*** chlong has joined #tripleo | 00:23 | |
*** ayoung has quit IRC | 00:24 | |
*** kjw3 has joined #tripleo | 00:26 | |
*** limao has joined #tripleo | 00:35 | |
*** artom has quit IRC | 00:39 | |
*** artom has joined #tripleo | 00:40 | |
openstackgerrit | Steve Baker proposed openstack/tripleo-heat-templates: Containerized Services for Composable Roles https://review.openstack.org/330659 | 00:45 |
---|---|---|
openstackgerrit | Steve Baker proposed openstack/tripleo-heat-templates: WIP use overcloud-full instead of atomic-image https://review.openstack.org/395880 | 00:45 |
openstackgerrit | Steve Baker proposed openstack/tripleo-common: Create new docker command hook https://review.openstack.org/312723 | 00:46 |
*** jerrygb has quit IRC | 00:48 | |
*** jerrygb has joined #tripleo | 00:49 | |
*** jerrygb has quit IRC | 00:49 | |
*** jerrygb has joined #tripleo | 00:49 | |
*** saneax is now known as saneax-_-|AFK | 00:53 | |
*** afazekas_ has quit IRC | 01:08 | |
*** kjw3 has quit IRC | 01:10 | |
*** rajinir has quit IRC | 01:16 | |
*** kjw3 has joined #tripleo | 01:21 | |
*** kjw3 has quit IRC | 01:34 | |
*** yamahata has quit IRC | 01:40 | |
*** kjw3 has joined #tripleo | 01:49 | |
ggillies | fultonj: I was able to get my composable roles working fine | 01:51 |
ggillies | everything working | 01:51 |
*** tzumainn has quit IRC | 02:03 | |
*** tiswanso has joined #tripleo | 02:08 | |
*** gbarros has joined #tripleo | 02:08 | |
gbarros | anyone know what the hiera value for InternalApiNetworkVip is? | 02:08 |
gbarros | dsneddon ^ | 02:08 |
*** fzdarsky|afk has quit IRC | 02:10 | |
ggillies | gbarros: I've done hacks in the past using something like | 02:15 |
ggillies | %{hiera('heat::api::bind_host')} | 02:15 |
ggillies | which will resolve to internal api vip | 02:15 |
gbarros | ty ggillies, will give t a try | 02:16 |
fultonj | ggillies: thanks for the update regarding your composable role. was there anything you had to do that was unexpected? Or was it simply a matter of OS::TripleO::DellCompute::Net::SoftwareConfig: nic-configs/compute.yaml ? | 02:17 |
ggillies | fultonj: with network isolation enabled there was a lot I had to add | 02:17 |
fultonj | ggillies: normally for network isolation I use something like this https://github.com/RHsyseng/hci/blob/master/custom-templates/network.yaml | 02:19 |
ggillies | I have http://paste.openstack.org/show/588643/ | 02:19 |
ggillies | because I am using static ip assignment | 02:19 |
ggillies | you will need all the ports definitions | 02:20 |
ggillies | like I have | 02:20 |
ggillies | except they need to point to the ones without *_from_pool.yaml | 02:20 |
ggillies | if you don't specify port definitions they will just default to ctlplane | 02:20 |
ggillies | and you will get no network isolation | 02:20 |
ggillies | it's insanely fiddly | 02:21 |
fultonj | ggillies: that's the trick, thanks! | 02:21 |
ggillies | I do also recommend you switch to predictable ip assignment if you can | 02:21 |
ggillies | we use it for all our environments production or otherwise | 02:21 |
ggillies | same with predicatable host assignment | 02:21 |
*** artom has quit IRC | 02:22 | |
*** artom has joined #tripleo | 02:23 | |
*** cdearborn has quit IRC | 02:24 | |
*** mrunge has quit IRC | 02:24 | |
*** mrunge has joined #tripleo | 02:24 | |
*** dmacpher-afk is now known as dmacpher | 02:25 | |
*** cwolferh has quit IRC | 02:25 | |
*** fabbione has quit IRC | 02:29 | |
*** fabbione has joined #tripleo | 02:29 | |
fultonj | thanks ggillies. | 02:33 |
*** fultonj has quit IRC | 02:33 | |
*** Nagendra has quit IRC | 02:34 | |
*** thrash is now known as thrash|g0ne | 02:51 | |
*** kjw3 has quit IRC | 02:52 | |
*** kjw3 has joined #tripleo | 02:54 | |
*** lblanchard has quit IRC | 03:19 | |
*** tiswanso has quit IRC | 03:25 | |
openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Cleanup yumdownloader repos https://review.openstack.org/395921 | 03:26 |
openstackgerrit | RedHat RDO CI proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/365449 | 03:30 |
*** fragatin_ has joined #tripleo | 03:32 | |
*** jerrygb has quit IRC | 03:33 | |
*** fragatina has quit IRC | 03:35 | |
*** fragatina has joined #tripleo | 03:36 | |
*** fragatin_ has quit IRC | 03:37 | |
*** fragatina has quit IRC | 03:40 | |
*** cmyster has quit IRC | 03:47 | |
*** kjw3 has quit IRC | 03:50 | |
*** cmyster has joined #tripleo | 03:50 | |
*** cmyster has quit IRC | 03:50 | |
*** cmyster has joined #tripleo | 03:50 | |
*** kjw3 has joined #tripleo | 03:51 | |
*** skramaja has quit IRC | 04:00 | |
*** afazekas has joined #tripleo | 04:04 | |
*** skramaja has joined #tripleo | 04:07 | |
*** fzdarsky|afk has joined #tripleo | 04:07 | |
*** chlong has quit IRC | 04:13 | |
*** coolsvap has joined #tripleo | 04:18 | |
*** links has joined #tripleo | 04:23 | |
*** numans has joined #tripleo | 04:24 | |
*** cwolferh has joined #tripleo | 04:34 | |
*** jerrygb has joined #tripleo | 04:34 | |
*** jerrygb has quit IRC | 04:39 | |
*** chandankumar has joined #tripleo | 04:42 | |
*** dsariel has joined #tripleo | 04:54 | |
*** bkreitch has quit IRC | 05:11 | |
*** skramaja_ has joined #tripleo | 05:12 | |
*** liverpooler has quit IRC | 05:36 | |
*** jaosorior has joined #tripleo | 05:51 | |
openstackgerrit | Ian Wienand proposed openstack/diskimage-builder: Cleanup yumdownloader repos https://review.openstack.org/395921 | 06:03 |
*** abregman has joined #tripleo | 06:05 | |
*** lmiccini has joined #tripleo | 06:17 | |
*** rcernin has joined #tripleo | 06:18 | |
*** jaosorior has quit IRC | 06:34 | |
*** jaosorior has joined #tripleo | 06:35 | |
*** jerrygb has joined #tripleo | 06:35 | |
*** iranzo has joined #tripleo | 06:40 | |
*** jerrygb has quit IRC | 06:42 | |
*** skramaja has quit IRC | 06:51 | |
*** pcaruana has joined #tripleo | 06:51 | |
*** rcernin has quit IRC | 06:56 | |
*** rcernin has joined #tripleo | 06:56 | |
*** chlong has joined #tripleo | 06:58 | |
*** oshvartz has joined #tripleo | 07:08 | |
*** chlong has quit IRC | 07:09 | |
*** owalsh has quit IRC | 07:13 | |
*** tesseract has joined #tripleo | 07:20 | |
*** tesseract is now known as Guest40216 | 07:20 | |
*** yamahata has joined #tripleo | 07:21 | |
jaosorior | any reviews for this? https://review.openstack.org/#/c/392519/ | 07:22 |
*** rasca has joined #tripleo | 07:25 | |
openstackgerrit | Sagi Shnaidman proposed openstack/instack-undercloud: Remove deprecated network range 192.0.2.0 https://review.openstack.org/395702 | 07:30 |
*** florianf has joined #tripleo | 07:31 | |
*** jtomasek has joined #tripleo | 07:32 | |
*** ealcaniz has joined #tripleo | 07:37 | |
*** rasca has quit IRC | 07:37 | |
*** rasca has joined #tripleo | 07:37 | |
matbu | o/ | 07:38 |
marios | \o | 07:39 |
chem | o_ | 07:40 |
*** ebarrera has joined #tripleo | 07:44 | |
panda|zZ | _o | 07:45 |
*** panda|zZ is now known as panda | 07:45 | |
rasca | _o_ | 07:46 |
*** openstackgerrit has quit IRC | 07:48 | |
*** openstackgerrit has joined #tripleo | 07:48 | |
*** jlinkes has joined #tripleo | 07:53 | |
*** jprovazn has joined #tripleo | 07:55 | |
openstackgerrit | Merged openstack/tripleo-validations: Validation stonith device exists in OpenStack Platform HA cluster https://review.openstack.org/360102 | 07:57 |
*** mcornea has joined #tripleo | 07:58 | |
*** bvandenh has joined #tripleo | 08:00 | |
*** rasca has quit IRC | 08:00 | |
openstackgerrit | Dougal Matthews proposed openstack/python-tripleoclient: Reset the parameter_defaults between deployments via the CLI https://review.openstack.org/394419 | 08:00 |
*** ohamada has joined #tripleo | 08:02 | |
*** rasca has joined #tripleo | 08:02 | |
*** ccamacho has joined #tripleo | 08:04 | |
*** hewbrocca_afk is now known as hewbrocca | 08:05 | |
openstackgerrit | Merged openstack/puppet-tripleo: Better way to ensure keepalived before haproxy. https://review.openstack.org/395633 | 08:08 |
hewbrocca | Morning all | 08:09 |
d0ugal | Morning! | 08:12 |
shardy | g'morning all | 08:13 |
*** zoli|gone is now known as zoli|wfh | 08:14 | |
*** jpena|off is now known as jpena | 08:21 | |
*** b00tcat has quit IRC | 08:25 | |
*** amoralej|off is now known as amoralej | 08:26 | |
hewbrocca | Have we got ceilometer to stop TRUMPing our CI yet? | 08:26 |
*** abehl has joined #tripleo | 08:27 | |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Introduce branch-based scenario tests https://review.openstack.org/396008 | 08:27 |
*** cylopez has joined #tripleo | 08:28 | |
*** abregman has quit IRC | 08:29 | |
jaosorior | shardy: could you give me your opinion on this? ^^ | 08:29 |
*** abregman has joined #tripleo | 08:32 | |
ccamacho | morning! | 08:36 |
jaosorior | ccamacho: hey dude, how's it going? | 08:36 |
ccamacho | jaosorior hey man! yesterday was a Holiday in Madrid :) I was going to review your patch in the morning but already landed | 08:38 |
jaosorior | no biggie :) | 08:38 |
jaosorior | ccamacho: this one is still missing reviews :D https://review.openstack.org/#/c/392473/ and it's pretty simple | 08:38 |
*** jerrygb has joined #tripleo | 08:38 | |
*** arxcruz|brb is now known as arxcruz | 08:40 | |
panda | hewbrocca: I don't think that any of the services analyzed on the last session is seriously affecting performance, unless pradk or jd__ confirm that the notify in ceilometermiddleware was causing bottlenecks. The only big thing that remains unsolved (that is until periodic jobs from this morning finish) is the memory allocation error from Cinder. It's still, happening, not very often, but still happening. | 08:40 |
panda | I'd like to ask derek what is the memory overcommitment for rh1, when he come online, as it seems happening when the test env is under high load. | 08:40 |
dsariel | hi is it a way to validate tripleo heat templates before deployment that is very lengthy? | 08:41 |
*** jpich has joined #tripleo | 08:41 | |
panda | dsariel: you mean this ? https://github.com/openstack/tripleo-validations | 08:43 |
*** apetrich has quit IRC | 08:43 | |
hewbrocca | panda: that's great. So we have mostly fixed the gnocchi vs. redis death match problem? | 08:43 |
*** apetrich has joined #tripleo | 08:44 | |
dsariel | panda, thanks for the reply, I've seen this and have seen also https://github.com/openstack/tripleo-heat-templates/tree/master/validation-scripts | 08:44 |
*** jerrygb has quit IRC | 08:44 | |
dsariel | panda, so the question is which is better? | 08:44 |
panda | shadower: ^ what you think ? :) | 08:44 |
dsariel | panda, maybe I will elaborate what exactly I need | 08:45 |
*** aufi has joined #tripleo | 08:45 | |
dsariel | panda, I have a yaml that overrides default parameters like cinder backups to be ceph instead of swift. I'm specifying to overcloud depoy this yaml file with -e flag. And before actual deployment I need something that validates the overridden parameters | 08:47 |
panda | hewbrocca: well, gnocchi vs redis, probably. gnocchi vs rest of the world, I don't think so. There was that patch that stopped metricd from retrying frantically for redis, but here for example metricd is retrying frantically because of other problems. This is one of the deployment that fails because of the cinder memory allocation problems. | 08:48 |
panda | http://logs.openstack.org/79/395379/2/check-tripleo/gate-tripleo-ci-centos-7-ovb-ha/5cdde75/logs/overcloud-controller-1/var/log/gnocchi/metricd.txt.gz | 08:48 |
hewbrocca | I see | 08:48 |
hewbrocca | seems like we need to pay some careful attention to gnocchi failure mode | 08:49 |
*** paramite has joined #tripleo | 08:52 | |
*** paramite has quit IRC | 08:52 | |
*** abregman has quit IRC | 08:52 | |
*** paramite has joined #tripleo | 08:52 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: Add firewall rules for manila api service https://review.openstack.org/395769 | 08:53 |
hewbrocca | failure modes | 08:53 |
shardy | dsariel: if you need to validate the heat templates then tripleo-validations won't help - we already do basic validation of the heat syntax when the plan is created (before the deployment starts) | 08:54 |
shardy | that basically uses the heat template-valdate call under the hood | 08:54 |
shardy | jaosorior: sorry, opinion on what? | 08:54 |
panda | mh, that's interesting, I didn't notice it before, cinder allocation error shows on all three controllers. Not at the same time, but sequentially. It's like heat tries with each controller, and they all can't allocate memory. This is another point in favor of memory overcommitment in rh1 causing problems. | 08:55 |
*** nyechiel has joined #tripleo | 08:55 | |
panda | updating bug. | 08:55 |
jaosorior | shardy: this commit https://review.openstack.org/396008 | 08:55 |
shardy | dsariel: if you can elaborate on what failure you're trying to validate I can possibly suggest ways to pre-validate, but some errors can only be caught at runtime (e.g when they depend on some value) | 08:55 |
*** zoli|wfh is now known as zoli|afk | 08:56 | |
shardy | jaosorior: wouldn't it be easier to just put the $newton_and_below services in some subset of the existing scenarios? | 08:57 |
shardy | and add new scenarios which are only run on >= $new_release? | 08:57 |
shardy | I'm not super keen on the duplication implied by copying all-the-tests every release | 08:57 |
*** gfidente has joined #tripleo | 08:58 | |
*** rasca has quit IRC | 08:59 | |
*** rasca has joined #tripleo | 09:00 | |
jaosorior | shardy: but then we'll have this same issue for every release as services are added | 09:02 |
dsariel | shardy, thanks a lot for the help. The script is here; http://paste.openstack.org/show/588669/. I'm trying to run openstack overcloud deploy and if fails because of override.yaml (without override.yaml evrything is ok, but cinder is using swift inctead of ceph backend) | 09:02 |
shardy | jaosorior: sure, but that could still be handled by having a matrix which grows over time, and only the latest releases run scenarion >= n | 09:05 |
shardy | and that would avoid duplicating every single scenario for every release | 09:05 |
jaosorior | shardy: but then newer releases would have to run more and more scenarios, as some integration of some services will require scenarios that contain pretty much the same services. | 09:05 |
shardy | jaosorior: so, you're trying to solve the problem that every release, new services are added | 09:06 |
shardy | but you don't want to run more scenarios on newer releases? | 09:06 |
* shardy is confused | 09:06 | |
*** fabbione has quit IRC | 09:07 | |
shardy | jaosorior: Perhaps there's some other way to make the scenarios more composable | 09:08 |
*** abregman has joined #tripleo | 09:08 | |
shardy | my main complaint is the copy/paste of the templates | 09:08 |
openstackgerrit | Adriano Petrich proposed openstack/tripleo-heat-templates: GATE TEST, please ignore https://review.openstack.org/365449 | 09:08 |
shardy | I had the same complaint when EmilienM started the scenario stuff, and this makes it worse AFAICS | 09:09 |
shardy | let me think about it and see if I can offer a way around it | 09:09 |
jaosorior | shardy: a scenario that only contains barbican wouldn't make any sense. and as features are added in each release, pike will run scenarios for ocata, pike, and below... | 09:09 |
jaosorior | sure | 09:10 |
jaosorior | shardy: now the issue is that we can't add more services cause that breaks the stable branches. And the current proposal is to stop running the offending scenarios entirely. | 09:11 |
panda | dam, I will never get https://review.openstack.org/382515 merged. Each recheck, at least one random job red .. :( | 09:11 |
shardy | jaosorior: Yeah, I'm just saying lets juggle the scenarios so we can say e.g only run scenario 1 & 2 on < newton and 1, 2 3 on > newton | 09:12 |
jaosorior | so I having scenarios is not about having a test check-list; but about testing something that makes sense | 09:12 |
shardy | not that we'd run a scenario per service | 09:12 |
shardy | there's already a bunch of overlap between scenarios | 09:12 |
shardy | Like we could add a scenario 4 | 09:12 |
shardy | and just disable it for stable branches | 09:12 |
shardy | it'd run some subset of the existing stuff plus $new_shiny_for_ocata | 09:13 |
shardy | jaosorior: maybe drop a message to the list, definitely requires some wider discussion I think | 09:13 |
jaosorior | so, by "subset" you just mean "a specific set of scenarios" ? | 09:13 |
shardy | https://github.com/openstack-infra/tripleo-ci#service-testing-matrix | 09:13 |
*** fabbione has joined #tripleo | 09:14 | |
shardy | each scenario already tests some subset of services | 09:14 |
jaosorior | can you rephrase what you said about the subset? I'm honestly getting lost here. | 09:14 |
* jaosorior needs to brew more coffee | 09:14 | |
*** b00tcat has joined #tripleo | 09:15 | |
shardy | jaosorior: I'm not sure how to better explain it, if you give me a list of the services you expect to be new for ocata, I'll push a patch illustrating my suggestion to the test matrix | 09:16 |
shardy | All I'm saying is rather than copying all the scenarios, lets just not run them all on stable branches | 09:17 |
jaosorior | shardy: barbican, mistral & zaqar | 09:18 |
jaosorior | I had initially added barbican in scenario 2, since that scenario already does a volume and attaches it to an instance, so I want to eventually be able to test encrypted volumes. | 09:18 |
jaosorior | I don't know enough about zaqar and mistral to know how those were meant to be tested in the scenarios though | 09:19 |
*** dtantsur|afk is now known as dtantsur | 09:21 | |
*** lucas-afk is now known as lucasagomes | 09:23 | |
*** ohamada has quit IRC | 09:24 | |
*** ohamada has joined #tripleo | 09:24 | |
*** tremble has joined #tripleo | 09:27 | |
*** fabbione has quit IRC | 09:28 | |
*** yamahata has quit IRC | 09:31 | |
*** yolanda has quit IRC | 09:31 | |
*** milan has joined #tripleo | 09:31 | |
*** yolanda has joined #tripleo | 09:31 | |
d0ugal | jaosorior: did you have any luck with your issue yesterday? | 09:32 |
jaosorior | d0ugal: so, I cannot reproduce it in master. Waiting for confirmation if someone else can reproduce it OSP10 | 09:33 |
d0ugal | jaosorior: k, thanks | 09:33 |
jaosorior | mcornea: you were running a job to check if its reproduceable in OSP10 right? the issue with updates from yesterday | 09:34 |
shardy | jaosorior: ack, OK so heat actually supports conditionals now, so we could add coverage to the scenario2 template with those resources conditionally enabled based on the branch name | 09:34 |
shardy | heat supports resources for all those services already | 09:34 |
mcornea | jaosorior: yes, let me see the result | 09:34 |
*** akrivoka has joined #tripleo | 09:34 | |
mcornea | jaosorior: looks good | 09:34 |
jaosorior | shardy: uhm... I like that idea. | 09:34 |
jaosorior | shardy: I was already working on adding a scenario004 if that's the way people wanna go | 09:34 |
shardy | jaosorior: I'll comment on the patch with more details | 09:35 |
shardy | yeah either would work I think, given that the scenarios are conditionally triggered | 09:35 |
jaosorior | mcornea: lets wait to see if the dude that reported the issue still has it with the latest OSP puddle | 09:35 |
jaosorior | mcornea: can you pass a reference to the job definition? | 09:37 |
jaosorior | shardy: what should go first for a scenario, the project-config commit or the tripleo-ci commit? | 09:37 |
*** fabbione has joined #tripleo | 09:38 | |
mcornea | jaosorior: I'll add it to the bz | 09:38 |
shardy | jaosorior: I think the tripleo-ci commit has to come first, so make the project-config patch Depends-On it | 09:44 |
*** milan has quit IRC | 09:45 | |
*** milan has joined #tripleo | 09:46 | |
chem | gfidente: hi, I've got this strange problem where using a blockstorage with ceph backend doesn't seem to work at all | 09:46 |
chem | gfidente: I've documented there https://bugs.launchpad.net/tripleo/+bug/1640730 | 09:46 |
openstack | Launchpad bug 1640730 in tripleo "Installing overcloud with block/object storage node and ceph backend doesn't work." [Undecided,New] | 09:46 |
*** zoli|afk is now known as zoli|wfh | 09:48 | |
*** kaslcrof has joined #tripleo | 09:48 | |
jaosorior | mcornea: right, but can you pass a link to the code that actually executes that job? | 09:48 |
*** social has quit IRC | 09:48 | |
mcornea | jaosorior: I manually trigger it | 09:49 |
jaosorior | shardy: having gotten more coffee and thought about it. Actually having extra scenarios might be the way to go. Would probably be less of a burden to maintain | 09:49 |
*** pcaruana has quit IRC | 09:49 | |
jaosorior | mcornea: I see, and, what about the job definition? | 09:49 |
*** limao has quit IRC | 09:53 | |
jaosorior | mcornea: wow, so from what I can see, that job pretty much does the same thing than what the guy that reported the bug is doing | 09:56 |
jaosorior | mcornea: I think it's just a matter of updating the puddle indeed | 09:56 |
*** social has joined #tripleo | 09:57 | |
gfidente | chem, so that is the expected behaviour | 09:58 |
gfidente | blockstorage nodes run the cinder/lvm driver only | 09:58 |
mcornea | jaosorior: yeah, unless there are some environmental things that I miss I think it should work | 09:58 |
gfidente | in newton it shouldn't have been different | 09:59 |
*** katkapilatova has joined #tripleo | 09:59 | |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Introduce scenario004 https://review.openstack.org/396056 | 09:59 |
gfidente | chem, so you should have the cephstorage nodes running ceph-osd and the blockstorage nodes running cinder-volume | 10:01 |
openstackgerrit | Merged openstack/tripleo-heat-templates: Defaults kernel.pid_max to 1048576 https://review.openstack.org/395608 | 10:01 |
chem | gfidente: that's corrrect | 10:01 |
*** pcaruana has joined #tripleo | 10:01 | |
gfidente | chem, and if you do cinder service-list you should see a single tripleo_ceph backend | 10:02 |
gfidente | and multiple tripleo_iscsi agents (one per each blockstorage node) | 10:02 |
*** sudipto has joined #tripleo | 10:04 | |
*** sudipto_ has joined #tripleo | 10:04 | |
*** athomas has joined #tripleo | 10:06 | |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Introduce scenario004 https://review.openstack.org/396056 | 10:09 |
*** jaosorior is now known as jaosorior_lunch | 10:10 | |
lucasagomes | folks, can someone take a look at this: https://review.openstack.org/#/c/370773/ It's a tiny patch (1 LOC) and have been hanging around for a long time. Thanks | 10:10 |
dtantsur | lucasagomes, we should not patch this code any more.. | 10:11 |
dtantsur | well, if you want to fix it for older releases (pre-newton), then yes | 10:11 |
jaosorior_lunch | dtantsur, lucasagomes well, if you wanna backport it it needs an LP bug | 10:12 |
dtantsur | jaosorior_lunch, there no stable branches there | 10:12 |
jaosorior_lunch | no? | 10:12 |
jaosorior_lunch | ah | 10:12 |
jaosorior_lunch | ok | 10:12 |
lucasagomes | dtantsur, this patch is from September, I was mostly tyring to get rid of my list of patches... but still, if not used anymore I will just abandon it | 10:13 |
dtantsur | lucasagomes, mind fixing https://github.com/openstack/tripleo-common/blob/master/tripleo_common/utils/nodes.py#L165 instead please? this is what tripleo really uses. and please create an LP bug, so that we can backport it later (tripleo-common does have stable branches) | 10:13 |
lucasagomes | dtantsur, on it | 10:14 |
openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Select bootstrap node by list index not name https://review.openstack.org/395699 | 10:27 |
dtantsur | shardy, do you by chance have any plans to continue with heat support for ironic? | 10:28 |
openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Select bootstrap node by list index not name https://review.openstack.org/395699 | 10:28 |
shardy | dtantsur: I don't atm but I think ricolin was planning to revive it | 10:28 |
shardy | dtantsur: for the tripleo use-case, I think defining a mistral workflow which deploys the nodes is somewhat more interesting | 10:29 |
shardy | (that workflow could then be triggered via heat) | 10:29 |
shardy | https://review.openstack.org/#/c/313048/ was a first pass at that, but I've not had time to get it fully working | 10:29 |
shardy | dtantsur: did you have a reason for wanting those resources? | 10:34 |
openstackgerrit | Lucas Alvares Gomes proposed openstack/tripleo-common: Replace ucs_hostname with ucs_address https://review.openstack.org/396086 | 10:34 |
lucasagomes | jaosorior_lunch, dtantsur ^ | 10:34 |
dtantsur | shardy, creating a scenario job exercising ironic in overcloud, primarily | 10:39 |
*** jerrygb has joined #tripleo | 10:40 | |
shardy | dtantsur: Ack, Ok we'll have to think how to do that - we could revive the heat resources, but I reached the conclusion that they won't be that useful without the workflow to drive deployment (which is a better fit for mistral than heat I think) | 10:43 |
shardy | One option would be to drive the interaction with Ironic with a OS::Mistral::Workflow resource, but that doesn't run the workflow (it's triggered via an alarm or user signal) | 10:45 |
*** jerrygb has quit IRC | 10:45 | |
openstackgerrit | Chris Jones proposed openstack/puppet-tripleo: Improve failed mysql node removal time in HA deploys. https://review.openstack.org/396092 | 10:45 |
dtantsur | hmm | 10:45 |
dtantsur | unrelated: could you please review one more small ocata-1/netwon bug fix: https://review.openstack.org/#/c/392181/ | 10:46 |
shardy | That looked directed at me, but I've already +2'd it so other folks please review :) | 10:47 |
*** abregman is now known as abregman|afk | 10:47 | |
dtantsur | yeah, sorry, it was a general request not related to the discussion | 10:48 |
dtantsur | just trying to clean up my bug list for newton downstream | 10:48 |
openstackgerrit | Anshul Jain proposed openstack/diskimage-builder: DIB element to support cinder local attach/detach functionality https://review.openstack.org/385880 | 10:50 |
*** tosky has joined #tripleo | 10:50 | |
*** chandankumar has quit IRC | 10:52 | |
*** chandankumar has joined #tripleo | 10:58 | |
*** gfidente has quit IRC | 10:58 | |
*** owalsh has joined #tripleo | 11:03 | |
openstackgerrit | Chris Jones proposed openstack/puppet-tripleo: Improve failed mysql node removal time in HA deploys. https://review.openstack.org/396092 | 11:03 |
*** mgould|afk is now known as mgould | 11:11 | |
*** gfidente has joined #tripleo | 11:17 | |
*** hjensas has quit IRC | 11:20 | |
*** jerrygb has joined #tripleo | 11:25 | |
*** hjensas has joined #tripleo | 11:30 | |
*** abregman|afk is now known as abregman | 11:34 | |
*** thrash|g0ne is now known as thrash | 11:36 | |
*** jkilpatr has quit IRC | 11:37 | |
*** sudipto has quit IRC | 11:38 | |
*** sudipto_ has quit IRC | 11:38 | |
shardy | thrash: Hey, good morning - am I right in thinking that you've been doing some work on the overcloud node delete command? | 11:40 |
thrash | shardy: sort of... rbrady-afk did that one. | 11:42 |
shardy | thrash: ah, OK no worries | 11:42 |
thrash | shardy: wait... | 11:42 |
thrash | shardy: he used some ideas of mine from stack update. | 11:42 |
shardy | I raised https://bugs.launchpad.net/tripleo/+bug/1640770 with a few usability issues I noticed, was looking for a volunteer to fix them ;) | 11:42 |
openstack | Launchpad bug 1640770 in tripleo "openstack node delete usability issues" [Medium,Triaged] | 11:42 |
thrash | shardy: sure. I'll take a look. | 11:42 |
shardy | not super urgent, I can probably get to it myself in due course if you're busy with other stuff | 11:43 |
*** chlong has joined #tripleo | 11:43 | |
thrash | shardy: not at all. It's on a list of things to improve the cli this cycle. | 11:44 |
*** chlong has quit IRC | 11:44 | |
*** cdearborn has joined #tripleo | 11:47 | |
openstackgerrit | Katerina Pilatova proposed openstack/tripleo-validations: callback_plugins/validation_output.py: improved output https://review.openstack.org/396135 | 11:48 |
d0ugal | thrash: I have done something similar to the usability issues | 11:55 |
d0ugal | thrash: https://review.openstack.org/#/c/395522/ | 11:55 |
*** bkreitch has joined #tripleo | 11:55 | |
d0ugal | thrash: I have only taken into account the first part | 11:56 |
thrash | d0ugal: that goes a long way to solve shardy's main complaint | 11:56 |
d0ugal | and mine too :) | 11:56 |
d0ugal | thrash: I can update it to default to 'overcloud' | 11:56 |
thrash | d0ugal: +1 | 11:57 |
thrash | relocating to RHT... | 11:57 |
*** thrash is now known as thrash|biab | 11:57 | |
d0ugal | thrash|biab: k, i'll do that, then you just need to do the 2nd part | 11:57 |
*** jkilpatr has joined #tripleo | 11:58 | |
marios | reviews please https://review.openstack.org/#/c/395053/ | 11:59 |
*** paramite has quit IRC | 12:03 | |
*** jpena is now known as jpena|lunch | 12:04 | |
openstackgerrit | Merged openstack/tripleo-specs: Spec: undercloud-ntp-server - Add NTP services to Undercloud https://review.openstack.org/350729 | 12:06 |
*** paramite has joined #tripleo | 12:07 | |
openstackgerrit | Merged openstack/tripleo-specs: TripleO Repo Management Tool https://review.openstack.org/389237 | 12:08 |
*** fragatina has joined #tripleo | 12:11 | |
openstackgerrit | Dougal Matthews proposed openstack/python-tripleoclient: Default the stack name to 'overcloud https://review.openstack.org/395522 | 12:11 |
*** fragatina has quit IRC | 12:11 | |
*** fragatina has joined #tripleo | 12:12 | |
marios | /win 29 | 12:13 |
d0ugal | so much win | 12:13 |
akrivoka | lol | 12:14 |
*** cylopez has quit IRC | 12:19 | |
*** dprince has joined #tripleo | 12:19 | |
*** cylopez has joined #tripleo | 12:20 | |
openstackgerrit | Dougal Matthews proposed openstack/python-tripleoclient: Default the stack name to 'overcloud' in overcloud node delete https://review.openstack.org/395522 | 12:20 |
openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Normalize civetweb binding address if IPv6 https://review.openstack.org/395785 | 12:21 |
*** arxcruz is now known as arxcruz|brb | 12:23 | |
*** masco has quit IRC | 12:24 | |
*** bkopilov has quit IRC | 12:31 | |
*** dmacpher has quit IRC | 12:33 | |
*** dmacpher has joined #tripleo | 12:33 | |
*** jaosorior_lunch is now known as jaosorior | 12:34 | |
gfidente | marios good call on the chcon | 12:34 |
gfidente | I think we need to find an agreement | 12:35 |
*** thrash|biab is now known as thrash | 12:35 | |
*** pkovar has joined #tripleo | 12:36 | |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Deployed server: switch to apply-config hook https://review.openstack.org/396171 | 12:39 |
jaosorior | is there anybody with some extra cycles to help me out debugging a heat deployment over httpd? | 12:39 |
*** chandankumar has quit IRC | 12:41 | |
*** lucasagomes is now known as lucas-hungry | 12:41 | |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Handle null role_data in services https://review.openstack.org/396175 | 12:43 |
thrash | d0ugal: can you tag your patch with Partial-Bug (I think that's the right one) | 12:45 |
therve | jaosorior, What's up? | 12:48 |
*** trown|outtypewww is now known as trown | 12:48 | |
*** maticue has joined #tripleo | 12:50 | |
d0ugal | therve: I thought I did. Checking. | 12:50 |
d0ugal | therve: I almost did | 12:50 |
therve | thrash, ^^ | 12:50 |
therve | :) | 12:50 |
jaosorior | therve: hey, so, I'm trying to deploy both heat and heat-cfn over httpd in TripleO's undercloud. And I ended up writing the necessary stuff in puppet-heat (which pretty much mimics how other services do it) https://review.openstack.org/#/c/394835/ ; however, when doing overcloud deploy, the deployment times out... In this attempt I'm only deploy heat-api over httpd https://review.openstack.org/#/c/394837/3 | 12:50 |
thrash | therve: haha... | 12:51 |
d0ugal | therve: sorry | 12:51 |
openstackgerrit | Dougal Matthews proposed openstack/python-tripleoclient: Default the stack name to 'overcloud' in overcloud node delete https://review.openstack.org/395522 | 12:51 |
*** panda has quit IRC | 12:51 | |
therve | d0ugal, No worries | 12:51 |
d0ugal | thrash: I had used "Parial-Bug" lol | 12:51 |
jaosorior | therve: this attempt is failing on the multinode job only; when I was trying to deploy heat-api and heat-cfn also, the ovb jobs were failing as well. | 12:51 |
thrash | d0ugal: haha | 12:51 |
openstackgerrit | Honza Pokorny proposed openstack/tripleo-specs: Add gui logging spec https://review.openstack.org/395138 | 12:52 |
jaosorior | therve: now, I tried it in a local deployment, and it both the heat endpoint and the heat-cfn endpoint seem to work :/ .... so I'm not entirely sure what part is messed up | 12:52 |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Neutron L3 service cleanups for hiera json hook https://review.openstack.org/393262 | 12:54 |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Horizon service cleanups for hiera json hook https://review.openstack.org/393258 | 12:54 |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Hiera optimization: use a new hiera hook https://review.openstack.org/384757 | 12:54 |
therve | jaosorior, Have you managed to extract anything from the failures? | 12:54 |
jaosorior | that's the issue, no logs indicating failure :/ | 12:55 |
jaosorior | I'm doing another local install now to reproduce it | 12:56 |
therve | I just see 'CREATE aborted' on the controller | 12:56 |
*** kjw3 has quit IRC | 12:56 | |
jaosorior | therve: yep, it times out | 12:56 |
*** panda has joined #tripleo | 12:57 | |
dprince | shardy: this seemed to work for me with my role_data issue https://review.openstack.org/#/c/396175/ | 12:57 |
jaosorior | haven't figured out why :/ | 12:57 |
therve | Oh | 12:57 |
therve | jaosorior, I was confusing hours and minutes | 12:58 |
therve | It's actually sitting there 1h20 doing nothing | 12:58 |
jaosorior | any idea what's it waiting for? | 12:58 |
jaosorior | panda:regarding https://review.openstack.org/#/c/396008/ I mention Ocata in the commit message | 12:59 |
jaosorior | panda: anyway, in that scheme, master would be tested without a prefix | 13:00 |
*** ealcaniz has quit IRC | 13:00 | |
jaosorior | panda: so that's one proposition, the other one https://review.openstack.org/#/c/396056/ is to have more scenarios | 13:00 |
jaosorior | panda: which at the moment I'm starting to favor more | 13:00 |
*** zoli|wfh is now known as zoli|lunch | 13:00 | |
*** ealcaniz has joined #tripleo | 13:01 | |
jaosorior | panda: so in this case, scenario004 wouldn't run in newton and below, this way, we can safely add new servicse. | 13:01 |
jaosorior | *services | 13:01 |
*** panda has quit IRC | 13:02 | |
*** jpena|lunch is now known as jpena | 13:03 | |
*** jayg|g0n3 is now known as jayg | 13:03 | |
*** egafford has joined #tripleo | 13:04 | |
openstackgerrit | Tom Barron proposed openstack/tripleo-heat-templates: Add firewall rules for manila api service https://review.openstack.org/396183 | 13:04 |
*** panda has joined #tripleo | 13:06 | |
openstackgerrit | Jiri Tomasek proposed openstack/tripleo-specs: GUI Deployment Configuration improvements https://review.openstack.org/393365 | 13:08 |
jtomasek | florianf: ^ I hope it makes more sense now | 13:08 |
therve | jaosorior, Oh, that changed the *undercloud* heat? | 13:09 |
therve | I misunderstood | 13:09 |
therve | OK so yeah os collect config fails to talk to heat | 13:10 |
jaosorior | therve: would that be heat-api or heat-cfn? | 13:10 |
therve | jaosorior, api | 13:10 |
florianf | jtomasek: thanks, lgtm | 13:10 |
florianf | jtomasek: thanks for the changes. | 13:10 |
jaosorior | ok | 13:11 |
jaosorior | therve: and what would that imply? we can see that we can begin the stack creation through heat-api. And I'm not changing any ports for heat either :/ | 13:11 |
*** bfournie has quit IRC | 13:11 | |
*** Vijayendra has quit IRC | 13:11 | |
therve | jaosorior, I'm not exactly sure yet, but it makes more sense :D | 13:12 |
*** panda is now known as panda|afk | 13:12 | |
therve | jaosorior, See http://logs.openstack.org/37/394837/3/check/gate-tripleo-ci-centos-7-nonha-multinode/6479036/logs/var/log/httpd/heat_api_wsgi_access.txt.gz we're getting 404 on metadata calls, that's the issue | 13:12 |
jaosorior | therve: I can discard selinux being an issue cause I'm seeting I'm seeing the same issue in a local deployment running on permissive mode | 13:13 |
*** liverpooler has joined #tripleo | 13:13 | |
*** coolsvap has quit IRC | 13:13 | |
*** rbrady-afk is now known as rbrady | 13:13 | |
jaosorior | therve: so... trying to access heat-cfn (which is the metadata service...right?) is returning 404? | 13:15 |
*** arxcruz|brb is now known as arxcruz | 13:15 | |
therve | jaosorior, No that's heat-api | 13:15 |
jaosorior | Oook | 13:15 |
therve | jaosorior, I believe the issue is with the %2F in there, I don't know what it is doing | 13:16 |
shardy | therve, jaosorior: we use the request collector by default, so we poll swift not heat-api | 13:16 |
shardy | we do still use CFN signaling tho | 13:17 |
jaosorior | therve: which %2F? | 13:17 |
jaosorior | I see | 13:17 |
therve | jaosorior, "GET /v1/e17833ec2d13432e9ccb74483530aeaf/stacks/overcloud-Controller-nlbcp2mngxpo-0-zpjcit3idr4y-Controller-3lfxsulsbqt3-deployed-server-h72riaxxmln4%2F0bc4f9f6-ba41-4d4c-ad47-0c708049c7f0/resources/deployed-server-config/metadata HTTP/1.1" 404 | 13:17 |
therve | The %2F in there | 13:17 |
Ng | jaosorior: would you care to look at the newton backport of the haproxy/mysql patch you +2'd? It's identical :) https://review.openstack.org/#/c/396092/2 | 13:17 |
therve | shardy, What's talking to metadata then? | 13:17 |
Ng | jaosorior: much appreciated, thank you | 13:18 |
shardy | therve: Oh, I guess this is the multinode job? | 13:18 |
therve | shardy, Yes | 13:19 |
shardy | ./overcloud-resource-registry-puppet.j2.yaml: SoftwareConfigTransport: POLL_TEMP_URL | 13:20 |
therve | Oh I think I know | 13:20 |
shardy | I don't see us set that in tripleo-ci | 13:20 |
*** kjw3 has joined #tripleo | 13:21 | |
therve | jaosorior, Where's the puppet code that configure httpd? Or the resulting config? | 13:21 |
akrivoka | rbrady: jtomasek: ping, I'd like to discuss the import and export of plans, when you have a minute | 13:21 |
jtomasek | akrivoka: yep, ready | 13:22 |
jaosorior | therve: /etc/httpd/conf.d/10-heat_api_wsgi.conf | 13:23 |
shardy | Can I get another review for https://review.openstack.org/#/c/395699/ please? | 13:23 |
therve | jaosorior, I meant in the gate logs :) | 13:23 |
therve | But I think I found it | 13:23 |
shardy | fixes a problem with node replacement that I want to backport to newton asap | 13:23 |
akrivoka | rbrady: jtomasek: basically I'd like to write blueprints for plan and environment import and export from gui and need to understand what exactly we need on the front and back end | 13:24 |
therve | jaosorior, It's missing AllowEncodedSlashes | 13:24 |
jaosorior | therve: duuuude yes! | 13:24 |
shardy | therve: nice work spotting that :) | 13:24 |
therve | shardy, Nothing liking finding a %2F in a 240 chars URL :) | 13:25 |
shardy | hehe :) | 13:25 |
jaosorior | therve: thanks a lot dude! I've been struggling with this | 13:25 |
therve | jaosorior, No problem! | 13:26 |
jtomasek | akrivoka: so for import I think all we need it a way to identify the mistral environment. as that would most probably be exported as json, then for import I think a convention of calling the file mistral-environment.json is enough. This is the same approach we use to identify capabilities-map.yaml | 13:26 |
*** lucas-hungry is now known as lucasagomes | 13:27 | |
jtomasek | akrivoka: so when creating the plan, tripleo-common would find that the swift container includes that file and updates mistral environment with it's content | 13:27 |
jtomasek | akrivoka: (same for updating the plan) | 13:27 |
*** udesale_ has joined #tripleo | 13:28 | |
d0ugal | akrivoka: I have been looking into this from the mistral workflow side. | 13:28 |
d0ugal | jtomasek: ^ | 13:29 |
jtomasek | akrivoka: the problem I see here is that tripleo-common would have to take care of removing that file from swift once it updates mistral environment with its contents | 13:29 |
shardy | Can I also get reviews on https://review.openstack.org/#/q/Id240709e256b057bbdff038aefc7fcacb0d339b8,n,z please | 13:29 |
shardy | one failed the ha ovb job but that file isn't tested in CI anyway | 13:29 |
*** amoralej is now known as amoralej|lunch | 13:30 | |
jtomasek | akrivoka: exporting the plan is a matter of downloading all the files from swift container in tar + include mistral environment contents as mistral-environment.json | 13:30 |
d0ugal | jtomasek, akrivoka - it is, but a workflow should provide that | 13:30 |
jtomasek | d0ugal: yep, agree | 13:30 |
d0ugal | jtomasek, akrivoka - importing is just "create", we have that. It might need to be extended slightly | 13:31 |
d0ugal | jtomasek, akrivoka - and then an export workflow - I am not sure how it gives you the files yet... | 13:31 |
d0ugal | jtomasek, akrivoka - but I am just walking out the door to get lunch - back in a bit. | 13:31 |
*** redhatkj has joined #tripleo | 13:32 | |
*** fragatina has quit IRC | 13:33 | |
akrivoka | jtomasek: regarding removing the mistral-environement.json after the update - how does are we handing that with capabilities-map.yaml? | 13:34 |
openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Normalize civetweb binding address if IPv6 https://review.openstack.org/395785 | 13:34 |
bandini | chem: did you manage to conclude your tls/nontls testing on https://review.openstack.org/#/c/395454/ ? if yes, can you throw in the results in the review? | 13:34 |
jtomasek | akrivoka: regarding GUI for import no changes should be needed and for export, we only need to add button to plans list which triggers an action to run the export workflow | 13:34 |
*** fultonj has joined #tripleo | 13:34 | |
jtomasek | akrivoka: capabilities-map.yaml is different in this regards, it is not being changed | 13:35 |
chem | bandini: didn't I ... sorry, doing it. all is green and good | 13:35 |
*** jerrygb has quit IRC | 13:35 | |
bandini | chem: awesome, thanks ;) | 13:35 |
*** jerrygb has joined #tripleo | 13:36 | |
jtomasek | akrivoka: mistral-environment.json is just a tool which transports its content to actual mistral environment. then it is not needed | 13:36 |
*** rhallisey has joined #tripleo | 13:36 | |
*** kjw3 has quit IRC | 13:36 | |
jtomasek | akrivoka: if we left it in the swift container and user would update the plan with additional plan, tripleo-common would overide the mistral environment with that file again, which is not what we want. this is why we need to remove it after it is read | 13:37 |
akrivoka | jtomasek: is there any harm in just leaving it in the swift container? | 13:37 |
akrivoka | jtomasek: ah I see :) | 13:37 |
jtomasek | akrivoka: or we would have to keep it in sync with mistral environment which would be more problematic and totally not necessary | 13:37 |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Deploy heat APIs over httpd https://review.openstack.org/394837 | 13:38 |
*** morazi has joined #tripleo | 13:38 | |
akrivoka | jtomasek: so there simply needs to be a step (action?) in the workflow to remove this file from the container | 13:38 |
chem | bandini: but on the tls setup, but I had an unfortunate vpn disconnection during convergence. Replying to the review anyway | 13:38 |
jtomasek | akrivoka: yes | 13:38 |
*** jerrygb has quit IRC | 13:40 | |
jtomasek | akrivoka: pdate the plan -> is there mistral-environment.json? -> I read it and store it's content in mistral environment -> remove mistral-environment.json from swift container | 13:40 |
akrivoka | jtomasek: got it | 13:41 |
*** jprovazn has quit IRC | 13:42 | |
bandini | chem: oh I see. I guess if it was during convergence we *should* be good as I cannot possibly think how our review affects anything but the major-upgrade step | 13:42 |
*** rhallisey has quit IRC | 13:43 | |
*** oshvartz has quit IRC | 13:43 | |
akrivoka | jtomasek: d0ugal: ok, I'll try to formalize this in a spec, I think one spec for both import and export would be ok, then we can create blueprints and bugs off of that as needed. sounds good? | 13:43 |
chem | bandini: ack, updated the review accordingly | 13:43 |
openstackgerrit | Tomas Sedovic proposed openstack/tripleo-docs: Add tripleo-validations user documentation https://review.openstack.org/396197 | 13:43 |
openstackgerrit | Merged openstack/tripleo-docs: Link to RDO built images https://review.openstack.org/394602 | 13:44 |
bandini | chem: thank you sir ;) | 13:44 |
chem | bandini: hehe, you're welcome. | 13:45 |
openstackgerrit | Merged openstack/tripleo-docs: Remove the replace controller docs https://review.openstack.org/381469 | 13:46 |
*** jerrygb has joined #tripleo | 13:47 | |
*** bfournie has joined #tripleo | 13:48 | |
*** bogdando has quit IRC | 13:50 | |
*** rlandy has quit IRC | 13:52 | |
openstackgerrit | Dan Prince proposed openstack/tripleo-puppet-elements: Move nova.conf truncation into separate element https://review.openstack.org/392559 | 13:53 |
*** rhallisey has joined #tripleo | 13:56 | |
*** cmyster has quit IRC | 13:56 | |
*** jcoufal has joined #tripleo | 13:57 | |
*** cmyster has joined #tripleo | 13:57 | |
*** cmyster has quit IRC | 13:57 | |
*** cmyster has joined #tripleo | 13:57 | |
openstackgerrit | Joe Talerico proposed openstack/tripleo-heat-templates: Increase reserved memory for computes when enabling DVR https://review.openstack.org/382594 | 13:57 |
*** rlandy has joined #tripleo | 13:57 | |
jaosorior | sshnaidm: for reference, this is the bug about the network range https://bugs.launchpad.net/tripleo/+bug/1553222 | 13:58 |
openstack | Launchpad bug 1553222 in tripleo "Default undercloud control plane network violates rfc5737" [Undecided,In progress] - Assigned to Juan Antonio Osorio Robles (juan-osorio-robles) | 13:58 |
openstackgerrit | Katerina Pilatova proposed openstack/tripleo-validations: undercloud-tokenflush: changed group to pre-introspection https://review.openstack.org/396212 | 13:59 |
*** athomas has quit IRC | 13:59 | |
openstackgerrit | Steven Hardy proposed openstack/tripleo-docs: Add note with boot config image options https://review.openstack.org/336078 | 13:59 |
sshnaidm | jaosorior, ok | 13:59 |
openstackgerrit | Merged openstack/tripleo-docs: Fix OpenStack client invocation https://review.openstack.org/382353 | 14:03 |
*** Goneri has joined #tripleo | 14:04 | |
*** athomas has joined #tripleo | 14:04 | |
afazekas | Is a current tripleo setup able to boot up a vm in an ipv6 setup ? | 14:04 |
*** amoralej|lunch is now known as amoralej | 14:05 | |
sshnaidm | jaosorior, trying to find why instack-undercloud tests fail.. | 14:06 |
*** tzumainn has joined #tripleo | 14:07 | |
*** jprovazn has joined #tripleo | 14:07 | |
jaosorior | sshnaidm: what was the patch? | 14:09 |
sshnaidm | jaosorior, https://review.openstack.org/#/c/395702/ | 14:09 |
jaosorior | sshnaidm: alright, trying to reproduce | 14:11 |
jaosorior | shardy: here are the patches for the scenario004 that will only run on on newer releases than newton https://review.openstack.org/#/c/396062/ https://review.openstack.org/#/c/396056/ | 14:12 |
openstackgerrit | Dan Prince proposed openstack/os-collect-config: Add support for Zaqar websockets https://review.openstack.org/342486 | 14:13 |
sshnaidm | jaosorior, how do you run them? I fail to run tests with tox -v -epy27 | 14:14 |
*** cdearborn has quit IRC | 14:14 | |
jaosorior | sshnaidm: have you tried re-creating the tox environment? | 14:14 |
jaosorior | sshnaidm: tox -r -e py27 | 14:14 |
*** lblanchard has joined #tripleo | 14:14 | |
*** panda|afk is now known as panda | 14:15 | |
sshnaidm | jaosorior, the same, everything fails | 14:16 |
jaosorior | sshnaidm: well, I can reproduce it in my local environment, I'll take a look | 14:16 |
*** tiswanso has joined #tripleo | 14:17 | |
sshnaidm | jaosorior, do you have FAILED (id=5, failures=15 (+15)) ? | 14:17 |
jaosorior | I have 15 failures | 14:18 |
jaosorior | different id, but that's to be expected | 14:18 |
*** jerrygb has quit IRC | 14:19 | |
jaosorior | the issue is this: https://review.openstack.org/#/c/395702/3/instack_undercloud/undercloud.py we're setting the sample_default, not the actual default value | 14:19 |
*** sudipto_ has joined #tripleo | 14:19 | |
*** sudipto has joined #tripleo | 14:19 | |
jaosorior | sshnaidm: fixing | 14:20 |
jaosorior | fixed | 14:20 |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack/instack-undercloud: Remove deprecated network range 192.0.2.0 https://review.openstack.org/395702 | 14:20 |
jaosorior | sshnaidm: ^^ | 14:20 |
d0ugal | jtomasek, akrivoka - back. | 14:20 |
*** jkilpatr has quit IRC | 14:20 | |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: WIP: Test SoftwareConfigTransport: ZAQAR_MESSAGE https://review.openstack.org/344365 | 14:20 |
jaosorior | sshnaidm: this is the only thing that changed https://review.openstack.org/#/c/395702/3..4/instack_undercloud/undercloud.py | 14:20 |
d0ugal | jtomasek, akrivoka - I am trying to get a spec together for "plan management". so I expect i'll touch on this stuff too, but I wont mention the GUI at all | 14:21 |
panda | bandini: are you a mysql god ? | 14:21 |
sshnaidm | jaosorior, thanks, not obvious for me :) | 14:21 |
*** jkilpatr has joined #tripleo | 14:22 | |
jtomasek | d0ugal: cool | 14:22 |
*** jkilpatr has quit IRC | 14:22 | |
*** zoli|lunch is now known as zoli|wfh | 14:22 | |
bandini | panda: lol I barely know how to spell it. but thankfully we have dciabrin and mike around ;) what's the problem? | 14:23 |
jaosorior | shardy: is it cool if I specify which scenario is run on what branch in another CR? | 14:24 |
jtomasek | d0ugal: could you please add me as a reviewer when you have the patch ready? | 14:24 |
d0ugal | jtomasek: sure, hoping for a first draft early tomorrow | 14:25 |
jaosorior | sshnaidm: no biggie, glad to help | 14:25 |
jtomasek | d0ugal: thanks | 14:25 |
Ng | jaosorior: ccamacho: thanks for the +2s, 396092 now has (sufficiently) green CI :) | 14:25 |
openstackgerrit | Dan Prince proposed openstack/tripleo-puppet-elements: Move nova.conf truncation into separate element https://review.openstack.org/392559 | 14:26 |
shardy | jaosorior: sure | 14:26 |
panda | bandini: then .. die! (force lightning) | 14:26 |
ccamacho | Ng awesome, I was going to make the review in the morning but was already landed :) | 14:26 |
Ng | ccamacho: :D | 14:27 |
bandini | panda: rotfl | 14:27 |
sshnaidm | panda, can you please try to recreate your dev env? I wonder if "nova boot" works for you.. | 14:27 |
*** zoli|wfh is now known as zoli|brb | 14:27 | |
openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Use j2 loops in post.j2.yaml https://review.openstack.org/396230 | 14:27 |
b00tcat | hey y'all - when building the undercloud images I'm getting: "ironic-python-agent already found in services list. Check ironic-agent for duplicate service name." and cannot continue, does this ring a bell? | 14:28 |
b00tcat | I could build overcloud-full just fine before | 14:28 |
panda | sshnaidm: ok | 14:29 |
*** jkilpatr has joined #tripleo | 14:29 | |
b00tcat | sorry, overcloud iamges* | 14:30 |
*** links has quit IRC | 14:31 | |
openstackgerrit | Juan Antonio Osorio Robles proposed openstack-infra/tripleo-ci: Add note about scenario004 in README https://review.openstack.org/396238 | 14:32 |
jaosorior | shardy: what about this? ^^ | 14:32 |
openstackgerrit | Dan Prince proposed openstack/tripleo-heat-templates: Hiera optimization: use a new hiera hook https://review.openstack.org/384757 | 14:33 |
panda | jaosorior: where exactly are we using scenarios ? | 14:33 |
*** jaosorior has quit IRC | 14:34 | |
openstackgerrit | Jiri Tomasek proposed openstack/tripleo-specs: GUI Deployment Configuration improvements https://review.openstack.org/393365 | 14:34 |
*** jaosorior has joined #tripleo | 14:35 | |
*** b00tcat has quit IRC | 14:36 | |
jpich | d0ugal: I'm talking about https://bugs.launchpad.net/tripleo/+bug/1640815 with florianf but there seems to be a Mistral bug in there as well (in api.log) - might be of interest! | 14:37 |
openstack | Launchpad bug 1640815 in tripleo "Selecting too many environments breaks Mistral (?)" [High,Triaged] | 14:37 |
d0ugal | jpich: oooh, fun | 14:37 |
panda | sshnaidm: it's taking a lot of time to spawn | 14:38 |
d0ugal | jpich: TypeError: __init__() takes exactly 4 arguments (2 given) | 14:38 |
panda | sshnaidm: but worked | 14:38 |
d0ugal | jpich: that bit? | 14:38 |
jpich | d0ugal: Yes | 14:38 |
sshnaidm | panda, so nova boot worked> | 14:38 |
d0ugal | jpich: that sounds horrific. I'll take a look | 14:38 |
sshnaidm | panda, which version of nova client do you have? | 14:38 |
jpich | d0ugal: That's what gets returned to the API consumer which is not the bestest :) Sorry for the crappy formatting on the bug, I'll see if I can do better | 14:38 |
jpich | florianf: Did you see the same output about the SizeLimitException in the engine.log? | 14:39 |
jaosorior | panda: what do you mean? | 14:39 |
d0ugal | jpich: lol, it's fine. I'll see what I can find out. | 14:39 |
*** jcoufal has quit IRC | 14:39 | |
panda | sshnaidm: using the one in jumphost (6.0.0-1.el7) | 14:40 |
florianf | jpich: let me see. | 14:40 |
sshnaidm | panda, do you run this from jumphost? | 14:40 |
panda | jaosorior: do we have gate or periodic jobs for those ? are they used somewhere else ? | 14:40 |
panda | sshnaidm: yes | 14:40 |
sshnaidm | panda, and it's rh1, right? | 14:40 |
jaosorior | panda: they run on the gate | 14:40 |
jaosorior | panda: and only run when specific files are modified, depending on the file it's the scenario we run | 14:41 |
florianf | jpich: yep, got that error too | 14:41 |
openstackgerrit | yolanda.robla proposed openstack/tripleo-quickstart: Create directories with root https://review.openstack.org/384892 | 14:41 |
jaosorior | panda: here's the definition https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L1553 | 14:42 |
*** zoli|brb is now known as zoli|wfh | 14:43 | |
*** numans has quit IRC | 14:43 | |
*** numans has joined #tripleo | 14:43 | |
panda | jaosorior: ah, thanks. | 14:44 |
d0ugal | jpich: I can't see how that error is possible | 14:44 |
florianf | jpich: OK, it doesn't seem to have to do with the TLS endpoint env. if I select an environment whose file has a similar size (like network iso), it hits the sizelimit error too | 14:45 |
florianf | d0ugal: ^ | 14:45 |
d0ugal | florianf: What do you mean by "select an environment"? | 14:45 |
jpich | d0ugal: I've got it in both a Newton and Ocata env (<- could give you access to that one?), let me know if I can give you any additional information/logs? Should I file a "FYI" Mistral bug? | 14:46 |
jpich | d0ugal: "enable" - it's checkboxes from the UI side :) | 14:46 |
d0ugal | jpich: Yeah, probably a good idea. We should never get tracebacks in the api.log | 14:46 |
jpich | florianf: Ok, I shall try adjusting the Mistral conf and see if that helps... just need to figure out which configuration setting that could be | 14:46 |
florianf | d0ugal: checking the box for an environment in the ui | 14:47 |
d0ugal | jpich, florianf: is the size limit an issue with the size of the workflow input? or something else? | 14:47 |
jpich | d0ugal: I'm not sure... seems to be related to task execution from the log extract? | 14:48 |
tzumainn | jpich, florianf, just a heads up - the project I'm working on got a similar error half a year ago, we had to adjust the max_json_body_size in /etc/heat/heat.conf | 14:48 |
florianf | jpich: sorry, forget what I said, I didn't save my overall settings, so one of the tls endpoint envs was still checked. | 14:48 |
jpich | tzumainn: In heat, huh. This one seems to be coming from Mistral but I'll have a look there too | 14:49 |
jpich | florianf: Ok! | 14:49 |
tzumainn | jpich, I'm just wondering if you guys will hit it once you try and deploy an overcloud with a lot of environments (in other words, once the mistral issue is cleared up) | 14:49 |
tzumainn | if you do, it's probably ^ | 14:50 |
jpich | tzumainn: Bleh, thanks for the heads-up | 14:50 |
*** jcoufal has joined #tripleo | 14:50 | |
florianf | d0ugal, jpich: arent these supposed to be valid json strings? https://github.com/openstack/tripleo-heat-templates/blob/master/environments/tls-endpoints-public-ip.yaml | 14:53 |
*** derekh has joined #tripleo | 14:53 | |
jaosorior | Any +As for this? https://review.openstack.org/#/c/392473/ :D | 14:54 |
jpich | florianf: I'm not sure, that's a good point | 14:55 |
d0ugal | tzumainn: we have people deploying with lots of envs via the CLI I believe. | 14:55 |
d0ugal | florianf: are they not? | 14:56 |
jpich | florianf: As a dictionary it seems a bit broken? Might be worthwhile comparing with other yaml and see? | 14:56 |
d0ugal | florianf: well, I mean, it's yaml. /me is confused. | 14:56 |
jaosorior | shardy: could you check this out? https://review.openstack.org/#/c/392473/ | 14:56 |
florianf | d0ugal: no, the params in this template look like a python dict notation | 14:57 |
florianf | d0ugal: but that isn't valid json | 14:57 |
shardy | florianf: it's a yaml map, not a json string | 14:57 |
d0ugal | florianf: it's valid YAML. | 14:59 |
d0ugal | yeah | 14:59 |
jpich | d0ugal: Surprised to see the dictionary keys are not between quotes, but that is valid yaml as well then? | 14:59 |
jpich | cool | 14:59 |
florianf | shardy, d0ugal: OK, thanks. It was just a wild guess... | 14:59 |
d0ugal | In yaml, almost everything is valid. it is scary | 14:59 |
jpich | Understood! | 14:59 |
* jpich goes back to config parameter hunting | 14:59 | |
jpich | florianf: Did you manage to reproduce the error with other environments in the end? | 15:00 |
florianf | jpich: nope | 15:00 |
jpich | florianf: Interesting... Would you mind adding that information to the bug? | 15:00 |
florianf | jpich: yep | 15:01 |
jpich | florianf: :D | 15:01 |
florianf | jpich: uh... no, I wouldn't, of course :-) | 15:01 |
*** dsneddon has quit IRC | 15:04 | |
*** noslzzp has quit IRC | 15:04 | |
jaosorior | therve: is zaqar supported in newton and below? | 15:04 |
jaosorior | huh | 15:05 |
jaosorior | sorry | 15:05 |
jaosorior | thrash: is the zaqar tripleo profile supported in newton and below? | 15:05 |
*** noslzzp has joined #tripleo | 15:05 | |
thrash | jaosorior: no, it hasn't even landed yet. | 15:05 |
*** lblanchard has quit IRC | 15:05 | |
thrash | jaosorior: Is that why barbican was taken out? | 15:06 |
jaosorior | thrash: so, for the barbican profile I had an issue that... it merged and broke CI... and that was because the scenarios run for all branches, so if the service is not available in an earlier release, it will crash | 15:06 |
thrash | jaosorior: gotcha | 15:06 |
jaosorior | so you might wanna wait until the scenario004 merges | 15:06 |
jaosorior | I've set that up so it runs in branches newer than newton | 15:06 |
thrash | jaosorior: ok. Will have to change layout.yaml... | 15:07 |
openstackgerrit | Merged openstack/tripleo-heat-templates: Select bootstrap node by list index not name https://review.openstack.org/395699 | 15:07 |
jaosorior | thrash: this is the commit to project-config https://review.openstack.org/#/c/396062/ | 15:07 |
thrash | jaosorior: can I just move the zaqar and mistral stuff in that patch? | 15:08 |
jpich | florianf: Updating the Mistral configuration fixed this, I'll update the bug and possibly propose a patch | 15:08 |
thrash | jaosorior: or if you want to do it. Would be easier if it was done in one fell swoop. | 15:09 |
jpich | florianf: I still think there's something funky if folks never hit anything similar when deploying with a bunch of environments with the CLI | 15:09 |
*** lblanchard has joined #tripleo | 15:09 | |
thrash | jaosorior: https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L1569-L1570 and https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L1579-L1580 | 15:10 |
jaosorior | thrash: you can do it if you want to | 15:10 |
thrash | jaosorior: ok. i'll hijack it. :) | 15:10 |
*** numans has quit IRC | 15:10 | |
*** yolanda has quit IRC | 15:11 | |
*** rbrady is now known as rbrady-brb | 15:12 | |
openstackgerrit | Katerina Pilatova proposed openstack/tripleo-validations: callback_plugins/validation_output.py: improved output https://review.openstack.org/396135 | 15:12 |
jaosorior | ccamacho: could you check out these two? https://review.openstack.org/#/c/392090/ https://review.openstack.org/#/c/392095/ | 15:13 |
thrash | jaosorior: updated. | 15:14 |
jaosorior | thrash: commented | 15:16 |
*** yamahata has joined #tripleo | 15:17 | |
thrash | jaosorior: ack | 15:17 |
openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Normalize civetweb binding address if IPv6 https://review.openstack.org/395785 | 15:17 |
openstackgerrit | James Slagle proposed openstack/tripleo-heat-templates: Select bootstrap node by list index not name https://review.openstack.org/396280 | 15:19 |
florianf | jpich: yeah, that's strange. I would have expected it to be reproducable with other files as well. but maybe I just didn't try enough... | 15:19 |
jpich | florianf: I might have an idea why it doesn't reproduce from the CLI, it looks like all the environments get bunched up together in one big "user-environment.yaml" file | 15:20 |
*** jaosorior has quit IRC | 15:20 | |
*** skramaja has joined #tripleo | 15:21 | |
openstackgerrit | yolanda.robla proposed openstack/python-tripleoclient: Support whole disk images in TripleO https://review.openstack.org/394426 | 15:22 |
d0ugal | jpich: yeah, that will likely hide it | 15:22 |
openstackgerrit | Katerina Pilatova proposed openstack/tripleo-validations: callback_plugins/validation_output.py: improved output https://review.openstack.org/396135 | 15:24 |
*** ealcaniz has quit IRC | 15:24 | |
openstackgerrit | Brad P. Crochet proposed openstack-infra/tripleo-ci: Add Zaqar to scenario004 https://review.openstack.org/365026 | 15:24 |
openstackgerrit | Brent Eagles proposed openstack/os-net-config: Add support for name replacement in OVS_EXTRA https://review.openstack.org/395795 | 15:24 |
florianf | jpich, d0ugal: Do we mind that cli and ui do this differently? Just asking because in general we want to keep the differences minimal... | 15:24 |
*** yolanda has joined #tripleo | 15:26 | |
jpich | florianf: Personally I like the UI way better because it's easier to see what envs are included just by peeking into Mistral, but I don't know why the CLI is doing it this way and/or how painful it would be to do it differently. Ideally they'd be more similar though, as you say :) | 15:26 |
gbarros | #tripleo, having some issues with a deployment after adding a few extraconfig, see this error: http://pastebin.com/b8077rzQ | 15:26 |
d0ugal | florianf, jpich - the CLI only does this for backwards compat issues | 15:27 |
florianf | d0ugal: which means, we shouldn't touch it? | 15:27 |
openstackgerrit | Martin Mágr proposed openstack/tripleo-heat-templates: Fix typo in Keystone Sensu subscription https://review.openstack.org/394824 | 15:27 |
paramite | Guys, can we get the following patch merged, so I can backport it to stable/newton (and have that merged also)? ^ | 15:27 |
d0ugal | florianf: shouldn't touch the CLI? :) | 15:28 |
d0ugal | florianf: it means we want to do it the same way as the GUI probably, but can't at the moment | 15:28 |
florianf | d0ugal: :-D | 15:28 |
trown | gbarros: that looks like the extraconfig has syntax issues, could you post the environment with that too? | 15:28 |
openstackgerrit | Brad P. Crochet proposed openstack-infra/tripleo-ci: Add Mistral to scenario004 https://review.openstack.org/368805 | 15:28 |
openstackgerrit | Brent Eagles proposed openstack/tripleo-heat-templates: WIP: Add fail_mode workaround for connectivity issues on reboot https://review.openstack.org/396285 | 15:28 |
florianf | d0ugal: OK, that's what I wanted to know | 15:28 |
florianf | :-) | 15:28 |
d0ugal | florianf, jpich - I can share you the nitty gritty details if you like ;) | 15:28 |
florianf | d0ugal: Sure! | 15:29 |
jpich | d0ugal: I'm kinda curious :) | 15:29 |
openstackgerrit | Brent Eagles proposed openstack/tripleo-heat-templates: WIP: Add fail_mode workaround for connectivity issues on reboot https://review.openstack.org/396285 | 15:30 |
gbarros | trown, here is the env. The new section is NovaComputeExtraConfig http://pastebin.com/FquyXRWk | 15:30 |
d0ugal | florianf, jpich - haha, okay, give me a few mins | 15:31 |
*** jerrygb has joined #tripleo | 15:31 | |
*** jerrygb has quit IRC | 15:31 | |
jpich | d0ugal, florianf: "Long, long ago, when TripleO was but a little owlet..." | 15:31 |
chem | any core for validating https://review.openstack.org/#/c/395053/ ? | 15:31 |
*** jerrygb has joined #tripleo | 15:32 | |
*** abregman has quit IRC | 15:33 | |
trown | gbarros: line 61/62 look suspicious there | 15:34 |
gbarros | trown, i was thinking the same thing | 15:35 |
*** b00tcat has joined #tripleo | 15:36 | |
panda | any network guys here ? | 15:36 |
*** abehl has quit IRC | 15:37 | |
*** sudipto_ has quit IRC | 15:37 | |
*** sudipto has quit IRC | 15:37 | |
panda | I wonder if all the bad VRRP packets here are normal http://logs.openstack.org/92/388592/2/check-tripleo/gate-tripleo-ci-centos-7-ovb-nonha/2959df7/logs/overcloud-controller-0/var/log/messages | 15:37 |
d0ugal | jpich, florianf: Okay, so with the CLI deploy command you can do something like this "openstack overcloud deploy --templates -e myenv.yaml" | 15:37 |
d0ugal | jpich, florianf: that would use the default templates with a custom env. simple enough. | 15:37 |
tzumainn | d0ugal, use the default templates without the need for a 'plan'? | 15:38 |
d0ugal | tzumainn: this is historical information only :) but yeah | 15:38 |
tzumainn | I like stories too, DAD | 15:39 |
*** jerrygb has quit IRC | 15:39 | |
d0ugal | jpich, florianf: However, the custom env can include relative file paths like this: https://github.com/openstack/tripleo-heat-templates/blob/master/environments/net-bond-with-vlans.yaml | 15:39 |
openstackgerrit | Julie Pichon proposed openstack/instack-undercloud: Increase Mistral Task Size limit https://review.openstack.org/348216 | 15:39 |
d0ugal | jpich, florianf: so one env could reference files anywhere locally - for this to continue to work, we need to use heatclient to discover all the referenced files but then we end up with a merged environment | 15:39 |
*** panda has quit IRC | 15:40 | |
d0ugal | jpich, florianf: hopefully that makes sense | 15:40 |
d0ugal | tzumainn: then go read a book :P | 15:40 |
*** panda has joined #tripleo | 15:40 | |
tzumainn | lol | 15:40 |
tzumainn | d0ugal, wait, just to confirm - openstack overcloud deploy *right now* doesn't require a mistral plan to exist? | 15:41 |
d0ugal | tzumainn: no, it does | 15:41 |
d0ugal | tzumainn: every deploy uses plans | 15:41 |
tzumainn | okay, so that's what you meant by historical | 15:41 |
tzumainn | I GOT IT | 15:41 |
d0ugal | tzumainn: but the CLI creates plans on the fly | 15:41 |
d0ugal | tzumainn: to maintain the existing deploy command | 15:41 |
tzumainn | d0ugal, ah, so if the plan doesn't exist, it creates it | 15:41 |
d0ugal | tzumainn: Yeah, or updates it | 15:41 |
florianf | d0ugal: thanks! it does | 15:42 |
*** skramaja has quit IRC | 15:42 | |
*** ayoung has joined #tripleo | 15:43 | |
jtomasek | d0ugal: question is, do we want to support this or deprecate it and get rid of that functionality? | 15:44 |
jpich | d0ugal: Thank you! Want to add this as a historical note on that bug too, so folks understand why the behaviour is different? 0=) | 15:44 |
jtomasek | shardy, d0ugal, akrivoka: see last comment here https://bugs.launchpad.net/tripleo/+bug/1626955 is it crazy idea? | 15:45 |
openstack | Launchpad bug 1626955 in tripleo "No way to override roles_data.yaml" [High,In progress] - Assigned to Steven Hardy (shardy) | 15:45 |
d0ugal | jtomasek: we want to deprecate it | 15:45 |
jtomasek | ok | 15:45 |
d0ugal | jpich: wich bug? | 15:45 |
jpich | d0ugal: https://bugs.launchpad.net/tripleo/+bug/1640815 , I note the difference between UI and CLI in comment #3 | 15:45 |
openstack | Launchpad bug 1640815 in tripleo "Selecting too many environments breaks Mistral (?)" [High,In progress] - Assigned to Julie Pichon (jpichon) | 15:45 |
d0ugal | jpich: sure, I can add something | 15:46 |
d0ugal | jpich: actually, we need a CLI bug for this, maybe I'll open one and that can be referenced? | 15:46 |
d0ugal | jtomasek: it doesn't seem that crazy to me :) | 15:47 |
jpich | d0ugal: Sounds good :) | 15:47 |
shardy | jtomasek: yes we could store the data in the mistral environment, but IMO we don't need a parameter because there's $RoleServices parameters already for defining the list of services | 15:47 |
jpich | I might move the bug I just mentioned to Critical though, as it's impossible to deploy anything with more than 2/3 environments without it... | 15:48 |
shardy | jtomasek: the main advantage of keeping the sample roles_data in tripleo-heat-templates is it's obvious where to update it when there's a new service added | 15:48 |
shardy | so personally I favor keeping it there, even if we copy the data into the mistral environment during the plan creation | 15:48 |
jtomasek | shardy: ok, yeah, that answers the services my question | 15:48 |
jtomasek | shardy: yeah, the proposal is just about renaming that file to plan.yaml and include more information there, not just roles stuff and use this file to populate mistral environment when plan is created | 15:49 |
jtomasek | shardy: so the file would be part of the repository but not part of plan (stored in swift) since it's content would be in mistral environment | 15:50 |
slagle | akrivoka: jtomasek jrist: can you comment on https://review.openstack.org/#/c/391799/ ? | 15:50 |
*** ayoung has quit IRC | 15:50 | |
slagle | there are some questions there about whether bumping the patch version for tripleo-ui is appropriate | 15:50 |
slagle | or if we should bump the minor version instead | 15:50 |
jrist | slagle: looking | 15:50 |
jrist | well see UI is in a tough position | 15:51 |
jrist | we had to scramble to get our patches up | 15:52 |
trozet | dprince, slagle: can you please review https://review.openstack.org/#/c/394640/ ? | 15:53 |
dprince | trozet: +2 +A | 15:54 |
b00tcat | in the latest releases of TripleO, let's assume that I want to add a custom service to a node that is being deployed - how would I do this? | 15:54 |
trozet | dprince: ty. Also this one if you have time https://review.openstack.org/#/c/394651/ | 15:54 |
openstackgerrit | Bogdan Dobrelya proposed openstack/tripleo-quickstart: Add support of deploy on a VM w/o nested virt https://review.openstack.org/393720 | 15:54 |
jtomasek | slagle: replied, jrist: I cherrypicked/backported only bugfix patches, not all of them have bug referenced, sorry about that | 15:55 |
jrist | thanks jtomasek | 15:55 |
shardy | jtomasek: that sounds reasonable, what sort of extra data are you expecting to put in there? | 15:55 |
*** bogdando has joined #tripleo | 15:55 | |
shardy | jtomasek: I'd suggest maybe we rename the file in t-h-t to default_plan.yaml or something | 15:56 |
b00tcat | I see that all the services are listed in overcloud-resource-registry-puppet.j2.yaml | 15:56 |
shardy | jtomasek: the name doesn't really matter provided the mistral action reads the content and passes it to jinja2 | 15:56 |
*** Guest40216 has quit IRC | 15:56 | |
shardy | b00tcat: you can just copy the list of default services from overcloud-resource-registry-puppet.j2.yaml for the role you want to customize, and add your new service | 15:57 |
shardy | e.g: | 15:57 |
jtomasek | shadower: yep, the name of that file needs to be standardized that is all. So in theory, that file could include roles_data.yaml, capabilities-map.yaml and potentially what mistral environment can include | 15:58 |
jtomasek | shardy: ^ | 15:58 |
*** mcornea has quit IRC | 15:59 | |
shardy | b00tcat: http://paste.openstack.org/show/588760/ | 15:59 |
shardy | http://hardysteven.blogspot.co.uk/2016/08/tripleo-composable-services-101.html | 15:59 |
shardy | also see that blog post (sorry tripleo-docs still todo) | 15:59 |
b00tcat | thanks shardy - if I have the time I might even contribute to the docs myself :D | 16:00 |
shardy | b00tcat: feel free! :) | 16:00 |
jtomasek | shardy: exporting the plan would then be a matter of providing tar which includes swift container contents + plan.yaml which would include contents of mistral environment | 16:00 |
shardy | b00tcat: if you do start working on them please let me know then I'll know not to do it myself | 16:00 |
shardy | I've been meaning to do them for a few weeks | 16:01 |
shardy | jtomasek: ack, that would be fine I think | 16:01 |
shardy | you could just write the roles_data.yaml data into plan.yaml regardless of any renaming | 16:01 |
b00tcat | sure! | 16:02 |
beagles | general heads up... I'm getting a lot of checksum errors with ssh sessions between undercloud and overcloud guests | 16:02 |
*** rajinir has joined #tripleo | 16:02 | |
*** BLOODBATH has joined #tripleo | 16:02 | |
*** BLOODBATH is now known as Guest13773 | 16:03 | |
beagles | let me know if anybody else runs into this kind of thing .. if for no other reason than to let me know I'm not going mad :) | 16:03 |
*** ebarrera has quit IRC | 16:03 | |
*** jkilpatr_ has joined #tripleo | 16:03 | |
beagles | I haven't been able to work out the cause yet.... symptoms would include really long time to ssh from one node to the other | 16:04 |
*** ayoung has joined #tripleo | 16:04 | |
jtomasek | shardy: on export, yes. having it all in one file in tht would probably simplify tripleo-common as it would just look for single file and populate the mistral environment with it when plan is created. Rather than looking separately for 3 files. Also maybe it would be simpler to the user as he'd know that all 'plan metadata' are in plan.yaml | 16:04 |
beagles | (particulars, virtual environment with virtio drivers only) | 16:05 |
jtomasek | (by 3 files I mean roles_data.yaml, capabilities-map.yaml and mistral-environment.yaml - which would be needed to introduce for exporting plan) | 16:05 |
jtomasek | shardy: ^ | 16:06 |
shardy | jtomasek: ack, but I think we'll need to support roles_data.yaml for backwards compatibility anyway, so we can just look for plan.yaml (or that content in the mistral environment), then fall back to reading roles_data.yaml | 16:06 |
jtomasek | shardy: oh, right. should I create spec for it or blueprint is enough? | 16:07 |
shardy | jtomasek: I think a blueprint is enough | 16:07 |
jtomasek | shardy: ok, I'll create one, thanks | 16:07 |
jtomasek | d0ugal: actually this blueprint might fit in your plans management spec, right? | 16:08 |
*** links has joined #tripleo | 16:09 | |
gbarros | trown that wasnt it, same error | 16:09 |
d0ugal | jtomasek: Yeah, maybe. I am probably going to do a blueprint btw | 16:09 |
jtomasek | d0ugal: ok | 16:09 |
*** udesale_ has quit IRC | 16:11 | |
*** rcernin has quit IRC | 16:11 | |
*** apetrich has quit IRC | 16:15 | |
*** iranzo has quit IRC | 16:15 | |
*** apetrich has joined #tripleo | 16:16 | |
*** [1]cdearborn has joined #tripleo | 16:18 | |
*** ramishra has quit IRC | 16:18 | |
*** ramishra has joined #tripleo | 16:20 | |
*** chandankumar has joined #tripleo | 16:21 | |
*** nyechiel has quit IRC | 16:21 | |
*** tremble has quit IRC | 16:22 | |
openstackgerrit | Merged openstack/tripleo-validations: callback_plugins/validation_output.py: improved output https://review.openstack.org/396135 | 16:24 |
dhill_ | hi guys | 16:25 |
*** dsneddon has joined #tripleo | 16:25 | |
*** chandankumar has quit IRC | 16:31 | |
*** rcernin has joined #tripleo | 16:32 | |
*** links has quit IRC | 16:37 | |
ansiwen | shardy, mwhahaha, chem: when you have a calm minute, you might look at https://review.openstack.org/#/c/336562/ I'd love to move forward with this neverending story | 16:37 |
*** mhenkel has joined #tripleo | 16:39 | |
chem | ansiwen: this is working for you rigth ? | 16:39 |
*** TSCHAK has joined #tripleo | 16:41 | |
shardy | ansiwen: is there a corresponding t-h-t patch which consumes the profile? | 16:41 |
*** rbrady-brb is now known as rbrady | 16:44 | |
ansiwen | shardy: well, that will follow | 16:45 |
ansiwen | chem: I have no idea yet, there is no consumer. | 16:46 |
chem | gfidente: for the blockstorage stuff, I tried those two environments http://paste.openstack.org/show/588774/ to disable ceph and enable iscsi but nothing get passed down in the hieradata : I just re-run the convergence step with the env added | 16:46 |
ansiwen | shardy: so, should I make the THT patch as well and make it depending on this? | 16:46 |
chem | gfidente: if you don't have anything in mind regarding this, I'm just going to remove the blockstorage from my test env for now (as it may not be a requirement) | 16:47 |
chem | ansiwen: ack | 16:47 |
gfidente | chem you don't see them in hieradata at all? | 16:47 |
ansiwen | chem, mwhahaha: thanks for the review. I just wanted to make sure you guys don't see an obvious mistake, and that my general approach is correct. | 16:48 |
chem | gfidente: nope nothing move in this directory | 16:48 |
*** chandankumar has joined #tripleo | 16:48 | |
chem | gfidente: the /etc/puppet/hieradata/extraconfig.yaml desperately stays empty | 16:49 |
*** ccamacho has quit IRC | 16:51 | |
chem | gfidente: so this doesn't work BlockStorageExtraConfig: {CinderEnableIscsiBackend: true, CinderEnableRbdBackend: false} (in openstack stack environment show overcloud) | 16:51 |
gfidente | chem the file should be named | 16:52 |
gfidente | volume_extraconfig | 16:52 |
chem | gfidente: ah, yes, sorry | 16:52 |
gfidente | but you can only pass the puppet params via extraconfig | 16:53 |
gfidente | not the template params | 16:53 |
chem | gfidente: oki, I've just got success :) You have to use the puppet parameter directly | 16:53 |
*** abehl has joined #tripleo | 16:53 | |
gfidente | chem yep | 16:54 |
chem | gfidente: in the paste this is the first version. | 16:54 |
chem | gfidente: oki, parameters are there, waiting for the converge to happen and then I'll add it as a workaround. | 16:54 |
chem | gfidente: thanks a lot. | 16:55 |
gfidente | chem cool, ty | 16:55 |
chem | gfidente: oki, step4 -> enabled_backends = tripleo_iscsi, | 16:56 |
*** ohamada has quit IRC | 16:56 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: POC: Use tripleo-repos in tripleo.sh https://review.openstack.org/395813 | 16:57 |
gfidente | chem cool | 16:57 |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Always use cached images in updates job https://review.openstack.org/386713 | 16:59 |
*** abehl has quit IRC | 17:02 | |
*** hewbrocca is now known as hewbrocca_afk | 17:03 | |
*** aufi has quit IRC | 17:06 | |
*** dhill_ has quit IRC | 17:06 | |
*** jpich has quit IRC | 17:06 | |
*** dtantsur is now known as dtantsur|afk | 17:07 | |
openstackgerrit | Steven Hardy proposed openstack/tripleo-heat-templates: Fix inconsistent Manila service naming https://review.openstack.org/396317 | 17:07 |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: Always use cached images in updates job https://review.openstack.org/386713 | 17:08 |
*** ccamacho has joined #tripleo | 17:08 | |
*** chandankumar has quit IRC | 17:10 | |
*** dhill_ has joined #tripleo | 17:11 | |
bnemec | Ouch, zuul restart. rh1 is about to have a bad day. | 17:11 |
sshnaidm | Have anybody seen error in Ci jobs like: Failed to call refresh: Could not restart Service[httpd] Execution of '/usr/bin/systemctl restart httpd' returned 1: Job for httpd.service failed because a timeout was exceeded | 17:12 |
openstackgerrit | Giulio Fidente proposed openstack/puppet-tripleo: Normalize civetweb binding address if IPv6 https://review.openstack.org/395785 | 17:12 |
slagle | bnemec: do we need to do anything in response to the restart? | 17:13 |
slagle | clear out all jobs/envs or anything? | 17:13 |
bnemec | slagle: I don't think so, but everything that was running will need to be rechecked. | 17:13 |
bnemec | IIRC anyway. | 17:13 |
slagle | yea ok, that's what i thought as well | 17:13 |
slagle | we'll have to keep an eye on it to just make sure new stuff is getting scheduled | 17:13 |
bnemec | I'm wondering if this would be a good time to merge https://review.openstack.org/#/c/384731/ | 17:14 |
shardy | ansiwen: yes, then you'll have a way to test the profile | 17:14 |
shardy | testing it without a tripleo patch will be pretty inconvenient I think | 17:14 |
trozet | mwhahaha: when you have a minute: https://review.openstack.org/#/c/396318/ | 17:14 |
*** paramite has quit IRC | 17:15 | |
*** abehl has joined #tripleo | 17:15 | |
mwhahaha | trozet: sure | 17:15 |
trozet | mwhahaha: thanks man | 17:15 |
*** _milan_ has joined #tripleo | 17:19 | |
panda | dsneddon: I wonder if all the bad VRRP packets here are normal http://logs.openstack.org/92/388592/2/check-tripleo/gate-tripleo-ci-centos-7-ovb-nonha/2959df7/logs/overcloud-controller-0/var/log/messages | 17:19 |
dsneddon | panda, That doesn't look normal to me. But 192.0.2.6 is a strange address to see there. Did you modify the 192.0.2.2/192.0.2.3 defaults in undercloud.conf? | 17:20 |
dsneddon | panda, Because 192.0.2.6 conflicts with the normal DHCP range of 192.0.2.5-24 | 17:21 |
*** milan has quit IRC | 17:22 | |
dsneddon | panda, Oh wait, I think I'm misreading these logs. These are from the Overcloud controller? | 17:22 |
panda | dsneddon: yes | 17:22 |
dsneddon | panda, Not sure what is going on there, but those aren't normal log messages. | 17:23 |
panda | dsneddon: I'll see how often we get those | 17:23 |
dsneddon | panda, Is 192.0.2.6 the address of the Control Plane VIP? | 17:24 |
panda | bnemec: this finally passed on all gates https://review.openstack.org/382515 | 17:24 |
*** jlinkes has quit IRC | 17:24 | |
dsneddon | panda, You can run "neutron port-list" against the undercloud to figure out which IPs are assigned on that network. | 17:24 |
bnemec | panda: Yes, but I still don't want it. :-) | 17:24 |
panda | bnemec: ... | 17:24 |
derekh | bnemec: would this be a good time to test drive https://review.openstack.org/#/c/384731/ ? | 17:25 |
derekh | given the restart and all | 17:25 |
bnemec | derekh: <bnemec> I'm wondering if this would be a good time to merge https://review.openstack.org/#/c/384731/ | 17:25 |
trozet | mwhahaha: fixed the whitespace | 17:25 |
bnemec | I was thinking the same thing. | 17:25 |
panda | dsneddon: It's a bit hard to reproduce in a live environment, but I can try | 17:26 |
dsneddon | panda, I'm just trying to interpret the log messages, and it would help if I knew what 192.0.2.6 was. | 17:26 |
ansiwen | shardy: so, this is my first "optional" feature, so I will not hardwire it into the controller node role. can you point me to a minimal as possible example of such an optional service? | 17:27 |
dsneddon | panda, If that's an IP of a controller, that would be different than if it was the IP of the VIP. | 17:27 |
panda | bnemec: can you answer to this then stating you don't want this to be merged ? http://lists.openstack.org/pipermail/openstack-dev/2016-October/105934.html | 17:28 |
shardy | ansiwen: Nothing is hardwired into the controller role since back in Newton | 17:28 |
shardy | ansiwen: https://github.com/openstack/tripleo-heat-templates/tree/master/puppet/services | 17:28 |
shardy | there are many examples in there | 17:28 |
ansiwen | shardy: thanks! | 17:28 |
shardy | all follow basically the same pattern, so you can pick any of them | 17:28 |
bnemec | panda: I left an early review on https://review.openstack.org/#/c/382515/ explaining why I don't like it. | 17:29 |
shardy | ansiwen: also see http://docs.openstack.org/developer/tripleo-docs/developer/tht_walkthrough/tht_walkthrough.html | 17:29 |
shardy | and | 17:29 |
*** dsariel has quit IRC | 17:29 | |
ansiwen | shardy: oh, this I know of course... but I have to reference the yaml file somehow, so that it get's tested | 17:29 |
shardy | http://hardysteven.blogspot.co.uk/2016/08/tripleo-composable-services-101.html | 17:29 |
shardy | ansiwen: yes, you'll add it to overcloud-resource-registry-puppet.j2.yaml | 17:30 |
ansiwen | shardy: I implemented the composable service for nova-compute already... but as I remember I inserted that into the compute role | 17:30 |
ansiwen | shardy: ok, so everything there will be picked up for testing? | 17:30 |
shardy | ansiwen: yeah if it's disabled by default you'll add it to overcloud-resource-registry-puppet.j2.yaml and not to the roles_data.yaml | 17:31 |
shardy | we'll need to add it to a CI scenario | 17:31 |
ansiwen | shardy: got it, thanks! | 17:31 |
shardy | but first step it get it working locally | 17:31 |
*** _milan_ has quit IRC | 17:32 | |
panda | dsneddon: it's the controller VIP | 17:32 |
ansiwen | shardy: "locally", well, on my 64GB tarox I guess ;-) | 17:32 |
ansiwen | shardy: can I test this without a full tripleo deployment? | 17:32 |
*** ccamacho has quit IRC | 17:34 | |
*** katkapilatova has left #tripleo | 17:35 | |
panda | bnemec: yeah, I partially agreed with you, but nobody answered, that's why I created that post, hoping to start a discussion, detailing what I was trying to do. What else I can do now ? | 17:35 |
*** akrivoka has quit IRC | 17:36 | |
*** ccamacho has joined #tripleo | 17:36 | |
panda | I will probably take another month before planets align correctly and I get green on all the gates. | 17:36 |
bnemec | panda: I want to push to get the updates job back in the gate. | 17:37 |
bnemec | We need it anyway, and it solves the ipv6 coverage problem too. | 17:37 |
bnemec | With only two jobs the best we can do is trade one CI hole for another. | 17:37 |
bnemec | slagle: Would you be okay merging https://review.openstack.org/#/c/384731/ to keep the flood of rechecks under control? | 17:38 |
openstackgerrit | Dan Prince proposed openstack/tripleo-puppet-elements: Disable old hiera.yaml o-a-c script https://review.openstack.org/392615 | 17:39 |
slagle | bnemec: i hesitate a bit...to make any changes. has it been tested in any live environment? | 17:40 |
*** hjensas has quit IRC | 17:40 | |
*** hjensas has joined #tripleo | 17:41 | |
*** lucasagomes is now known as lucas-afk | 17:41 | |
sshnaidm | fyi https://bugs.launchpad.net/tripleo/+bug/1640879 | 17:42 |
openstack | Launchpad bug 1640879 in tripleo "CI: apache fails to start on overcloud controller" [Undecided,New] | 17:42 |
*** pkovar has quit IRC | 17:43 | |
bnemec | slagle: There's not really any way to test it live without deploying it to all the te workers on rh1. derekh and I have both tested the semaphore behavior standalone though. | 17:44 |
*** trown is now known as trown|lunch | 17:45 | |
*** owalsh has quit IRC | 17:46 | |
hrybacki | hey all. I'm seeing an 'Enviornment not found <stack> ' when calling `openstack overcloud node delete --stack <stack>' Verified that the stack id is correct -- any pointers? https://paste.fedoraproject.org/477444/79992414/ | 17:48 |
hrybacki | Newton ^^ | 17:48 |
slagle | bnemec: so posix_ipc.Semaphore will block indefinitely? | 17:51 |
slagle | or until we hit the job timeout i guess | 17:52 |
*** derekh has quit IRC | 17:55 | |
bnemec | slagle: Yes: "A timeout of None (the default) implies no time limit. The call will not return until its wait condition is satisfied. " | 17:57 |
bnemec | From the acquire docstring: http://semanchuk.com/philip/posix_ipc/#semaphore | 17:57 |
*** morazi has quit IRC | 17:58 | |
*** thrash is now known as thrash|f00dz | 18:00 | |
openstackgerrit | Merged openstack-infra/tripleo-ci: Redirect image build stderr to stdout https://review.openstack.org/382641 | 18:01 |
openstackgerrit | Merged openstack/python-tripleoclient: Reset the parameter_defaults between deployments via the CLI https://review.openstack.org/394419 | 18:04 |
openstackgerrit | Ben Nemec proposed openstack/python-tripleoclient: Refactor creation of no_proxy list for postconfig https://review.openstack.org/391197 | 18:04 |
*** chandankumar has joined #tripleo | 18:06 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: Ensure heat-domain hiera is in nodes that contain keystone https://review.openstack.org/392519 | 18:07 |
openstackgerrit | Merged openstack/tripleo-heat-templates: Fixes incorrect reference to OpendaylightApiNetwork https://review.openstack.org/394640 | 18:08 |
*** morazi has joined #tripleo | 18:10 | |
openstackgerrit | Ben Nemec proposed openstack/python-tripleoclient: Add unit tests for bracket_ipv6 https://review.openstack.org/396345 | 18:12 |
*** chandankumar has quit IRC | 18:13 | |
*** numans has joined #tripleo | 18:13 | |
openstackgerrit | Attila Darazs proposed openstack/tripleo-quickstart: Force the use of python version 2.7 https://review.openstack.org/368013 | 18:14 |
*** jpena is now known as jpena|off | 18:14 | |
cmyster | .9 | 18:14 |
*** rasca has quit IRC | 18:14 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: POC: Use tripleo-repos in tripleo.sh https://review.openstack.org/395813 | 18:15 |
shardy | ansiwen: I'm not sure I understand, you have a 64G box, why not install a tripleo VM environment and test it on that? | 18:21 |
shardy | that's what most of us are doing | 18:21 |
* shardy only has 32G tho ;( | 18:21 | |
shardy | I don't think you can really develop tripleo profiles or composable services without testing them locally on tripleo | 18:21 |
openstackgerrit | Tim Rozet proposed openstack/tripleo-heat-templates: Fixes incorrect reference to OpendaylightApiNetwork https://review.openstack.org/396348 | 18:22 |
shardy | well, you could, but it will take forever | 18:22 |
openstackgerrit | Gabriele Cerami proposed openstack/tripleo-quickstart: Add configuration of HA IPv6 deployments https://review.openstack.org/380358 | 18:22 |
trozet | dprince, shardy: ^^ cherry pick to stable/newton | 18:23 |
*** chandankumar has joined #tripleo | 18:25 | |
*** chandankumar has quit IRC | 18:26 | |
*** panda_ has joined #tripleo | 18:26 | |
*** dsariel has joined #tripleo | 18:27 | |
*** panda has quit IRC | 18:29 | |
*** chandankumar has joined #tripleo | 18:30 | |
*** cdearborn has joined #tripleo | 18:35 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: Removes deprecated overcloud VIP outputs https://review.openstack.org/394651 | 18:36 |
openstackgerrit | Merged openstack/tripleo-heat-templates: set url_base option in static web middleware https://review.openstack.org/395654 | 18:36 |
*** [1]cdearborn has quit IRC | 18:39 | |
*** shardy has quit IRC | 18:41 | |
*** tosky has quit IRC | 18:43 | |
*** chandankumar has quit IRC | 18:46 | |
*** dprince has quit IRC | 18:48 | |
*** trown|lunch is now known as trown | 18:48 | |
*** thrash|f00dz is now known as thrash | 18:51 | |
*** Guest13773 has quit IRC | 18:51 | |
*** jkilpatr_ has quit IRC | 18:51 | |
*** athomas has quit IRC | 18:52 | |
*** jkilpatr has quit IRC | 18:52 | |
*** florianf has quit IRC | 18:54 | |
*** panda_ is now known as panda|bbl | 18:56 | |
*** chandankumar has joined #tripleo | 18:58 | |
*** amoralej is now known as amoralej|off | 18:58 | |
openstackgerrit | Harry Rybacki proposed openstack/tripleo-quickstart: [WIP] DO NOT MERGE -- Add tuned check to OOOQ https://review.openstack.org/396362 | 18:59 |
*** florianf has joined #tripleo | 18:59 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: Fix race during major-upgrade-pacemaker step https://review.openstack.org/395454 | 19:00 |
*** zoli|wfh is now known as zoli|brb | 19:00 | |
*** abehl has quit IRC | 19:01 | |
*** dprince has joined #tripleo | 19:03 | |
*** jkilpatr_ has joined #tripleo | 19:04 | |
*** Guest13773 has joined #tripleo | 19:05 | |
*** jkilpatr has joined #tripleo | 19:05 | |
*** fragatina has joined #tripleo | 19:08 | |
*** yamahata has quit IRC | 19:16 | |
openstackgerrit | yolanda.robla proposed openstack/tripleo-quickstart: Create directories with root https://review.openstack.org/384892 | 19:17 |
gbarros | #tripleo, im using '%{hiera(''keystone_admin_api_vip'')}' in one of my env's but it does not seem to expand to anything. Any pointers? | 19:24 |
*** numans has quit IRC | 19:24 | |
*** chandankumar has quit IRC | 19:25 | |
*** ebalduf has quit IRC | 19:28 | |
*** zoli|brb is now known as zoli|wfh | 19:30 | |
*** liverpooler has quit IRC | 19:43 | |
*** jerrygb has joined #tripleo | 19:43 | |
*** openstackgerrit has quit IRC | 19:48 | |
*** openstackgerrit has joined #tripleo | 19:48 | |
openstackgerrit | Merged openstack/instack-undercloud: Increase Mistral Task Size limit https://review.openstack.org/348216 | 19:49 |
*** bana_k has joined #tripleo | 19:53 | |
*** yamahata has joined #tripleo | 19:55 | |
*** owalsh has joined #tripleo | 19:58 | |
openstackgerrit | Honza Pokorny proposed openstack/tripleo-specs: Add network diagram spec https://review.openstack.org/396383 | 20:01 |
*** cylopez has quit IRC | 20:04 | |
*** zoli|wfh is now known as zoli|gone | 20:04 | |
*** chem has quit IRC | 20:12 | |
*** gregwork has quit IRC | 20:16 | |
*** gregwork has joined #tripleo | 20:18 | |
*** ccamacho1 has joined #tripleo | 20:18 | |
*** ccamacho has quit IRC | 20:20 | |
*** bana_k has quit IRC | 20:21 | |
*** bana_k has joined #tripleo | 20:23 | |
*** panda|bbl is now known as panda | 20:24 | |
jidar | can you fit a HA deployment in 32 gigs? | 20:25 |
jidar | (I assume yes) | 20:25 |
slagle | yes, but it's tight | 20:35 |
slagle | i assume you're asking about for a virt environment | 20:35 |
*** rcernin has quit IRC | 20:36 | |
*** jcoufal has quit IRC | 20:38 | |
*** jcoufal has joined #tripleo | 20:40 | |
openstackgerrit | Michele Baldessari proposed openstack/tripleo-heat-templates: Fix race during major-upgrade-pacemaker step https://review.openstack.org/395460 | 20:41 |
*** mhenkel has quit IRC | 20:43 | |
*** dprince has quit IRC | 20:50 | |
jidar | yea, on the one box for testing/development | 20:50 |
jidar | I keep meaning to put together a box at home for this but I keep having to do prod deployments so it's not as important (*sigh*) | 20:50 |
*** Guest13773 has quit IRC | 20:51 | |
*** jkilpatr has quit IRC | 20:51 | |
trown | 32G is enough if you can do KSM to overcommit a bit | 20:51 |
*** jkilpatr_ has quit IRC | 20:51 | |
trown | I dont think it actually fits anymore otherwise | 20:52 |
bnemec | CI uses exactly 32 G worth of memory for an HA deployment right now - 4x6G overcloud nodes + 8G undercloud | 20:54 |
mwhahaha | what's the best way to get the puppet logs from a deploy? | 20:55 |
*** lblanchard has quit IRC | 20:55 | |
*** jkilpatr has joined #tripleo | 20:55 | |
trown | ya, in order to get 32G of VMs on a 32G host though, overcommit is required :) | 20:56 |
*** iranzo has joined #tripleo | 20:56 | |
*** iranzo has joined #tripleo | 20:56 | |
trown | mwhahaha: from a failed deploy or a successful one? | 20:56 |
mwhahaha | yes? | 20:56 |
trown | hehe... well a failed one is much eaiser `openstack stack failures list overcloud` | 20:57 |
mwhahaha | well that only gives me a snippet | 20:57 |
mwhahaha | i want it all | 20:57 |
mwhahaha | which is why i asked | 20:57 |
mwhahaha | (which should be a thing i might add) | 20:58 |
trown | indeed | 20:58 |
trown | I dont think it is though... you would need to look up each SoftwareDeployment resource and do resource show on it | 20:58 |
slagle | it's in the os-collect-config log on each node | 20:59 |
trown | that is probably nicer :) | 20:59 |
slagle | which is accessible via journalctl | 20:59 |
mwhahaha | https://review.openstack.org/#/c/388210/ | 20:59 |
thrash | mwhahaha: actually, log into the node you are interested in and 'journalctl -u os-collect-config' | 20:59 |
mwhahaha | no i just want the puppet logs | 20:59 |
mwhahaha | that's really a terrible thing to have to do | 21:00 |
slagle | yes, but you asked for the best way | 21:00 |
thrash | :D | 21:00 |
mwhahaha | no i said the best way, not the only way :) | 21:01 |
trozet | dsneddon: can you please set workflow on this https://review.openstack.org/#/c/396348/ ? | 21:01 |
mwhahaha | it just so happens that it's the only way right now it appears, that's totally not the best way :) | 21:01 |
*** jkilpatr has quit IRC | 21:03 | |
trozet | slagle, beagles: ^^^ | 21:07 |
stevebaker | mwhahaha: `openstack stack failures list --long overcloud` should give the full puppet output | 21:09 |
openstackgerrit | Sagi Shnaidman proposed openstack-infra/tripleo-ci: TEST: DONT RECHECK: periodic jobs https://review.openstack.org/359215 | 21:10 |
mwhahaha | stevebaker: thanks, that's a bit better :) | 21:11 |
sshnaidm | weshay, there is weird problem with periodics, they run each time on the same delorean hash and promote it | 21:11 |
stevebaker | mwhahaha: also /var/run/heat-config/deployed will have all puppet logs if you're running a new enough image | 21:11 |
sshnaidm | weshay, sent a patch to check what is going on there.. | 21:11 |
*** florianf has quit IRC | 21:13 | |
weshay | sshnaidm, k | 21:13 |
*** jayg is now known as jayg|g0n3 | 21:15 | |
trown | sshnaidm: that actually might be because there was a stalled "consistent" repo in RDO for a few days | 21:15 |
trown | it is fixed as of today though | 21:15 |
*** ayoung has quit IRC | 21:15 | |
*** jkilpatr has joined #tripleo | 21:16 | |
sshnaidm | trown, oh, that's it then | 21:17 |
sshnaidm | trown, because today it has different hash | 21:17 |
*** bana_k has quit IRC | 21:18 | |
trown | sshnaidm: ya, rally added a new dep that needed to be packaged | 21:18 |
*** iranzo has quit IRC | 21:19 | |
*** jkilpatr has quit IRC | 21:21 | |
panda | so upstream periodic jobs didn't promote anything even if they were all green for the past few days ? | 21:22 |
*** Goneri has quit IRC | 21:22 | |
trown | they promoted the same thing over and over :P | 21:24 |
tbarron | slagle: when you have a minute, stable/newton backport https://review.openstack.org/#/c/396183/ is ready for you | 21:25 |
openstackgerrit | Gabriele Cerami proposed openstack/tripleo-quickstart: Add configuration of HA IPv6 deployments https://review.openstack.org/380358 | 21:28 |
dsneddon | trozet, That review looked good, I +2, +1'd it. | 21:29 |
*** ccamacho1 has quit IRC | 21:31 | |
*** maticue has quit IRC | 21:32 | |
*** gfidente has quit IRC | 21:32 | |
*** bana_k has joined #tripleo | 21:37 | |
*** zaneb has quit IRC | 21:39 | |
tbarron | slagle: ty | 21:41 |
slagle | tbarron: yw | 21:48 |
*** jkilpatr has joined #tripleo | 21:50 | |
*** jkilpatr has joined #tripleo | 21:51 | |
*** ccamacho has joined #tripleo | 21:51 | |
*** owalsh has quit IRC | 21:53 | |
*** rhallisey has quit IRC | 21:57 | |
openstackgerrit | Ben Nemec proposed openstack/puppet-tripleo: Fix puppet lint failure https://review.openstack.org/396419 | 21:59 |
*** bana_k has quit IRC | 22:00 | |
*** panda has quit IRC | 22:01 | |
*** panda_ has joined #tripleo | 22:01 | |
*** jprovazn has quit IRC | 22:02 | |
*** lmiccini has quit IRC | 22:02 | |
openstackgerrit | Ben Nemec proposed openstack-infra/tripleo-ci: POC: Use tripleo-repos in tripleo.sh https://review.openstack.org/395813 | 22:03 |
*** panda_ has quit IRC | 22:05 | |
*** panda has joined #tripleo | 22:12 | |
*** jcoufal has quit IRC | 22:15 | |
*** trown is now known as trown|outtypewww | 22:15 | |
*** aufi has joined #tripleo | 22:17 | |
*** jerrygb_ has joined #tripleo | 22:21 | |
mwhahaha | bnemec: the lint failure slipped in because puppet-tripleo was using an older puppet-lint (via puppet-openstack_spec_helper) and we just merged the version increase. | 22:21 |
mwhahaha | bnemec: https://review.openstack.org/#/c/353488/ ccamacho forgot to recheck puppet-tripleo :o | 22:22 |
bnemec | mwhahaha: Okay, I suspected it was something like that but I don't know enough about how the puppet jobs work. :-) | 22:22 |
*** ccamacho has quit IRC | 22:22 | |
*** colonwq has quit IRC | 22:22 | |
*** jerrygb has quit IRC | 22:24 | |
*** jerrygb has joined #tripleo | 22:25 | |
*** jerrygb_ has quit IRC | 22:26 | |
*** tiswanso has quit IRC | 22:26 | |
*** jerrygb_ has joined #tripleo | 22:27 | |
*** jerrygb has quit IRC | 22:30 | |
*** jerrygb_ has quit IRC | 22:31 | |
bnemec | Huh, vlan-aware vms merged to neutron in newton? Cool. | 22:34 |
*** fragatin_ has joined #tripleo | 22:34 | |
* bnemec needs to rebuilt his ovb environment | 22:34 | |
bnemec | Or rebuild even | 22:34 |
*** fragatina has quit IRC | 22:37 | |
openstackgerrit | Merged openstack/tripleo-heat-templates: Fixes incorrect reference to OpendaylightApiNetwork https://review.openstack.org/396348 | 22:45 |
openstackgerrit | Merged openstack/tripleo-heat-templates: Select bootstrap node by list index not name https://review.openstack.org/396280 | 22:46 |
openstackgerrit | Ben Nemec proposed openstack/instack-undercloud: Nothing to see here https://review.openstack.org/386648 | 22:47 |
openstackgerrit | Merged openstack/tripleo-heat-templates: Add firewall rules for manila api service https://review.openstack.org/396183 | 22:51 |
*** aufi has quit IRC | 22:55 | |
*** mhenkel has joined #tripleo | 22:59 | |
*** limao has joined #tripleo | 23:01 | |
*** fragatin_ has quit IRC | 23:04 | |
*** fragatina has joined #tripleo | 23:05 | |
openstackgerrit | Pradeep Kilambi proposed openstack/puppet-tripleo: Fix barbican server name to not use aodh hiera https://review.openstack.org/396434 | 23:20 |
*** Alex___1234 has joined #tripleo | 23:22 | |
openstackgerrit | Pradeep Kilambi proposed openstack/puppet-tripleo: Add panko service support https://review.openstack.org/396435 | 23:22 |
*** ayoung has joined #tripleo | 23:22 | |
*** pmannidi has quit IRC | 23:23 | |
Alex___1234 | Hi Everyone. I'm attempting to deploy an overlcoud and getting errors with no available hosts found. I have two nodes that are supposed to be used but I think they are failing introspection. Does anyone have a few minutes to help me dig deeper into this? | 23:24 |
*** pmannidi has joined #tripleo | 23:24 | |
Alex___1234 | One thing I should mention is that all 3 parts (undercloud, controller node, and compute node) are VMs | 23:25 |
*** limao has quit IRC | 23:28 | |
*** fultonj has quit IRC | 23:35 | |
*** cmyster has quit IRC | 23:36 | |
openstackgerrit | Pradeep Kilambi proposed openstack/tripleo-heat-templates: Add panko api support to service templates https://review.openstack.org/396439 | 23:37 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!