** fgimenez has quit IRC17:08
*** dshulyak has quit IRC17:10
*** michchap_ has joined #tripleo17:11
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Use instack-qcow2 from cache if available and appropriate
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Upload the instack qcow image after periodic jobs
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: [NO MERGY] Test a fake periodic job
*** yamahata__ has joined #tripleo17:13
*** nico_auvv has joined #tripleo17:14
*** pleia2_ has joined #tripleo17:15
*** dmacpher is now known as dmacpher-afk17:16
*** george_goh_ has joined #tripleo17:17
*** Nakato_ has joined #tripleo17:17
*** r-mibu_ has joined #tripleo17:18
*** stevebak` has joined #tripleo17:19
*** panda has quit IRC17:20
*** r-mibu has quit IRC17:20
*** nico_auv has quit IRC17:20
*** Tristitia has quit IRC17:20
*** michchap has quit IRC17:20
*** Nakato has quit IRC17:20
*** isq_ has quit IRC17:20
*** onovy has quit IRC17:20
*** lucasagomes has quit IRC17:20
*** bswartz has quit IRC17:20
*** afazekas has quit IRC17:20
*** pleia2 has quit IRC17:20
*** stevebaker has quit IRC17:20
*** george_goh has quit IRC17:20
*** onovy_ has joined #tripleo17:20
*** absubram has joined #tripleo17:20
*** afazekas has joined #tripleo17:21
*** onovy_ is now known as onovy17:21
*** absubram_ has joined #tripleo17:21
openstackgerritMerged openstack/tripleo-common: Add requirement for gcc to delorean setup
*** lucasagomes has joined #tripleo17:22
*** lucasagomes_ has joined #tripleo17:22
*** lucasagomes has quit IRC17:22
*** lucasagomes_ is now known as lucasagomes17:23
*** Tristitia has joined #tripleo17:23
*** absubram has quit IRC17:25
*** absubram_ is now known as absubram17:25
*** panda has joined #tripleo17:27
*** dshulyak has joined #tripleo17:27
*** ohamada has quit IRC17:31
*** bvandenh has quit IRC17:38
openstackgerritDan Sneddon proposed openstack/tripleo-heat-templates: Add IPv6 Support to Isolated Networks
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: [NO MERGY] Test a fake periodic job
*** dshulyak has quit IRC17:46
*** eil397 has joined #tripleo17:48
*** athomas has quit IRC17:49
openstackgerritJiri Stransky proposed openstack/tripleo-heat-templates: Nova RPC unpinning
*** eil397 has left #tripleo17:49
*** dtantsur is now known as dtantsur|afk17:49
jistrhave a nice weekend all o/17:49
*** jistr has quit IRC17:49
*** tosky has quit IRC17:59
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Source undercloud environment variable from a file
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Split the deploy script into its own file
*** derekh has quit IRC18:00
*** mcornea has quit IRC18:02
*** Marga_ has joined #tripleo18:10
*** Marga__ has joined #tripleo18:11
*** Marga__ has quit IRC18:12
*** Marga_ has quit IRC18:12
*** Marga_ has joined #tripleo18:13
*** Marga_ has quit IRC18:13
*** jcoufal has quit IRC18:14
*** trown|lunch is now known as trown18:14
*** Marga_ has joined #tripleo18:14
*** xinwu has quit IRC18:25
*** mgould has quit IRC18:28
*** tiswanso has quit IRC18:33
*** tiswanso has joined #tripleo18:34
dprincebnemec: so network isolation is working for you locally?18:34
dprincebnemec: if so I'm curious which network/config templates you are using18:35
*** dshulyak has joined #tripleo18:35
bnemecdprince: Yeah, I did a deployment with the SoftwareConfigTransport: POLL_TEMP_URL set in my resource registry earlier this week.18:35
bnemecdprince: I'm actually using custom templates because my environment is weird.18:35
dprincebnemec: okay, so there is a difference18:36
dprincebnemec: if you get a chance to kick off a stock upstream template set I'd be interested if it works for you18:36
bnemecdprince: I can't use any of the stock upstream templates.  They don't work in any of my environments.18:37
bnemecLike I said, weird. :-)18:37
dprincebnemec: single nic with vlans wouldn't work for you?18:38
dprincebnemec: oh, perhaps your switch doesn't do vlans?18:38
bnemecdprince: I have a vlan-capable switch, but only my baremetal boxes are on it, so I have to untag the external traffic in order to give my VM undercloud access to it.18:39
bnemecdprince: The templates I use in this environment are untagged-external from
dprincebnemec: I see, so maybe the issue is totally specific to single nic w/ vlans18:42
*** paramite is now known as paramite|afk18:43
dprincebnemec: anyways, it doesn't seemed to have helped get gfidente's netiso CI patch passing :/18:45
*** dprince has quit IRC18:49
*** absubram has quit IRC18:50
openstackgerritPradeep Kilambi proposed openstack/tripleo-heat-templates: Deploy Aodh services, replacing Ceilometer Alarm
*** bswartz has joined #tripleo18:55
*** xinwu has joined #tripleo19:05
*** leanderthal is now known as leanderthal|afk19:08
slaglebnemec: in my local test with, auth_url still doesn't have /v2.0 on the end in nova.conf19:13
slagleit's doomed19:13
slaglei wonder if puppet-neutron has the needed change in stable/liberty?19:14
*** dprince has joined #tripleo19:14
bnemecslagle: Hmm, I recalled that the fix there was to _remove_ the suffix and let the keystoneclient code figure out the version.19:15
bnemecI don't think that change touches nova.conf at all.19:15
bnemecMaybe there's a separate fix needed here?19:16
slaglei thought puppet-neutron was supposed to add the /v2.019:16
slagleEmilienM: do you recall?19:16
EmilienMlet me check what we did in liberty19:17
EmilienMliberty is so far :P19:17
EmilienMslagle: by default, we have this
bnemecYeah, FWIW I don't know for sure that what I proposed ever worked.19:17
EmilienMbut I'm sure THT overrides it19:18
bnemecI just backported the change that fixed master.19:18
slagleEmilienM: maybe it's in puppet-nova actually19:18
slaglethis is auth_url in the [neutron] section in nova.conf19:19
EmilienMthat's the default in puppet-nova19:19
EmilienMbut again I'm sure you override it in tripleo19:19
slaglewe do19:19
slaglewith the right IP19:19
EmilienMyou need the /v2.0 I think19:20
EmilienMlet me check our CI manifests19:20
slagleyea, so this is wrong then
slaglewe should not switch to using uri_no_suffic19:20
EmilienMin liberty ^19:20
*** dprince has quit IRC19:21
slaglelet me try locally if just adding the /v2.0 works19:21
*** dprince has joined #tripleo19:21
slaglethat worked19:24
*** paramite|afk is now known as paramite19:25
openstackgerritJames Slagle proposed openstack/tripleo-heat-templates: Set project name for nova -> neutron notifications
slaglebnemec: let's see if that passes ^19:29
*** paramite has quit IRC19:32
*** leanderthal|afk is now known as leanderthal19:44
*** absubram has joined #tripleo19:53
*** cinerama has quit IRC20:00
*** mbound has joined #tripleo20:03
*** leanderthal is now known as leanderthal|afk20:29
*** Goneri has quit IRC20:40
*** absubram has quit IRC20:54
openstackgerritDan Sneddon proposed openstack/tripleo-heat-templates: Make External Load Balancer templates work with IPv6
*** absubram has joined #tripleo21:07
*** jdob has quit IRC21:13
openstackgerritDan Prince proposed openstack/tripleo-image-elements: Correct ping test when using Swift temp_urls
*** tiswanso has quit IRC21:24
*** tiswanso has joined #tripleo21:25
*** absubram has quit IRC21:28
*** mkovacik has joined #tripleo21:34
dprincebnemec, dsneddon:
dprincedsneddon: context is IPv4 network isolation has been broken on master with some of the config templates21:34
dprincedsneddon: on a different note did you get a chance to create the etherpad w/ the required IPv6 patches?21:35
dsneddondprince, Yeah, I added it to the existing Etherpad shortly after we spoke about it. Let me grab the URL.21:36
*** xinwu has quit IRC21:36
openstackgerritDan Prince proposed openstack-infra/tripleo-ci: Use netiso in the ha job
dprincedsneddon: thanks21:37
bnemecdprince: So this is going to be another upgrade problem, right?  Because the orc scripts are installed into the image at build time and not packaged.21:37
dprincebnemec: I would upgrade the ORC scripts with this:
*** xinwu has joined #tripleo21:37
dprincebnemec: tarball extracts the new ORC scripts into the existing images, boom21:38
dprincebnemec: works for puppet-modules, ORC scripts, anything. RPMs too21:38
dprincebnemec: anyways, the ping_metadata_ip check is newish. Users of really old overclouds won't have it anyways21:39
dprincethe metadata_ip ping tests is a good all around improvement though21:39
bnemecI bet it got backported to Kilo though.21:40
dprincebnemec: so long as the swift temp URL signal default in t-h-t didn't we are okay there21:40
dprincebnemec: upgrades, scale outs, etc. with the new upstream stuff could all fail without fixing the os-net-config script though21:41
bnemecdprince: Yeah, we just need to make sure we don't backport the signal change to a version where we don't have the artifact update functionality.21:41
bnemecAnd get the artifact update merged, of course. :-)21:41
dprincebnemec: I'm happy for it to merge as is21:42
*** rlandy has quit IRC21:42
bnemecdprince: Yeah, I'll take a closer look.21:42
dprincebnemec: anyways, fingers crossed gfidente's netiso CI patch passes this time21:43
dprincebnemec: then I try moving things to IPv6...21:43
bnemecFun stuff. :-)21:44
slaglei think liberty is broken21:45
slaglenothing has passed in 8 hours21:45
bnemecmv stable/liberty "stable"/liberty21:45
*** slagle changes topic to "TripleO | CI status: | Docs:"21:46
openstackgerritOpenStack Proposal Bot proposed openstack/tripleo-common: Updated from global requirements
*** cinerama has joined #tripleo21:48
openstackgerritMerged openstack/instack-undercloud: Make ceilometer hardware meters work out of the box
bnemecdprince: I think there may be a bug in the rpm install code.  Let me know if I'm missing something.21:51
dprincebnemec: I think you are right. will fix it. Thanks21:53
bnemecdprince: Cool.  It could probably use a more recent CI pass anyway since the last one was pre-pingtest.21:54
bnemecGiven that it already had a +2 though, I'd be okay with single approving it for such a small change.21:54
openstackgerritDan Prince proposed openstack/tripleo-heat-templates: Add support for DeployArtifactURLs
dprincebnemec: ^^, thanks21:56
*** spredzy has quit IRC21:56
*** spredzy has joined #tripleo21:57
*** dustins has quit IRC21:59
*** Goneri has joined #tripleo22:08
*** mbound has quit IRC22:11
*** rhallisey has quit IRC22:14
*** Marga_ has quit IRC22:16
*** Marga_ has joined #tripleo22:16
*** lblanchard has quit IRC22:20
*** dshulyak has quit IRC22:24
*** trown is now known as trown|outtypewww22:25
*** tiswanso has quit IRC22:26
*** yamahata has quit IRC22:37
*** marcusvrn_ has quit IRC22:47
*** Marga_ has quit IRC22:48
*** Marga_ has joined #tripleo22:49
openstackgerritMerged openstack-infra/tripleo-ci: Set overcloud deploy timeout
*** Marga_ has quit IRC22:53
*** jtomasek has quit IRC22:54
*** pradk has quit IRC22:56
*** rpothier has quit IRC23:06
*** mbound has joined #tripleo23:12
*** mkovacik has quit IRC23:13
*** yamahata has joined #tripleo23:13
*** Marga_ has joined #tripleo23:15
*** mbound has quit IRC23:17
*** Marga_ has quit IRC23:19
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Enable undercloud ssl on nonha job
openstackgerritMerged openstack/tripleo-heat-templates: Emits a different hostname for each network the node is on
*** jayg is now known as jayg|g0n323:33
openstackgerritMerged openstack/tripleo-heat-templates: Pin docker version for atomic at 1.8.2
openstackgerritMerged openstack/tripleo-heat-templates: Add NovaVNCProxy back into endpoint_map
*** panda has quit IRC23:37
*** panda has joined #tripleo23:38
openstackgerritBen Nemec proposed openstack-infra/tripleo-ci: Add undercloud idempotency test to periodic job
*** nico_auvv has quit IRC23:54
*** Marga_ has joined #tripleo23:55
*** ibravo has joined #tripleo23:58

Generated by 2.14.0 by Marius Gedminas - find it at!