ScoopIcicle | Our challenge is not to articulate the verifiable knowledge management. Our challenge is to broaden a specific self-efficacy. | 00:00 |
---|---|---|
sdake | is ScoopIcicle like a bot? | 00:00 |
portdirect | ? | 00:00 |
sbezverk | kfox1111: mariadb<-servicepackagename> | 00:01 |
sbezverk | but what is I want to call not mariadb but mariadb-007 ;) | 00:01 |
sbezverk | just because I already use mariadb name for something different? | 00:02 |
kfox1111 | just asking generically, that you want to be launchign stand alone mariadb? | 00:03 |
kfox1111 | not launching them as part of a bigger service package? | 00:03 |
*** ruhe has joined #openstack-kolla | 00:03 | |
kfox1111 | fasinating.... https://github.com/fission/fission | 00:04 |
sbezverk | kfox1111: right I want to launch 2 services of mariadb and give then two differnet names | 00:04 |
kfox1111 | yeah. then either --set globals.kolla.all.element_name or --values ~/mysettings.yaml would probably be the best we can do... | 00:05 |
kfox1111 | or... | 00:05 |
*** david-lyle has quit IRC | 00:05 | |
kfox1111 | if we are giving element_name its own search path anyway, | 00:05 |
portdirect | kfox1111: can we not do it through setting the release name? | 00:06 |
kfox1111 | we can root just that one var all the way to the top of global. | 00:06 |
kfox1111 | so, --set global.element_name=foo | 00:06 |
sbezverk | kfox1111: works for me | 00:06 |
kfox1111 | hmm... | 00:07 |
*** adrian_otto has quit IRC | 00:07 | |
kfox1111 | yeah, I thin kthat would work. | 00:07 |
kfox1111 | might have some weirdness if someone launches a greater service pckage like neutron, with mariadb/memcached enabled. | 00:08 |
portdirect | its how every other chart ive seen solves this question | 00:08 |
kfox1111 | shouldn't use global.element_name in that case, but can do it lower in the tree to still be unique. | 00:08 |
portdirect | (if they have attempted to) | 00:08 |
ScoopIcicle | The innovators gradually manage Balanced Scorecards. The policy makers whiteboard a capability. | 00:09 |
sbezverk | kfox1111: I still not clear how mariadb running for neutron will get its name | 00:10 |
kfox1111 | neutron's values.yaml would have globals.kolla.mariadb.all.element_name=mariadb_neutron ? | 00:11 |
kfox1111 | the user shouldn't have to worry about it then. | 00:11 |
*** krtaylor has joined #openstack-kolla | 00:12 | |
*** adrian_otto has joined #openstack-kolla | 00:12 | |
sdake | portdirect sbezverk I cleaned up the final version a bit of the talk abstrract | 00:12 |
sdake | portdirect sbezverk mind taking a look plz | 00:12 |
sdake | portdirect sbezverk i think we need to pick a title now :) | 00:13 |
sbezverk | sdake: sorry we are in the middle of the crisis | 00:13 |
portdirect | i think ScoopIcicle may be trying to help there? | 00:13 |
sdake | its only a crisis if the sky is falling | 00:14 |
sdake | or hell has frozen over | 00:14 |
sdake | that said, how can i help if at all | 00:14 |
sbezverk | kfox1111: right now we have .Values.element_name used as a reference | 00:14 |
kfox1111 | sbezverk: I think we just add an elementSearchPath = ":global.kolla.mariadb.statefulset:global.kolla.mariadb.all:global.kolla.all:global" and look up element_name in there. if not set, default to resourceName. | 00:14 |
kfox1111 | users can then set global.element_name and it will work standalone. | 00:15 |
sbezverk | it is not mariadb, rabbitmq has the same issue | 00:15 |
kfox1111 | service packages can set global.kolla.mariadb.all.element_name, and it will work from there. | 00:15 |
*** srwilkers has quit IRC | 00:15 | |
kfox1111 | sbezverk: same there. just s/mariadb/rabbitmq/ | 00:15 |
kfox1111 | mariadb/memcached/rabbitmq will have the issue. | 00:15 |
kfox1111 | as they are multi instance. | 00:16 |
kfox1111 | the same fix should work on all 3 I think. | 00:16 |
sbezverk | kfox1111: I think other services will have the same issue too, basically all services need .global.element_name | 00:16 |
kfox1111 | I'm still not convinced thats true. | 00:17 |
sbezverk | in values.yaml | 00:17 |
kfox1111 | will we be laucnhing multiple glance-api's in the same namespace? | 00:17 |
kfox1111 | or neutron-servers? | 00:17 |
kfox1111 | I think the only exception might be nova for cells v2. | 00:18 |
kfox1111 | though no one I know has ever deployed that. so kind of shooting in the dark there. | 00:18 |
sbezverk | kfox1111: I am talking about using alternative name for service | 00:18 |
kfox1111 | sbezverk: yeah. and I'm saying, why support that? | 00:18 |
sbezverk | I do not think it will work without that modification | 00:18 |
kfox1111 | what value does it add? it does add more complexity. | 00:18 |
portdirect | kfox1111: beacuse a lot of people will want to use that functionaility | 00:18 |
kfox1111 | portdirect: for what? | 00:19 |
sbezverk | kfox1111: why then we introduced variables for names? then we should have hardcoded names | 00:19 |
sbezverk | I think it is very useful | 00:19 |
kfox1111 | the names are hardcoded in a lot of places. | 00:19 |
sbezverk | if I already have mariadb running | 00:19 |
kfox1111 | to be able to rename your nova foo in kube get pods? | 00:19 |
sbezverk | and I do not want to change name for the exisiting | 00:19 |
portdirect | kfox1111: not everyone will want to call things the same names - they may have existing infra they want to ingergrate with (like yourself) or have other reasons | 00:20 |
kfox1111 | mariadb is an exception. so is rabbit and memchached. | 00:20 |
sbezverk | when I deploy kolla-kube service, I wold just use a different name | 00:20 |
*** adrian_otto has quit IRC | 00:20 | |
sbezverk | ok cinder | 00:20 |
kfox1111 | just saying, what is the benifit of being able to rename every other major openstack service in one namespace? | 00:20 |
kfox1111 | portdirect: no one will have existing infra in k8s though? | 00:20 |
sbezverk | kfox1111: what was the reason to introduce variables for names? | 00:20 |
portdirect | kfox1111: a lot will, like you do | 00:21 |
kfox1111 | sbezverk: specifically, so you could write one values.yaml fijle and share it with all service packages, and get the value scoping right. | 00:21 |
sbezverk | if you are saying that it is useless why we did it at the first place? | 00:21 |
kfox1111 | portdirect: I have an existing openstack. I don't have an openstack on k8s. | 00:21 |
portdirect | yes - but you want to tie into your esisting infra dont you? | 00:21 |
kfox1111 | this only affects naming of k8s object in a k8s namespace. | 00:21 |
kfox1111 | yes. | 00:21 |
*** adrian_otto has joined #openstack-kolla | 00:22 | |
kfox1111 | sbezverk: which is useless? | 00:22 |
kfox1111 | elmeent names on openstack services? | 00:22 |
sbezverk | to have names as variables for all services we offer under kolla-kube | 00:22 |
portdirect | kfox1111: will it not also affect dns resoltion of services? | 00:22 |
sbezverk | portdirect: yep good point | 00:23 |
kfox1111 | portdirect: its namespaced. | 00:23 |
sbezverk | it will | 00:23 |
*** hfu has joined #openstack-kolla | 00:23 | |
kfox1111 | the name will be unique within the namespace. | 00:23 |
*** awiddersheim has joined #openstack-kolla | 00:23 | |
kfox1111 | its only a problem if you want to run kolla-kubernetes in the same namespace as something else already consuming those names. | 00:23 |
sbezverk | kfox1111: we are not forcing to have a separate namespace for kolla | 00:23 |
sbezverk | so people could deploy it in default namesapce | 00:24 |
portdirect | we should also support having services accrossm multiple namespaces | 00:24 |
sbezverk | I think we need to be consistent | 00:24 |
portdirect | it is madness to insatll production openstack in a single k8s namespace | 00:24 |
kfox1111 | yup. but we never said that you could pick a deployment called nova-api and have kolla-kubernetes avoid that. | 00:24 |
sbezverk | if we allow specify custom element_name then it should be for all or for nothing | 00:24 |
sbezverk | better for all ;) | 00:24 |
kfox1111 | disagree. | 00:25 |
kfox1111 | thats like saying, because you might ever want to cluster mariadb, you should cluster everything the same exact way. | 00:25 |
kfox1111 | mariadb/memcached/rabbit have very good reasons to have multiple instances in the same namespace. | 00:25 |
kfox1111 | so the added complexity for them makes sense to me. | 00:26 |
kfox1111 | but for glance-api? what does that buy you? | 00:26 |
sbezverk | no it is like if you buy a car, then you can expect that it has an engine | 00:26 |
kfox1111 | good coding rule, "the code is done, not when there is nothing more to add, but nothing more to remove" | 00:27 |
portdirect | but what if you want/need your glance api to be called glance-region1-api? | 00:27 |
kfox1111 | portdirect: the only reason I see that beeing nessisary, is if you are running 2 regions in the same namespace. | 00:27 |
kfox1111 | something we origionally agreed not to support. | 00:27 |
kfox1111 | is that really a desired feature? | 00:27 |
portdirect | from my perspective this is about dns and tls, not namespaceing | 00:28 |
portdirect | I will be spreading all services accross namesapces | 00:29 |
sbezverk | kfox1111: as you said once we already have 99% of code, we just need to add 1 line in service's values | 00:29 |
kfox1111 | dns in k8s isn't exported outside the cluster. | 00:29 |
kfox1111 | glance-api.region1.cluster.local | 00:29 |
kfox1111 | glance-api.region2.cluster.local | 00:29 |
kfox1111 | sbezverk: I don't think all the support is in every package today. | 00:30 |
kfox1111 | I think its a lot more work then you think to fully flesh out. | 00:30 |
portdirect | now think about how host those services externally with a tls proected endpoint with certs that work everywere | 00:30 |
portdirect | you want the same hostname externally as you do internally | 00:30 |
kfox1111 | portdirect: its exported out via lb or ingress. | 00:31 |
kfox1111 | it can have literally any name at that point. | 00:31 |
portdirect | the domain is irrelevant | 00:31 |
portdirect | not if your running internal tls and want to keep your sanity | 00:31 |
portdirect | but anyway ive said enoough | 00:31 |
* portdirect out | 00:31 | |
sbezverk | we should probably have a wider discussion | 00:32 |
sbezverk | since I am still note convinced and would like to hear what other kolla-kube folks think about it | 00:32 |
kfox1111 | I don't understand the concern there. | 00:33 |
kfox1111 | portdirect: not trying to be stubbern here. just trying to understand. | 00:33 |
kfox1111 | +1 to wider discussion. | 00:34 |
kfox1111 | if its really a requirement, lets support it. | 00:34 |
sdake | are we tlaking about internal tls? | 00:34 |
*** adrian_otto has quit IRC | 00:34 | |
kfox1111 | but a non trivial amount of work should be done for a well defined reason I think. | 00:34 |
portdirect | sdake: touching on it | 00:34 |
kfox1111 | sdake: talking about should we support renaming openstack resources in k8s. | 00:35 |
sdake | multi-region would be cool | 00:35 |
sbezverk | kfox1111: I mean I still rememeber when we came up with idea of element_name, and the reason was to be able to give different names, without any specification that it is for 3 services only | 00:35 |
kfox1111 | and or multiregion in the same k8s namespace. | 00:35 |
*** sacharya has quit IRC | 00:35 | |
sdake | just readin gthru scrollback :) | 00:35 |
sbezverk | otherwise we would have just hardcoded names, why bother use variables?? | 00:35 |
*** inc0 has quit IRC | 00:35 | |
sdake | what openstack resources would be renamed? | 00:35 |
*** denggw has quit IRC | 00:36 | |
sdake | like keystone, etc? | 00:36 |
kfox1111 | again, the global values to me was about allowing one file to be written to let youconfigure all the things. | 00:36 |
kfox1111 | with the amount of genericness or specificity you want. | 00:36 |
kfox1111 | sdake: yeah. being able to rename keystone/nova/glance/etc. | 00:36 |
sbezverk | sdake: example cinder-us or glance-region-1 | 00:36 |
sdake | i guess I dont understand the problem in detail - and it should probablybe oepned up to a wider audience, not debated at EOD US time on a hump day :) | 00:36 |
*** adrian_otto has joined #openstack-kolla | 00:36 | |
sdake | that seems reasonable to me (to call something glance-region-X | 00:37 |
sdake | whats the downside? | 00:37 |
kfox1111 | sdake: amount of work needed to implement it. | 00:37 |
sbezverk | kfox1111: I think it is not that much since we already have variables in place | 00:38 |
sdake | work is all relative, the key quesiton is *when* | 00:38 |
kfox1111 | sbezverk: I'm saying, I don't think we have all the variables in place. | 00:38 |
kfox1111 | sdake: and by *who* | 00:38 |
sdake | yup :) | 00:38 |
sbezverk | kfox1111: we have NO object with hardcoded name | 00:38 |
kfox1111 | and who's going to write the tests. :) | 00:38 |
sdake | well when defined who in many cases | 00:38 |
sdake | defines who that is | 00:38 |
sdake | so here is what i'd suggest | 00:38 |
*** jtriley has joined #openstack-kolla | 00:39 | |
sdake | we stick to the game plan laid out in 0.5.0 | 00:39 |
kfox1111 | sbezverk: sure, lots... look for how many have printf in the search path. | 00:39 |
sbezverk | sdake: I can take care of that if community decides | 00:39 |
sdake | we got a long way to go until 1.0.0 | 00:39 |
sdake | ok then lets talk about downsides | 00:39 |
sdake | if the only downsides are "who will write the tests and who will implement the code" that is surmountable | 00:39 |
kfox1111 | and extra code to manage. | 00:39 |
sdake | those are that is :) | 00:40 |
kfox1111 | but I think thats all managable. | 00:40 |
kfox1111 | so, lets just say, I won't block the work. I just think its very low on the priority list. | 00:40 |
sdake | I'd really like a lazer focus on 0.5.0 personally - wish I had more time to do implementation work | 00:40 |
sdake | as far as testing goes in teh gate, we could just test the default names | 00:41 |
kfox1111 | sdake: whats not tested is broken. | 00:41 |
sdake | i think we hae bigger fish to fry in the gate then figuring how to test multiregion | 00:41 |
sdake | broken/tech preview ;-) | 00:41 |
kfox1111 | agreed. | 00:41 |
sdake | tesitng multiregion and even tls is very complex | 00:42 |
sdake | kolla-ansible has no tls tests for 2 years | 00:42 |
sdake | while tls works fantasstically well with kolla-ansible | 00:42 |
sdake | actually testing tls would be relatively easy | 00:42 |
sdake | anyway thats an aside, I think we decided we weren't targeting tls for pre-PTG | 00:42 |
sdake | and weren't really targeting multiregion for pre-PTG | 00:43 |
kfox1111 | I may tackle that sooner rather then later. | 00:43 |
sbezverk | kfox1111: when we have config management sorted out it will be easier to test | 00:43 |
kfox1111 | production without tls is not a good plan. | 00:43 |
sdake | if someone wnats to do any work at any time, its not like they have to do what they are told tho | 00:43 |
jascott1 | kfox1111 thanks for fission link | 00:43 |
sdake | sup jascott1 | 00:43 |
kfox1111 | jascott1: yeah, looks pretty cool. :) | 00:43 |
*** portdirect has left #openstack-kolla | 00:43 | |
jascott1 | whassup sdake | 00:43 |
sdake | in kola people do what they want, and we sort it out later ;) | 00:44 |
sdake | are there other downsides besides it takes a person to do the work and someone to wriet teh tets | 00:44 |
sdake | are there technical downsides? | 00:44 |
sdake | the above are not technical downsides - but rather timing downsides | 00:44 |
kfox1111 | just slightly more code and a bunch of touching files. | 00:45 |
kfox1111 | like, all the files. | 00:45 |
sdake | best time to touch a bunch of files, even all of them is early in a project | 00:45 |
kfox1111 | not nessisarily. | 00:45 |
sdake | kolla (the original repo) docker files were redesigned from scratch atleast 6-7 times | 00:45 |
sdake | its better then later :) | 00:45 |
kfox1111 | the best time to get important features in is early in the project too. | 00:45 |
kfox1111 | ;) | 00:45 |
kfox1111 | priorities and all. | 00:46 |
sdake | ya there is a duality there | 00:46 |
kfox1111 | orchestration I think will be more important to users then this. | 00:46 |
sdake | anyway if nobody objects to the work why is it a crisis? | 00:46 |
kfox1111 | not sure it really is. | 00:46 |
sdake | sbezverk crisis averted ;) | 00:46 |
*** gfidente has quit IRC | 00:46 | |
kfox1111 | but we haven't gotten a good explination of why some folks think its really important. | 00:47 |
sbezverk | sdake: not yet ;) | 00:47 |
kfox1111 | portdirect things it may affect internal tls. | 00:47 |
sdake | i heard multiregion - which makes perfect sense to me | 00:47 |
kfox1111 | but I don't see how. | 00:47 |
sdake | i heard tls, which we need | 00:47 |
sdake | internal tls - not sure how that is valuable | 00:47 |
sdake | kolla-ansible doesn't use internal tls | 00:47 |
sdake | we are aiming for parity with kolla-ansible in many ways on feature set | 00:48 |
kfox1111 | I think internal tls will be valueable, but still don't see how fixed naming will affect it. | 00:48 |
sdake | id' be happy with external tls | 00:48 |
sdake | here is the deal folks - code evolve | 00:48 |
sdake | s | 00:48 |
sdake | as long as we dont break apis as we go | 00:48 |
sdake | it doesn't matter if variables are static or dynamic | 00:49 |
sdake | we do want to be as flexible as possible | 00:49 |
sdake | that is what sets kolla apart from others | 00:49 |
kfox1111 | agreed. | 00:49 |
sdake | that is why we hvea gone to so much trouble to do the helm microservices, helm services, etc | 00:49 |
kfox1111 | we should e able to add this feature as an api addition at any time. | 00:49 |
sdake | kfox1111 i'd challenge you to prove that assertion :) | 00:50 |
kfox1111 | all I'm saying is I think there is a huge number of other things that should come first. thats all. just my opinioin. | 00:50 |
sdake | come befoer multiregion? if so, I agree | 00:50 |
sdake | like the service packages, like operators | 00:50 |
sdake | we need to get our fundametal architecture off the ground | 00:51 |
kfox1111 | before supporing object renaming of all objects. | 00:51 |
kfox1111 | I do think we need configmap renaming soon. | 00:51 |
kfox1111 | but being able to rename the deployment object for glance-api doesn't seem terribly important. | 00:51 |
kfox1111 | if someone can explain why that may be problem though, I'm totally open to change my priorities. | 00:52 |
sdake | sbezverk my recommendation wuld be to start a ml thread | 00:52 |
sbezverk | sdake: well this issue got surfaced while working on the service package | 00:52 |
*** hfu has quit IRC | 00:52 | |
sdake | sbezverk plz expand | 00:52 |
kfox1111 | sbezverk: surfaced on a multiinstance package. | 00:52 |
*** harlowja has joined #openstack-kolla | 00:52 | |
kfox1111 | I totally agree it needs to be fixed, asap on those. | 00:52 |
sdake | sbezverk if its blocking service packages, we need to fix that part at minimum | 00:53 |
kfox1111 | mariadb/memcached/rabbitmq. | 00:53 |
kfox1111 | the rest, not convinced need it right away. | 00:53 |
sdake | where doth portdirect go | 00:53 |
kfox1111 | the same fix could be applied to the rest, its just a lot more work. | 00:53 |
*** hfu has joined #openstack-kolla | 00:54 | |
*** sayantani01 has joined #openstack-kolla | 00:55 | |
sbezverk | {{- $elementService := .Values.element_name }} | 00:56 |
sbezverk | {{- $searchPath := printf ":global.kolla.%s.svc:global.kolla.%s.all:global.kolla.rabbitmq.svc:global.kolla.rabbitmq.all:global.kolla.all" $elementName $elementName }} | 00:56 |
sbezverk | irony | 00:56 |
sbezverk | element_name is used to build search path | 00:57 |
kfox1111 | yes. | 00:57 |
kfox1111 | it works when launched as a microservice. | 00:57 |
sbezverk | when itself needed to be found in the path | 00:57 |
kfox1111 | just not as a service, without specifying them all indivdiualy. | 00:57 |
kfox1111 | didn't predict that use case. :/ | 00:57 |
kfox1111 | my cristal ball is pretty good, but not quite that good aparanetly. :/ | 00:57 |
kfox1111 | I figured only the openstack service packages would use it, | 00:58 |
kfox1111 | so mariadb.element_name would work form within the service package. | 00:58 |
kfox1111 | I did not think we'd put an intermediate mariadb wrapper around it. | 00:58 |
kfox1111 | though that makes perfect sense. | 00:59 |
sbezverk | kfox1111: ok, for sure that would be hard to forsee.. | 01:00 |
*** cuongnv has joined #openstack-kolla | 01:00 | |
sbezverk | now the solution | 01:00 |
*** zhurong has joined #openstack-kolla | 01:00 | |
kfox1111 | so, I think we just need to do $elementSearchPath := ":global.kolla.rabbitmq.svc:global.kolla.rabbitmq.all:global.kolla.all:global" | 01:00 |
kfox1111 | search for element_name in there. | 01:01 |
kfox1111 | if not set, elmenet_name = resourceName, | 01:01 |
kfox1111 | and then continue building searchPath as normal. | 01:01 |
sbezverk | and the service's values.yaml will have at the second level element_name | 01:02 |
kfox1111 | service package could set global.kolla.rabbitmq.all.element_name and it would work. | 01:02 |
kfox1111 | mariadb service package could use global.element_name and it would work. | 01:03 |
kfox1111 | or mariadb.mariadb-init-element.element_name would work too. lots of ways to set it very specifically if you need to. | 01:04 |
kfox1111 | srwilkers_: https://review.openstack.org/#/c/422147/ looks ready? | 01:04 |
*** duonghq_ is now known as duonghq | 01:04 | |
kfox1111 | heh. if gerrit didn't crash. | 01:04 |
*** salv-orlando has joined #openstack-kolla | 01:05 | |
sbezverk | kfox1111: but would it not break ability to insert neutron's search path before? | 01:06 |
kfox1111 | each microservcie has its own search path. | 01:07 |
sbezverk | so instead of searching in rabbitmq or mariadb branch, brach from neutron would be prefered since it is | 01:07 |
sbezverk | before the "general" service path | 01:07 |
*** qiliang28 has joined #openstack-kolla | 01:07 | |
*** qiliang27 has quit IRC | 01:08 | |
*** qiliang28 is now known as qiliang27 | 01:08 | |
kfox1111 | that was intended to be handled by making the rabbit branch completely on its own for element_named things. | 01:08 |
kfox1111 | so if you launch mariadb-nova, | 01:08 |
duonghq | morning guys | 01:09 |
kfox1111 | there is a whole tree of configurability at global.kolla.mariadb-nova.... | 01:09 |
kfox1111 | but if you want to set a setting that affects all mariadb, you can set it up a level at global.kolla.mariadb... | 01:09 |
*** salv-orlando has quit IRC | 01:10 | |
*** sayantani01 has quit IRC | 01:11 | |
*** sayantani01 has joined #openstack-kolla | 01:12 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Optimize reconfigure action for nova https://review.openstack.org/407416 | 01:16 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Detect openstack_release variable when it auto https://review.openstack.org/422324 | 01:16 |
Jeffrey4l | morning | 01:16 |
sbezverk | ok need to take off, will start tomorrow morning, bye folks | 01:16 |
sdake | sup Jeffrey4l | 01:17 |
sdake | night sbezverk | 01:17 |
sdake | sup duonghq | 01:17 |
duonghq | night sbezverk | 01:17 |
Jeffrey4l | nigh sbezverk | 01:17 |
Jeffrey4l | hey sdake | 01:17 |
kfox1111 | k. have a good one. | 01:17 |
Jeffrey4l | sdake, check this, which solve the pbr issue. | 01:18 |
Jeffrey4l | https://review.openstack.org/422324 | 01:18 |
*** zhubingbing has joined #openstack-kolla | 01:18 | |
sdake | Jeffrey4l tht looks cool what about pip installs? | 01:19 |
Jeffrey4l | sdake, it will return the exact kolla-ansible version you installed. | 01:19 |
sdake | with the import pbr thing? | 01:20 |
Jeffrey4l | pbr is installed and required when installing kolla-ansible. | 01:20 |
sdake | Jeffrey4l does it work as is? | 01:20 |
sdake | if so looks like a winner | 01:20 |
Jeffrey4l | it is in the kolla-ansible's requirements.txt file too. | 01:20 |
sdake | we can remove all the nonsense in the docs about workflows for devs vs ops | 01:20 |
Jeffrey4l | yep. | 01:21 |
Jeffrey4l | hrm. Let me test install kolla-ansible from pypi. it should works. | 01:21 |
sdake | Jeffrey4l wish I would have thought of that | 01:21 |
Jeffrey4l | auto vs latest? which is better? | 01:22 |
sdake | auto seems better | 01:22 |
*** tovin07 has joined #openstack-kolla | 01:22 | |
sdake | latest = the latest bits available from dockerhub | 01:22 |
sdake | auto = current bits | 01:22 |
sdake | maybe change auto to current | 01:22 |
Jeffrey4l | ok. | 01:22 |
sdake | Jeffrey4l tbh not sure- probably needs another reviewer to look at it for the best name | 01:23 |
sdake | or copious documentation in the config file about what it means | 01:24 |
sdake | auto is super ambiguous | 01:24 |
sdake | current is also slightly ambiguous | 01:24 |
Jeffrey4l | or just leave it empty? or just comment all openstack_release variable with empty value? | 01:25 |
Jeffrey4l | install from pip works. | 01:26 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 01:27 |
*** user__ has joined #openstack-kolla | 01:29 | |
*** saneax is now known as saneax-_-|AFK | 01:33 | |
sdake | dunno but I gotta roll for awhile | 01:37 |
sdake | Jeffrey4l ide arocks | 01:37 |
sdake | Jeffrey4l not sure what to name it | 01:37 |
sdake | Jeffrey4l idea that is | 01:37 |
Jeffrey4l | thanks ;) | 01:37 |
sdake | Jeffrey4l i'd suggest sending an emali to the ml asking for advice on naming convention | 01:38 |
sdake | although I don't know if thta will turn anything up | 01:38 |
sdake | or ping the core reviewers in EMEA when they start to appear later today | 01:38 |
Jeffrey4l | ok. | 01:38 |
sdake | i'll likely be back in a few hours - need to eat | 01:38 |
*** sdake has quit IRC | 01:38 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for rally https://review.openstack.org/421893 | 01:39 |
*** tovin07_ has joined #openstack-kolla | 01:40 | |
*** denggw has joined #openstack-kolla | 01:41 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 01:42 |
*** sacharya has joined #openstack-kolla | 01:46 | |
*** denggw- has joined #openstack-kolla | 01:47 | |
*** tovin07 has quit IRC | 01:50 | |
*** denggw has quit IRC | 01:51 | |
*** adrian_otto has quit IRC | 01:51 | |
*** schwicht has joined #openstack-kolla | 01:55 | |
ScoopIcicle | Measurements deliver maximum impact, while an integrativeness prioritizes the partners. Our resilient platforms influence the project leader. Leading-edge decision making is all about visual thinking. The group envisions images. Inventory-planning delivery frameworks incentivise the sales manager, while the senior support staff reinvests in our usage-based and progressive corporate governance. | 01:56 |
openstackgerrit | jimmygc proposed openstack/kolla: Use install-pip macro in most source images https://review.openstack.org/413861 | 01:58 |
*** salv-orlando has joined #openstack-kolla | 02:06 | |
*** unicell has quit IRC | 02:08 | |
*** rhallisey has quit IRC | 02:09 | |
*** salv-orlando has quit IRC | 02:11 | |
user__ | what is the difference between kolla and stackanetes? why would i chose one or the other? | 02:13 |
*** tovin07 has joined #openstack-kolla | 02:13 | |
*** user__ has left #openstack-kolla | 02:15 | |
*** ipsecguy_ has joined #openstack-kolla | 02:18 | |
*** ipsecguy has quit IRC | 02:19 | |
openstackgerrit | Cao Xuan Hoang proposed openstack/kolla-kubernetes: Remove support for py33 https://review.openstack.org/422342 | 02:26 |
openstackgerrit | Cuong Nguyen proposed openstack/kolla-kubernetes: Improve file handling with context manager https://review.openstack.org/422343 | 02:27 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Push images all the time https://review.openstack.org/422344 | 02:27 |
*** tovin07_ has quit IRC | 02:31 | |
*** goldyfruit has joined #openstack-kolla | 02:31 | |
*** awidders_ has joined #openstack-kolla | 02:32 | |
*** jascott1_ has joined #openstack-kolla | 02:33 | |
*** sacharya_ has joined #openstack-kolla | 02:33 | |
*** tonanhngo_ has joined #openstack-kolla | 02:34 | |
*** berendt_ has joined #openstack-kolla | 02:34 | |
*** smekel__ has joined #openstack-kolla | 02:34 | |
*** tovin07_ has joined #openstack-kolla | 02:35 | |
*** sayantan_ has joined #openstack-kolla | 02:38 | |
*** Aju has joined #openstack-kolla | 02:40 | |
*** denggw has joined #openstack-kolla | 02:41 | |
*** nihilifer has quit IRC | 02:41 | |
*** tovin07 has quit IRC | 02:41 | |
*** sacharya has quit IRC | 02:41 | |
*** sayantani01 has quit IRC | 02:41 | |
*** awiddersheim has quit IRC | 02:41 | |
*** krtaylor has quit IRC | 02:41 | |
*** tonanhngo has quit IRC | 02:41 | |
*** jascott1 has quit IRC | 02:41 | |
*** smekel_ has quit IRC | 02:41 | |
*** hyakuhei has quit IRC | 02:41 | |
*** wirehead_ has quit IRC | 02:41 | |
*** afranc has quit IRC | 02:41 | |
*** Kdecherf has quit IRC | 02:41 | |
*** berendt has quit IRC | 02:41 | |
*** saneax-_-|AFK has quit IRC | 02:41 | |
*** kklimonda has quit IRC | 02:41 | |
*** pron has quit IRC | 02:41 | |
*** abc__ has quit IRC | 02:41 | |
*** smekel__ is now known as smekel_ | 02:42 | |
*** kklimonda has joined #openstack-kolla | 02:42 | |
*** berendt_ is now known as berendt | 02:43 | |
*** denggw- has quit IRC | 02:44 | |
openstackgerrit | zhubingbing proposed openstack/kolla: Add vitrage images https://review.openstack.org/417593 | 02:47 |
*** Kdecherf has joined #openstack-kolla | 02:50 | |
*** pron has joined #openstack-kolla | 02:50 | |
*** krtaylor has joined #openstack-kolla | 02:50 | |
*** abc has joined #openstack-kolla | 02:51 | |
*** denggw- has joined #openstack-kolla | 02:52 | |
zhubingbing | hello guys | 02:53 |
*** sdake has joined #openstack-kolla | 02:54 | |
*** denggw has quit IRC | 02:56 | |
openstackgerrit | Duong Ha-Quang proposed openstack/kolla-kubernetes: Memcached service package https://review.openstack.org/421650 | 02:57 |
*** denggw has joined #openstack-kolla | 02:57 | |
duonghq | hi zhubingbing | 02:57 |
zhubingbing | hi duonghq | 02:58 |
*** denggw has quit IRC | 03:00 | |
*** denggw has joined #openstack-kolla | 03:00 | |
*** Jeffrey4l_ has joined #openstack-kolla | 03:00 | |
*** denggw- has quit IRC | 03:00 | |
*** hyakuhei has joined #openstack-kolla | 03:01 | |
*** Jeffrey4l has quit IRC | 03:04 | |
*** wirehead_ has joined #openstack-kolla | 03:04 | |
*** goldyfruit has quit IRC | 03:05 | |
*** denggw has quit IRC | 03:07 | |
*** denggw has joined #openstack-kolla | 03:07 | |
*** denggw has quit IRC | 03:09 | |
*** denggw has joined #openstack-kolla | 03:09 | |
*** saneax-_-|AFK has joined #openstack-kolla | 03:10 | |
*** sdake has quit IRC | 03:13 | |
*** denggw has quit IRC | 03:14 | |
*** sdake has joined #openstack-kolla | 03:15 | |
*** MasterOfBugs has quit IRC | 03:18 | |
sdake | duonghq your patch looks good | 03:18 |
sdake | is that port address correct for the gate? | 03:18 |
*** adrian_otto has joined #openstack-kolla | 03:19 | |
*** pramodrj07 has quit IRC | 03:19 | |
openstackgerrit | zhubingbing proposed openstack/kolla: Fix designate dockerfile https://review.openstack.org/422366 | 03:20 |
duonghq | sdake, the mariadb is fine, so I think the memcached one should fine | 03:20 |
sdake | yup just checked it out | 03:21 |
sdake | it looks good to me | 03:21 |
sdake | hows the gate look? | 03:21 |
*** nihilifer has joined #openstack-kolla | 03:21 | |
duonghq | the centos-binary-3-ceph-multi-nv and centos-binary-2-helm-entrypoint-nv still fail | 03:21 |
duonghq | the binary-3 failed with error Cannot overwrite digest sha256 | 03:22 |
duonghq | the other reported that 2017-01-18 15:04:53.629974 | containers failed to start. | 03:22 |
duonghq | http://logs.openstack.org/50/421650/3/experimental/gate-kolla-kubernetes-deploy-centos-binary-2-helm-entrypoint-nv/d19d6d9/console.html#_2017-01-18_15_04_53_629974 | 03:22 |
duonghq | sdake, I'm trying to rebase and recheck these tests | 03:23 |
duonghq | but seem that it doesn't relate to memcached portion | 03:23 |
openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes: Test of gate https://review.openstack.org/422368 | 03:25 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Push images all the time https://review.openstack.org/422344 | 03:26 |
duonghq | sdake, seem that your ps does not trigger test and experimental test? | 03:28 |
*** adrian_otto has quit IRC | 03:29 | |
*** kbaegis has joined #openstack-kolla | 03:34 | |
*** zhurong has quit IRC | 03:34 | |
*** adrian_otto has joined #openstack-kolla | 03:35 | |
*** prameswar has joined #openstack-kolla | 03:38 | |
*** prameswar has quit IRC | 03:43 | |
*** MasterOfBugs has joined #openstack-kolla | 03:47 | |
*** pramodrj07 has joined #openstack-kolla | 03:47 | |
*** adrian_otto has quit IRC | 03:54 | |
*** mdnadeem has joined #openstack-kolla | 03:55 | |
*** sacharya_ has quit IRC | 03:57 | |
*** adrian_otto has joined #openstack-kolla | 03:57 | |
*** jtriley has quit IRC | 03:58 | |
*** adrian_otto has quit IRC | 04:04 | |
*** sdake has quit IRC | 04:06 | |
*** sdake has joined #openstack-kolla | 04:07 | |
*** sudipto_ has joined #openstack-kolla | 04:07 | |
*** sudipto has joined #openstack-kolla | 04:07 | |
*** salv-orlando has joined #openstack-kolla | 04:07 | |
*** jtriley has joined #openstack-kolla | 04:08 | |
*** salv-orlando has quit IRC | 04:11 | |
openstackgerrit | zhang kun proposed openstack/kolla-ansible: Fix Bootstraping the mongodb replication set https://review.openstack.org/422377 | 04:16 |
*** caowei has joined #openstack-kolla | 04:19 | |
*** sp___ has joined #openstack-kolla | 04:20 | |
*** adrian_otto has joined #openstack-kolla | 04:22 | |
openstackgerrit | Merged openstack/kolla-kubernetes: Memcached service package https://review.openstack.org/421650 | 04:24 |
openstackgerrit | Merged openstack/kolla-kubernetes: Fix missing yum install -y in gate. https://review.openstack.org/422147 | 04:32 |
*** absubram has joined #openstack-kolla | 04:33 | |
*** kbaegis has quit IRC | 04:33 | |
openstackgerrit | Steven Dake proposed openstack/kolla-kubernetes: GATECHECK https://review.openstack.org/422380 | 04:35 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: install murano dashboard package in horizon https://review.openstack.org/395957 | 04:37 |
*** adrian_otto has quit IRC | 04:38 | |
*** adrian_otto has joined #openstack-kolla | 04:40 | |
openstackgerrit | Merged openstack/kolla-kubernetes: Fix typo in loopback file setup https://review.openstack.org/421204 | 04:41 |
sdake | sbezverk can you ack this https://review.openstack.org/#/c/421786/1 | 04:42 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for cinder https://review.openstack.org/422384 | 04:43 |
openstackgerrit | Merged openstack/kolla-kubernetes: Change 'docker' to 'Docker' in docs https://review.openstack.org/408626 | 04:47 |
*** sudipto_ has quit IRC | 04:49 | |
*** sudipto has quit IRC | 04:49 | |
*** krtaylor has quit IRC | 04:52 | |
*** sdake_ has joined #openstack-kolla | 04:56 | |
*** sacharya has joined #openstack-kolla | 04:57 | |
*** sdake has quit IRC | 04:59 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 05:00 |
*** absubram has quit IRC | 05:03 | |
*** sacharya has quit IRC | 05:03 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Rabbitmq service deployment with entrypoint https://review.openstack.org/422229 | 05:07 |
*** jtriley has quit IRC | 05:10 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Changing mariadb bootstart job name to init-element https://review.openstack.org/422389 | 05:12 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for cinder https://review.openstack.org/422384 | 05:16 |
*** adrian_otto1 has joined #openstack-kolla | 05:16 | |
*** adrian_otto has quit IRC | 05:16 | |
*** saneax-_-|AFK is now known as saneax | 05:24 | |
*** sudipto has joined #openstack-kolla | 05:27 | |
*** sudipto_ has joined #openstack-kolla | 05:27 | |
*** jtriley has joined #openstack-kolla | 05:35 | |
openstackgerrit | Wei Cao proposed openstack/kolla: Add dragonflow container https://review.openstack.org/418193 | 05:38 |
*** adrian_otto1 has quit IRC | 05:39 | |
*** adrian_otto has joined #openstack-kolla | 05:43 | |
*** adrian_otto has quit IRC | 05:43 | |
*** sdake_ has quit IRC | 05:47 | |
*** leseb has quit IRC | 05:48 | |
*** leseb has joined #openstack-kolla | 05:53 | |
*** skramaja has joined #openstack-kolla | 05:58 | |
*** sacharya has joined #openstack-kolla | 05:58 | |
*** unicell has joined #openstack-kolla | 05:59 | |
*** jtriley has quit IRC | 06:03 | |
*** sacharya has quit IRC | 06:03 | |
*** salv-orlando has joined #openstack-kolla | 06:08 | |
*** Sajauddin has joined #openstack-kolla | 06:10 | |
*** salv-orlando has quit IRC | 06:13 | |
Sajauddin | Hi | 06:15 |
Sajauddin | I have register a blue print "Kolla on Power server i.e., ppc64le architecture compatible" | 06:15 |
Sajauddin | It is still not approved, priority and directions are not yet assigned | 06:17 |
Sajauddin | Can anyone tell me what is the procedure to take this forward | 06:17 |
*** zhangqiankun has joined #openstack-kolla | 06:20 | |
*** jtriley has joined #openstack-kolla | 06:28 | |
duonghq | Sajauddin, you should discuss with sdake, inc0 or Jeffrey4l_ | 06:32 |
Jeffrey4l_ | Sajauddin, could u post the bp link? | 06:33 |
*** jtriley has quit IRC | 06:33 | |
*** jtriley has joined #openstack-kolla | 06:37 | |
*** nihilifer has quit IRC | 06:42 | |
Sajauddin | @Jeffrey4l: https://blueprints.launchpad.net/kolla/+spec/kolla-on-power-server | 06:46 |
Sajauddin | please check and let me know | 06:46 |
Jeffrey4l_ | Sajauddin, seem cool. approved. | 06:49 |
Jeffrey4l_ | Sajauddin, one question, have u ever thought adding Power gate for kolla? | 06:52 |
Sajauddin | Jeffrey4l, could u please elaborate about Power Gate. I dint get, what you are pointing to? | 06:55 |
openstackgerrit | Zeyu Zhu proposed openstack/kolla-kubernetes: Fix the issue in multi-node.rst https://review.openstack.org/421054 | 06:56 |
*** nihilifer has joined #openstack-kolla | 06:57 | |
openstackgerrit | Zeyu Zhu proposed openstack/kolla-kubernetes: Fix the issue in multi-node.rst https://review.openstack.org/421054 | 06:59 |
*** jascott1_ has quit IRC | 07:00 | |
*** zhurong has joined #openstack-kolla | 07:07 | |
*** jtriley has quit IRC | 07:10 | |
*** jascott1 has joined #openstack-kolla | 07:12 | |
*** tgoto has joined #openstack-kolla | 07:14 | |
tgoto | Hello, I'm having a queer behavior at building instance. | 07:15 |
tgoto | kolla-build has error as follows. | 07:15 |
tgoto | ERROR:kolla.image.build.base:Unknown error when building Traceback (most recent call last): File "/usr/lib/python2.7/site-packages/kolla/image/build.py", line 408, in builder for response in self.dc.build(path=image.path, File "/usr/lib/python2.7/site-packages/kolla/image/build.py", line 131, in dc self._dc = docker.Client(version='auto', **docker_kwargs) AttributeError: 'module' object has no attribute 'Client' | 07:15 |
tgoto | Does this error caused by docker-py library? I reckon this error persists from 3 days ago. | 07:16 |
tgoto | Thank you. | 07:16 |
*** tgoto has quit IRC | 07:18 | |
*** lrensing has joined #openstack-kolla | 07:23 | |
zhangkun_ | Client | 07:23 |
*** masber has quit IRC | 07:30 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 07:31 |
*** tgoto has joined #openstack-kolla | 07:32 | |
*** lrensing has quit IRC | 07:38 | |
*** zhangqiankun has quit IRC | 07:38 | |
tgoto | I was using latest docker binary, I'll try downgrading binary for that.Thank you for comment > zhangkun_ | 07:39 |
*** zhangqiankun has joined #openstack-kolla | 07:40 | |
*** sayantan_ has quit IRC | 07:41 | |
*** zhangqiankun has quit IRC | 07:42 | |
*** zhangqiankun has joined #openstack-kolla | 07:42 | |
*** bmace has quit IRC | 07:46 | |
*** bmace has joined #openstack-kolla | 07:46 | |
zhangkun_ | tgoto, just believe youself,you can solve your problem | 07:49 |
tgoto | well, you are right, zhangun_. I downgraded my docker binary, and it worked! | 07:50 |
tgoto | Thank you ! | 07:50 |
tgoto | I was using latest docker binary from "Docker Master Binaries" for docker client use. It was the root cause. | 07:51 |
*** matrohon has joined #openstack-kolla | 07:55 | |
openstackgerrit | Surya Prakash Singh proposed openstack/kolla: Code coverage improvement for kolla https://review.openstack.org/422444 | 07:56 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 07:57 |
*** sacharya has joined #openstack-kolla | 08:00 | |
*** haplo37_ has quit IRC | 08:00 | |
*** jtriley has joined #openstack-kolla | 08:01 | |
Jeffrey4l_ | Sajauddin, run test in the openstack gerrit. | 08:01 |
Jeffrey4l_ | for example, build the images/try to deploy openstack. | 08:01 |
*** haplo37_ has joined #openstack-kolla | 08:03 | |
*** tgoto has quit IRC | 08:04 | |
openstackgerrit | desheng proposed openstack/kolla: Modify punctuation https://review.openstack.org/422450 | 08:05 |
*** sacharya has quit IRC | 08:06 | |
*** salv-orlando has joined #openstack-kolla | 08:09 | |
*** tonanhngo_ has quit IRC | 08:09 | |
*** tonanhngo has joined #openstack-kolla | 08:11 | |
*** tonanhngo has quit IRC | 08:12 | |
*** salv-orlando has quit IRC | 08:14 | |
*** msimonin has joined #openstack-kolla | 08:14 | |
openstackgerrit | Cuong Nguyen proposed openstack/kolla-ansible: Use yaml.safe_load() instead of yaml.load() https://review.openstack.org/422460 | 08:20 |
*** jtriley has quit IRC | 08:31 | |
*** openstackgerrit has quit IRC | 08:33 | |
*** egonzalez has joined #openstack-kolla | 08:34 | |
*** masber has joined #openstack-kolla | 08:34 | |
*** zhurong has quit IRC | 08:35 | |
egonzalez | morning | 08:42 |
*** gfidente has joined #openstack-kolla | 08:42 | |
*** gfidente has quit IRC | 08:42 | |
*** gfidente has joined #openstack-kolla | 08:42 | |
zhangkun_ | good afternoon | 08:45 |
*** mnasiadka has joined #openstack-kolla | 08:47 | |
*** JeroenBo has joined #openstack-kolla | 08:48 | |
*** JeroenBo has left #openstack-kolla | 08:48 | |
*** JeroenBo has joined #openstack-kolla | 08:50 | |
Sajauddin | Jeffrey4l_, ok. will test by deploying openstack using Kolla | 08:52 |
duonghq | morning egonzalez | 08:52 |
Jeffrey4l_ | Sajauddin, just throw this question. i have no idea about this. openstack infra do not provide ppc64 node, imo. | 08:53 |
*** jtriley has joined #openstack-kolla | 09:00 | |
*** hfu has quit IRC | 09:02 | |
*** shardy_afk is now known as shardy | 09:03 | |
*** hfu has joined #openstack-kolla | 09:03 | |
*** salv-orlando has joined #openstack-kolla | 09:03 | |
*** rmart04 has joined #openstack-kolla | 09:08 | |
*** satyar has joined #openstack-kolla | 09:09 | |
*** jtriley has quit IRC | 09:09 | |
satyar | Hi jeffrey4l | 09:09 |
satyar | Hi jeffrey4l_ | 09:09 |
Jeffrey4l_ | sup satyar | 09:10 |
satyar | How are you | 09:10 |
Jeffrey4l_ | not bad ;) | 09:10 |
satyar | :P | 09:10 |
*** zhangkun_ is now known as hime | 09:11 | |
satyar | getting an issue with the current code | 09:11 |
satyar | http://paste.openstack.org/show/595570/ | 09:11 |
*** hime is now known as zhangkun_ | 09:11 | |
satyar | while deployment | 09:12 |
Jeffrey4l_ | satyar, your version of docker client ( docker-py ) do not match docker server. | 09:13 |
Jeffrey4l_ | satyar, Bad Request (\"client is newer than server (client API version: 1.24, server API version: 1.22)\") | 09:13 |
Jeffrey4l_ | either downgrade you client version or upgrade you docker engine. | 09:13 |
satyar | its same 1.22 | 09:13 |
satyar | docker --version | 09:14 |
satyar | Docker version 1.10.3, build 57bf6fd | 09:14 |
Jeffrey4l_ | check the docker-py version on the node. | 09:15 |
*** jascott1 has quit IRC | 09:15 | |
Jeffrey4l_ | it is a AIO deployment? | 09:15 |
satyar | no multinode | 09:15 |
*** openstackgerrit has joined #openstack-kolla | 09:15 | |
openstackgerrit | qiankun.zhang proposed openstack/kolla: Install tacker-horizon into horizon image https://review.openstack.org/407810 | 09:15 |
Jeffrey4l_ | check the target nodes' docker-py version. | 09:15 |
satyar | -bash: docker-py: command not found | 09:19 |
Jeffrey4l_ | satyar, pip freeze | grep docker-py | 09:19 |
satyar | docker-py==1.10.6 | 09:21 |
satyar | docker-pycreds==0.2.1 | 09:21 |
Jeffrey4l_ | docker engine version? satyar | 09:21 |
Jeffrey4l_ | 1.10.6 use 1.24 api version. i think you need downgrade the docker-py or upgrade docker engine. | 09:23 |
satyar | Sorry jeffrey4l_ whats the command to check that | 09:25 |
Jeffrey4l_ | satyar, "rpm -qa | grep docker" or "dpkg -l | grep docker" | 09:25 |
*** athomas has joined #openstack-kolla | 09:25 | |
satyar | rpm -qa | grep docke | 09:25 |
satyar | docker-engine-1.10.3-1.0.3.el7.x86_64 | 09:25 |
satyar | do i need to upgrade the engine? | 09:27 |
Jeffrey4l_ | i think so. but i have no idea which version match the docker-py client. ;( | 09:28 |
satyar | i did a yum upgrade | 09:28 |
satyar | docker-engine-1.12.6-1.0.1.el7.x86_64 | 09:28 |
Jeffrey4l_ | ya. that should work | 09:29 |
Jeffrey4l_ | satyar, check here https://docs.docker.com/engine/reference/api/docker_remote_api/ | 09:29 |
satyar | it seems 1.10 work with 1.22 | 09:30 |
satyar | 1.12 requires 1.24 | 09:30 |
Jeffrey4l_ | but your docker-py use 1.24 as default. | 09:30 |
Jeffrey4l_ | your docker-py is higher. | 09:31 |
*** msimonin has quit IRC | 09:31 | |
openstackgerrit | qiankun.zhang proposed openstack/kolla-ansible: Enable tacker_horizon when enable_tacker is yes https://review.openstack.org/407822 | 09:32 |
*** haplo37 has quit IRC | 09:32 | |
*** hyakuhei has quit IRC | 09:36 | |
*** hyakuhei has joined #openstack-kolla | 09:36 | |
*** hyakuhei has quit IRC | 09:36 | |
*** hyakuhei has joined #openstack-kolla | 09:36 | |
openstackgerrit | Cuong Nguyen proposed openstack/kolla-ansible: Close file-like object after used https://review.openstack.org/422504 | 09:37 |
*** haplo37 has joined #openstack-kolla | 09:41 | |
openstackgerrit | qiankun.zhang proposed openstack/kolla-ansible: Enable tacker_horizon when enable_tacker is yes https://review.openstack.org/407822 | 09:41 |
*** jascott1 has joined #openstack-kolla | 09:42 | |
*** msimonin has joined #openstack-kolla | 09:56 | |
*** Serlex has joined #openstack-kolla | 09:56 | |
*** tovin07_ has quit IRC | 09:57 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for cloudkitty https://review.openstack.org/422516 | 09:58 |
*** jrich523_ has joined #openstack-kolla | 10:04 | |
*** jrich523 has quit IRC | 10:07 | |
*** cuongnv has quit IRC | 10:11 | |
*** jascott1 has quit IRC | 10:14 | |
*** sambetts|afk is now known as sambetts | 10:17 | |
*** satyar has quit IRC | 10:17 | |
*** sdake has joined #openstack-kolla | 10:18 | |
berendt | o/ | 10:19 |
*** salv-orl_ has joined #openstack-kolla | 10:23 | |
*** msimonin1 has joined #openstack-kolla | 10:25 | |
*** masber has quit IRC | 10:26 | |
*** salv-orlando has quit IRC | 10:26 | |
*** msimonin has quit IRC | 10:27 | |
*** zhurong has joined #openstack-kolla | 10:30 | |
*** msimonin1 has quit IRC | 10:32 | |
*** nihilifer has quit IRC | 10:33 | |
*** kbaegis has joined #openstack-kolla | 10:33 | |
sdake | sup brad[] | 10:35 |
sdake | berendt :) | 10:35 |
sdake | morning folks | 10:35 |
kbaegis | Morning sdake | 10:36 |
sdake | sup kbaegis | 10:36 |
sdake | ever get gentoo working? :) | 10:36 |
*** nihilifer has joined #openstack-kolla | 10:36 | |
sdake | hey nihilifer | 10:36 |
openstackgerrit | Surya Prakash Singh proposed openstack/kolla: Code coverage improvement for kolla https://review.openstack.org/422444 | 10:36 |
kbaegis | sdake: Yes, actually. Switched to the 4.x ansible tree though and having to reinvent the wheel :) | 10:36 |
nihilifer | sdake: o/ | 10:37 |
*** zhangkun_ is now known as zkun | 10:41 | |
*** masber has joined #openstack-kolla | 10:41 | |
berendt | sdake: i missed the team meeting yesterday, nobody on the ops list replied to the debian deprecation pre warning mail, is it fine to open the vote on openstack-dev or should i wait for the next team meeting? | 10:42 |
sdake | berendt the process we followed for the last deprecation (which is documented somewhere in the follows:standard deprecation policy in the tc's repo | 10:43 |
sdake | notify ops ist, open vote from core review team, submit review with link to ops discussion in the gerrit review log or in thecommit log | 10:43 |
berendt | ok | 10:43 |
berendt | than i will open the vote to move this topic forward | 10:43 |
sdake | if the ops list pushes back anywhere in that chain, we have more work to do | 10:43 |
berendt | nobody replied on the ops list... | 10:44 |
sdake | either not do the deprecation | 10:44 |
sdake | or find something that works for ops | 10:44 |
sdake | berendt ack - obedient silent consent :) | 10:44 |
sdake | its sort of like when intel announced they would stop making memory | 10:45 |
sdake | all their customers were like "meh who cares" | 10:46 |
sdake | (back in the early 90s) | 10:46 |
berendt | lol | 10:46 |
sdake | although given your eperiencey ou probably remember th epentium floating point bug :) | 10:46 |
sdake | that almost cratered intel as a company | 10:46 |
openstackgerrit | Wei Cao proposed openstack/kolla-ansible: Add freezer ansible role https://review.openstack.org/402224 | 10:47 |
sdake | they recovered quite nicesely from it | 10:47 |
sdake | even though in reality nobody really needed that floating point functionality | 10:47 |
sdake | and by nobody, i mean 90% of the population :) | 10:47 |
sdake | or more | 10:47 |
sdake | obviously scientists needed it | 10:47 |
sdake | although I think scientists use bignum libraries | 10:47 |
*** DanyC has joined #openstack-kolla | 10:47 | |
*** caowei has quit IRC | 10:48 | |
*** pbourke has quit IRC | 10:48 | |
*** pbourke has joined #openstack-kolla | 10:49 | |
berendt | how long do we vote? 7 days? | 10:49 |
sdake | yup 7 days | 10:50 |
sdake | should jus tbe the kolla deliverable core team i htink | 10:50 |
sdake | although our policy is fuzzy atm | 10:50 |
berendt | i wrote "Kolla core reviewers" because i was not sure about it | 10:50 |
sdake | thats fine | 10:50 |
sdake | we just won't count kolla-kubernetes core reviewers for the sake of quorum | 10:51 |
berendt | sdake: http://paste.openstack.org/show/595595/ | 10:51 |
berendt | can you check? | 10:51 |
sdake | I think inc0 had suggested he had more reorg to do on the policies | 10:51 |
sdake | berendt looking | 10:51 |
sdake | berendt in th emial i'd recommend stating the voting deadline | 10:52 |
*** nihilifer has quit IRC | 10:52 | |
berendt | "The vote will be open for 7 days."" | 10:52 |
sdake | i mean in terms of the calendar day | 10:52 |
berendt | The vote will be open for 7 days (26.01.2017). | 10:52 |
sdake | 7 days, until | 10:52 |
sdake | yup | 10:52 |
sdake | sounds good :) | 10:52 |
berendt | i will send a link to the openstack-oeprators list, this way they can take care of it, too | 10:53 |
*** nihilifer has joined #openstack-kolla | 10:54 | |
sdake | berendt you are #19 on commits for the lifetime of opensstack | 10:56 |
sdake | berendt nice job ! :) | 10:56 |
berendt | haha | 10:56 |
*** DanyC has quit IRC | 10:56 | |
sdake | i'm 52 | 10:56 |
sdake | Jeffrey4l_ is 112 | 10:56 |
sdake | prtty cool alot of our core review team is in the top 5% | 10:57 |
sdake | btw, we talked about deprecating fedora and its supposed to deprecate this cycle | 10:57 |
sdake | did that hit the repo? | 10:57 |
berendt | sdake: https://review.openstack.org/#/c/359447/ | 10:58 |
berendt | i think so | 10:58 |
sdake | berendt to be brutally honest I am struggling for time to keep up with kolla and kolla-ansible deliverables | 10:58 |
sdake | berendt nice you did that ;) | 10:58 |
sdake | good work dude | 10:58 |
sdake | berendt anything we can do to reduce our support footprint is good - as long as it doesn't impact operators | 10:58 |
berendt | i think so, yes | 10:59 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Deprecate Debian based Docker images https://review.openstack.org/417874 | 11:00 |
openstackgerrit | Christian Berendt proposed openstack/kolla: Add short README file for contrib/template-override directory https://review.openstack.org/422552 | 11:04 |
*** masber has quit IRC | 11:06 | |
berendt | has anybody used the demos in contrib/demos (kolla repo) the last months? | 11:06 |
egonzalez | no | 11:09 |
*** zkun has quit IRC | 11:13 | |
*** zhugaoxiao has joined #openstack-kolla | 11:14 | |
*** zhubingbing has quit IRC | 11:20 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 11:24 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 11:24 |
openstackgerrit | Christian Berendt proposed openstack/kolla-ansible: Remove contrib directory https://review.openstack.org/422559 | 11:25 |
berendt | do we use the documentation inside kolla-ansible? | 11:26 |
*** athomas has quit IRC | 11:28 | |
openstackgerrit | Christian Berendt proposed openstack/kolla-ansible: Remove specs directory https://review.openstack.org/422562 | 11:29 |
*** zhugaoxiao has quit IRC | 11:29 | |
*** zhugaoxiao has joined #openstack-kolla | 11:29 | |
*** msimonin has joined #openstack-kolla | 11:32 | |
*** athomas has joined #openstack-kolla | 11:35 | |
ScoopIcicle | Solution-oriented, bottom-up, unique and market-driving strategic staircases diligently influence the Senior Director of Client Leadership. The thinkers/planners make it possible. Genuine system is all about solutions-based say/do ratio. | 11:36 |
openstackgerrit | Nenad Radojevic proposed openstack/kolla-ansible: Fix placeholder for 'service_auth' section in neutron_lbaas.conf https://review.openstack.org/422566 | 11:37 |
openstackgerrit | Christian Berendt proposed openstack/kolla-ansible: Cleanup docs https://review.openstack.org/422568 | 11:38 |
*** sudipto_ has quit IRC | 11:39 | |
*** sudipto has quit IRC | 11:39 | |
*** salv-orl_ has quit IRC | 11:42 | |
*** sp___ has quit IRC | 11:47 | |
*** satyar has joined #openstack-kolla | 11:56 | |
*** msimonin has quit IRC | 11:59 | |
*** msimonin has joined #openstack-kolla | 11:59 | |
*** msimonin has quit IRC | 12:00 | |
*** rhallisey has joined #openstack-kolla | 12:00 | |
*** msimonin has joined #openstack-kolla | 12:00 | |
*** msimonin has quit IRC | 12:01 | |
*** msimonin has joined #openstack-kolla | 12:01 | |
*** JoseMello has joined #openstack-kolla | 12:01 | |
*** msimonin has quit IRC | 12:01 | |
sdake | berendt the heat demo i use often | 12:02 |
*** msimonin has joined #openstack-kolla | 12:02 | |
sdake | berendt the magnum demo, not so much | 12:02 |
*** msimonin has quit IRC | 12:02 | |
sdake | ScoopIcicle must be a bot ;) | 12:03 |
*** zhurong has quit IRC | 12:05 | |
satyar | Hi jeffrey4l_ | 12:05 |
satyar | hi sdake | 12:05 |
sdake | sup satyar | 12:05 |
satyar | dueto recent change i am getting an issue with the kolla deployment | 12:06 |
satyar | the issue is at kolla_toolbox.py | 12:06 |
satyar | http://paste.openstack.org/show/595607/ | 12:06 |
*** awidders_ has quit IRC | 12:08 | |
egonzalez | satyar: can you use docker ps in the node is failing? | 12:08 |
satyar | nope | 12:09 |
*** krtaylor has joined #openstack-kolla | 12:10 | |
egonzalez | have you upgrade/downgraded to match both versions? | 12:10 |
sdake | egonzalez that error is caused when docker-py is too new for the server | 12:10 |
sdake | egonzalez although without a docker ps, its ipossible to tell what went wrong | 12:10 |
egonzalez | sdake: yeah, just to check if the issue is with kolla_toolbox or his docker version | 12:11 |
satyar | http://paste.openstack.org/show/595608/ | 12:12 |
*** tonanhngo has joined #openstack-kolla | 12:13 | |
*** tonanhngo has quit IRC | 12:14 | |
egonzalez | satyar: maybe have some other version cached or not correctly removed (from rpm/deb and pip) | 12:18 |
satyar | i created a new vm from which i am deploying and reimaged the servers | 12:19 |
*** masber has joined #openstack-kolla | 12:20 | |
*** kbaegis has quit IRC | 12:33 | |
*** shardy is now known as shardy_lunch | 12:34 | |
*** jtriley has joined #openstack-kolla | 12:34 | |
*** Sajauddin has quit IRC | 12:36 | |
openstackgerrit | Surya Prakash Singh proposed openstack/kolla: Code coverage improvement for kolla https://review.openstack.org/422444 | 12:39 |
*** srwilkers has joined #openstack-kolla | 12:45 | |
ScoopIcicle | We will go the extra mile to calibrate an on-message, enterprise-wide and executive-level win-win solution. As a result, the product manager targets our technologies. | 12:46 |
*** dave-mccowan has joined #openstack-kolla | 12:48 | |
*** srwilkers has quit IRC | 12:48 | |
*** saneax is now known as saneax-_-|AFK | 12:49 | |
*** msimonin has joined #openstack-kolla | 12:52 | |
*** msimonin1 has joined #openstack-kolla | 12:53 | |
*** srwilkers has joined #openstack-kolla | 12:54 | |
sdake | morning srwilkers | 12:55 |
sdake | up early :) | 12:55 |
*** msimonin has quit IRC | 12:56 | |
*** mgiles has joined #openstack-kolla | 13:00 | |
sdake | sup mgiles | 13:01 |
mgiles | morning sdake | 13:01 |
sdake | mgiles will you be making it to the PTG? | 13:01 |
mgiles | no | 13:01 |
sdake | mgiles boston summit? | 13:01 |
mgiles | Probably will be in Boston since it's close (I'm in NH) | 13:02 |
sdake | nice - NH alot like AZ :) | 13:02 |
mgiles | just the same, I'm sure. :) | 13:02 |
sdake | although AZ quiclky turning into AZ | 13:02 |
sdake | rather CA | 13:02 |
*** schwicht has quit IRC | 13:03 | |
openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Nova service deployment https://review.openstack.org/421457 | 13:09 |
sdake | mgiles re nova, would it be possible to split that review into two parts - one for compute and one for control? | 13:14 |
mgiles | I can do so if it would be helpful, although I think both are working at this point | 13:15 |
sdake | maybe even 3 parts - one for the microservices change | 13:15 |
sdake | mgiles ya - it just helps reviewers understand the change a little more easily | 13:15 |
sdake | and helps other people writing code copy your work | 13:15 |
sdake | the big thing is it helps other people copy your code easily or learn from it if its a unit | 13:16 |
sdake | you can use git rebase -i to edit a stack of patches | 13:16 |
sdake | if you do go down that route, keep the change ID intact on one of the patches (probably nova-controller) | 13:16 |
mgiles | Sure, I can split it out | 13:16 |
sdake | thanks! | 13:16 |
sdake | mgiles note I have a review to promote many experimental gate jobs to nonvoting jobs | 13:17 |
sdake | so check experiemntal will soon be a thing we wont have to do every patch | 13:17 |
mgiles | That'd make life a little simpler | 13:17 |
srwilkers | morning sdake -- im up at 4a CST every day ;) | 13:18 |
sdake | mgiles yup - and i'll stay on top of that going forward as eople add gate jobs | 13:19 |
sdake | (or do my best atleast, pretty busy with dayjob at present) | 13:19 |
*** shardy_lunch is now known as shardy | 13:20 | |
*** schwicht has joined #openstack-kolla | 13:25 | |
*** schwicht has quit IRC | 13:31 | |
*** jheroux has joined #openstack-kolla | 13:33 | |
*** satyar has quit IRC | 13:37 | |
*** schwicht has joined #openstack-kolla | 13:39 | |
sdake | i see the service level packages are making good progress | 13:40 |
sdake | so I amarked it as such | 13:40 |
*** sudipto has joined #openstack-kolla | 13:43 | |
*** sudipto_ has joined #openstack-kolla | 13:43 | |
*** salv-orlando has joined #openstack-kolla | 13:43 | |
*** sudipto has quit IRC | 13:45 | |
*** sudipto_ has quit IRC | 13:45 | |
*** saneax-_-|AFK is now known as saneax | 13:47 | |
*** schwicht has quit IRC | 13:51 | |
*** goldyfruit has joined #openstack-kolla | 13:58 | |
*** salv-orlando has quit IRC | 14:05 | |
*** salv-orlando has joined #openstack-kolla | 14:05 | |
ScoopIcicle | The key to landscape is breakthrough. Our validation generates highly satisfactory efficiencies as part of the plan. The key representatives potentiate our executions. The white-collar workers accelerate the strategy at the individual, team and organizational level, whereas a systematized requirement invigorates the brand manager. A non-deterministic, siloed, openness deepens our fine-grained business philosophy, whereas | 14:05 |
*** fguillot has joined #openstack-kolla | 14:05 | |
openstackgerrit | Nenad Radojevic proposed openstack/kolla-ansible: Add default roles used by Barbican https://review.openstack.org/422643 | 14:08 |
sbezverk | srwilkers: since you are up, could you check this short one? https://review.openstack.org/#/c/422389/ | 14:11 |
*** mdnadeem has quit IRC | 14:11 | |
*** mkoderer has joined #openstack-kolla | 14:14 | |
*** schwicht has joined #openstack-kolla | 14:15 | |
*** kelv has quit IRC | 14:16 | |
*** jtriley has quit IRC | 14:16 | |
mliima | morrrrrrrrrrrrrrrrrrrrrging guys | 14:16 |
mliima | \o | 14:16 |
egonzalez | mliima: wow woken up powerful, morning btw | 14:17 |
mliima | hehe | 14:18 |
srwilkers | sbezverk, of course | 14:25 |
sbezverk | srwilkers: thanks | 14:27 |
srwilkers | sbezverk, np man :) | 14:27 |
openstackgerrit | Dan Prince proposed openstack/kolla: Add nova-placement-api packages https://review.openstack.org/422185 | 14:28 |
*** satyar has joined #openstack-kolla | 14:28 | |
*** srwilkers has quit IRC | 14:28 | |
*** JeroenBo has left #openstack-kolla | 14:28 | |
sdake | mliima morning :) | 14:29 |
mliima | hey sdake :) | 14:30 |
openstackgerrit | Merged openstack/kolla-kubernetes: Changing mariadb bootstart job name to init-element https://review.openstack.org/422389 | 14:30 |
dmsimard | Hey guys, genuine question. What's the value in printing a literal dict to stdout with "kolla-build --list-dependencies" ? http://paste.openstack.org/raw/595628/ | 14:30 |
dmsimard | It's hard for a user to read, it's not easily machine-readable either .. Why not a table ? Or JSON ? Or anything else ? | 14:30 |
sdake | dmsimard i pretty much thought a json structure would be better too | 14:31 |
*** srwilkers has joined #openstack-kolla | 14:31 | |
sdake | dmsimard the dict is probably printed out for processing by other tools | 14:32 |
dmsimard | I'm trying to build my own tool around it and the dict output is horribly inconvenient | 14:32 |
dmsimard | :) | 14:32 |
sdake | dmsimard i hear ya | 14:33 |
dmsimard | Maybe the kolla CLI could leverage cliff (same framework as openstackclient) and then you could output to json, yaml and etc natively | 14:33 |
sdake | dmsimard I think there are likely soem products that depend on that functionality in the marketplace already | 14:33 |
sdake | dmsimard so i'm not sure changing it to a different formatting would make sense - perhaps a flag to specify format type tho would work | 14:33 |
sdake | dmsimard that sounds pretty good | 14:34 |
sdake | dmsimard i did write a tool which converted the dict into a python dictionary that coudl be used (iirc) | 14:34 |
sdake | dmsimard im not sure what I did with teh code | 14:34 |
sdake | dmsimard i'll hunt for it if you want to see it - its like 20 lines | 14:34 |
egonzalez | Jeffrey4l_: have you looked at gates error? linux init cgroups path: mkdir /docker: not a directory. | 14:36 |
sdake | ok gotta jet for next 1 hr while i drag my kid out of bed | 14:38 |
* sdake enjoys | 14:38 | |
*** srwilkers has quit IRC | 14:38 | |
*** jtriley has joined #openstack-kolla | 14:40 | |
kfox1111 | morning. | 14:41 |
*** srwilkers has joined #openstack-kolla | 14:42 | |
kfox1111 | srwilkers: hey | 14:42 |
srwilkers | hey kfox1111 :) hows it going man? | 14:44 |
kfox1111 | pretty good. just couldn't sleep. :/ | 14:44 |
kfox1111 | you? | 14:44 |
srwilkers | that sucks :( i also couldnt sleep last night. been a very busy week | 14:45 |
srwilkers | think my bloods probably 60% coffee at this point | 14:45 |
*** saneax is now known as saneax-_-|AFK | 14:46 | |
kfox1111 | sorry you couldn't also. | 14:46 |
kfox1111 | hehe. yeah. I get that. :) | 14:46 |
srwilkers | i saw the TC approved the new language process for projects | 14:47 |
dmsimard | sdake: http://paste.openstack.org/raw/595630/ | 14:48 |
*** breitz has quit IRC | 14:50 | |
sbezverk | kfox1111: moring.. please check this ps https://review.openstack.org/#/c/422229/4 it seems satisfy microservice and service packages.. | 14:50 |
*** breitz has joined #openstack-kolla | 14:50 | |
sdake | dmsimard looks good ;-) | 14:50 |
sbezverk | srwilkers: so when go is go in openstack ;-) | 14:51 |
srwilkers | that was going to be something i wanted to discuss us considering at some point ;) | 14:52 |
kfox1111 | sbezverk, without the element name in the search path, you can't write settings that specifically target individual rabbits from within one config file. | 14:52 |
srwilkers | i think its going to be an important topic at the PTG, especially wrt operators | 14:52 |
kfox1111 | also, you wanted global.element_name to work? | 14:53 |
kfox1111 | I think those two features mean the search path for the element_name needs to be differen then the search path for everything else? | 14:53 |
*** yingjun has joined #openstack-kolla | 14:59 | |
*** v1k0d3n has joined #openstack-kolla | 15:00 | |
*** tonanhngo has joined #openstack-kolla | 15:01 | |
*** mnasiadka has quit IRC | 15:02 | |
*** salv-orlando has quit IRC | 15:02 | |
*** tonanhngo has quit IRC | 15:03 | |
sbezverk | kfox1111: that was what I was suspecting, what if we have only one %s in the path | 15:03 |
sbezverk | and then hardcoded rabbitmq? | 15:03 |
sbezverk | in this case another service can insert element_name for rabbitmq before the rabbitmq own check | 15:04 |
*** TxGirlGeek has joined #openstack-kolla | 15:06 | |
sbezverk | {{- $searchPath := printf ":global.kolla.%s.all:global.kolla.rabbitmq.svc:global.kolla.rabbitmq.all:global.kolla.all" }} $elementName | 15:06 |
Jeffrey4l_ | hi egonzalez, no yet. | 15:06 |
sbezverk | in this case if element_name available, printf generates good search path and rabbitmq name can be found in branch | 15:07 |
sbezverk | kfox1111: if not available, then we fall back to element name defined under rabbitmq own branch | 15:08 |
kfox1111 | sbezverk: need global.kolla.%s.svc too | 15:08 |
kfox1111 | no, for microservices. | 15:08 |
kfox1111 | if you launch just rabbitmq-statefulset, it shoudl not rquire the user to specify elmenet_name. | 15:09 |
kfox1111 | so rabbitmq-statefulset: in all_values shoudl have global.element_name = rabbitmq | 15:09 |
openstackgerrit | Merged openstack/kolla: Modify punctuation https://review.openstack.org/422450 | 15:09 |
kfox1111 | same with the others. | 15:09 |
sbezverk | kfox1111: right I was just presenting the concept | 15:12 |
kfox1111 | k. I was a bit confused. I thought you were talking about values.yaml in the service level package. | 15:13 |
*** yangzb09 has joined #openstack-kolla | 15:13 | |
*** adrian_otto has joined #openstack-kolla | 15:14 | |
openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Nova-control service deployment https://review.openstack.org/421457 | 15:15 |
openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Nova-compute service deployment https://review.openstack.org/422688 | 15:16 |
*** schwicht has quit IRC | 15:17 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for congress https://review.openstack.org/422690 | 15:17 |
*** schwicht has joined #openstack-kolla | 15:18 | |
*** kristian__ has joined #openstack-kolla | 15:20 | |
*** MasterOfBugs has quit IRC | 15:23 | |
*** pramodrj07 has quit IRC | 15:23 | |
*** srwilkers__ has joined #openstack-kolla | 15:25 | |
*** narasimha_SV has joined #openstack-kolla | 15:26 | |
narasimha_SV | http://paste.openstack.org/show/595642/ - anyone facing this error | 15:27 |
narasimha_SV | my docker version is 1.10.3 | 15:27 |
narasimha_SV | docker-py version is 1.10.6 | 15:27 |
*** srwilkers has quit IRC | 15:28 | |
narasimha_SV | I am facing this issue since yesterday's master branch kolla code | 15:28 |
*** sacharya has joined #openstack-kolla | 15:30 | |
*** adrian_otto has quit IRC | 15:31 | |
narasimha_SV | http://paste.openstack.org/show/595644/ | 15:34 |
narasimha_SV | when i try the same code in kolla_toolbox.py in kolla-ansible library code manually | 15:35 |
narasimha_SV | i am getting the same error | 15:35 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for rabbitmq https://review.openstack.org/422705 | 15:44 |
*** lrensing has joined #openstack-kolla | 15:47 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix ironic inspector https://review.openstack.org/380774 | 15:57 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Fix ironic inspector https://review.openstack.org/420122 | 15:58 |
*** shardy has quit IRC | 16:00 | |
*** shardy has joined #openstack-kolla | 16:01 | |
*** salv-orlando has joined #openstack-kolla | 16:03 | |
*** adrian_otto has joined #openstack-kolla | 16:03 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Rabbitmq service deployment with entrypoint https://review.openstack.org/422229 | 16:05 |
*** david-lyle has joined #openstack-kolla | 16:06 | |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfigure action for horizon https://review.openstack.org/422719 | 16:09 |
*** rhallisey has quit IRC | 16:14 | |
*** chopmann has joined #openstack-kolla | 16:14 | |
*** rhallisey has joined #openstack-kolla | 16:14 | |
*** chopmann has quit IRC | 16:15 | |
*** tonanhngo has joined #openstack-kolla | 16:15 | |
*** yangzb09 has quit IRC | 16:17 | |
*** yangzb09 has joined #openstack-kolla | 16:18 | |
*** kelv has joined #openstack-kolla | 16:18 | |
*** denggw has joined #openstack-kolla | 16:19 | |
*** nihilifer has quit IRC | 16:22 | |
*** absubram has joined #openstack-kolla | 16:22 | |
*** nihilifer has joined #openstack-kolla | 16:23 | |
*** inc0 has joined #openstack-kolla | 16:23 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: [MASTER][TEST][KOLLA-ANSIBLE] https://review.openstack.org/422734 | 16:24 |
*** salv-orl_ has joined #openstack-kolla | 16:24 | |
inc0 | leeeeeroy jenkins! | 16:24 |
sdake | inc0 no shit ;) | 16:26 |
*** salv-orlando has quit IRC | 16:27 | |
kfox1111 | was there an update? | 16:27 |
srwilkers__ | what are we leroying about? | 16:28 |
*** denggw has quit IRC | 16:31 | |
*** rmart04 has quit IRC | 16:31 | |
*** denggw has joined #openstack-kolla | 16:32 | |
*** kristian__ has quit IRC | 16:33 | |
*** kristian__ has joined #openstack-kolla | 16:34 | |
*** salv-orl_ has quit IRC | 16:34 | |
sdake | no idea | 16:36 |
sdake | apparently inc0 has just started playing world of warcraft | 16:36 |
*** sayantani01 has joined #openstack-kolla | 16:37 | |
kfox1111 | anyone awake that has ever done self signed certs with openstack? | 16:37 |
sdake | kfox1111 dave-mccowan is your dude | 16:37 |
satyar | i love warcraft :) | 16:37 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Optimize reconfiguration for tacker https://review.openstack.org/422036 | 16:37 |
sdake | kfox1111 athere is a rockin playbook that creates selfsigned certs in kolla-ansible btw | 16:37 |
kfox1111 | I've only done it with regular certs. | 16:37 |
srwilkers__ | just started? late to the party | 16:37 |
sdake | srwilkers__ ya about 35 or 40 levels | 16:38 |
sdake | srwilkers__ lthough I dont' play anymore since it wasn't healthy for me :) | 16:38 |
*** kristian__ has quit IRC | 16:38 | |
sdake | kfox1111 i've got 10 mins before my son's assembly, i can piont you at the code or answer q's you may have | 16:38 |
srwilkers__ | sdake: i quit playing right before i started getting involved with kolla-k8s. wanted to replace the hours i was wasting on that game with something productive ;p | 16:38 |
sdake | srwilkers__ at one point in my life i played 35 hrs/week | 16:39 |
sdake | or more | 16:39 |
sdake | top 100 guild | 16:39 |
srwilkers__ | sdake: yep, same | 16:39 |
sdake | this was before there were raid guides and videos and stuff | 16:39 |
kfox1111 | sdake: just wondering how you load the ca certificate into the various client side of the services (nova/neutron) | 16:39 |
srwilkers__ | sdake: back when molten core was a weekend long ordeal | 16:39 |
kfox1111 | ie, so nova can talk to neutron's endpoint, etc. | 16:39 |
sdake | kfox1111 in kolla-ansible, we use haproxy to proxy the certificates | 16:39 |
kfox1111 | yeah. | 16:39 |
sdake | kfox1111 internally we don't support TLS | 16:39 |
kfox1111 | but nova has to talk through haproxy to... | 16:39 |
kfox1111 | ah. ok. | 16:39 |
sdake | kfox1111 nah we use an internal non-tls endpoint | 16:40 |
kfox1111 | was wondering how to do it with tls everywhere. | 16:40 |
sdake | i think that would be an evolutionary thing | 16:40 |
sdake | but something to discuss cross-deliverable during the ptg | 16:40 |
*** unicell has quit IRC | 16:40 | |
kfox1111 | easier to do it everywhere with the setup thats there. | 16:40 |
kfox1111 | unless you break haproxy up into two frontends. | 16:41 |
sdake | kfox1111 kolla-ansible is the best place to make improvements there, as well as fernet rotation | 16:41 |
sdake | and then we can semi-copy the implementation | 16:41 |
sdake | seems like a fine blueprint for pike | 16:41 |
*** denggw- has joined #openstack-kolla | 16:41 | |
kfox1111 | sdake: kolla-kubernetes is doing it differently, as it has kubernetes netowkring tools available. | 16:42 |
sdake | it took us 1 year of ansible before we even tackled tls | 16:42 |
kfox1111 | I want to tackle fernet token at some point too. we use it on our clouds. | 16:42 |
*** SatyaP has joined #openstack-kolla | 16:42 | |
sdake | yup | 16:42 |
kfox1111 | am currently thinking maybe cephfs would be a good option for backing it. | 16:42 |
sdake | my point is we have about what i consider 3 months of a larger community of developers with a common goal now | 16:42 |
kfox1111 | then the keystone pods can just mount it, and use a scheduledjob object to rotate. | 16:43 |
sdake | and we have 6 months of prototyping work that happened prior | 16:43 |
kfox1111 | we'd need a cephfs container and a ceph-mds container in kolla though. | 16:43 |
sdake | kolla had 1.5 years of protyping work | 16:43 |
sdake | it really took kolla about 2-2.5 years to implement tls externally and fernet rotation | 16:43 |
kfox1111 | + aggregate experience of devs. ;) | 16:43 |
sdake | lots o folks know how it works | 16:43 |
*** denggw- has quit IRC | 16:43 | |
sdake | i even know how the ansible parts work :) | 16:44 |
*** sayantani01 has quit IRC | 16:44 | |
sdake | although I think we have more pressing matters | 16:44 |
kfox1111 | more pressing then what? | 16:44 |
kfox1111 | fernet? | 16:44 |
*** sayantani01 has joined #openstack-kolla | 16:44 | |
sdake | yes like finishing the service layer for example | 16:44 |
kfox1111 | my goal's to get a production system running on kolla. | 16:44 |
sdake | although if you want to get fernet rolling the code is all there to use in contianers already | 16:44 |
kfox1111 | that requires stuff like tls/fernet. | 16:44 |
kfox1111 | service packages, not so much. | 16:45 |
kfox1111 | how is kolla-ansible doing it? | 16:45 |
sdake | doing which - tls? | 16:45 |
sdake | or fernet | 16:45 |
kfox1111 | fernet. | 16:45 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Rabbitmq service deployment with entrypoint https://review.openstack.org/422229 | 16:45 |
*** denggw has quit IRC | 16:45 | |
kfox1111 | just using ansible to push out the rolled changes? | 16:45 |
sdake | I can't explain it now (don't have time) but I'll be back later | 16:45 |
kfox1111 | k. | 16:45 |
sdake | kfox1111 no, we have a container that rotates the tokens | 16:45 |
sdake | and uses ssh iirc to do rotation | 16:45 |
kfox1111 | how does it push to the other nodes? | 16:45 |
kfox1111 | ah. ok. | 16:45 |
sdake | kfox1111 alot of this has already been handled by kolla-ansible | 16:46 |
inc0 | sdake, started? | 16:46 |
inc0 | noooo... | 16:46 |
sdake | kfox1111 no need to reinvent it :) | 16:46 |
kfox1111 | that might work well with the static env kolla-ansible has, | 16:46 |
sdake | kfox1111 although it may require some tweaking to get it to work in kolla-kubernetes | 16:46 |
kfox1111 | but maybe not for kolla-kubernetes.... | 16:46 |
kfox1111 | gota think through that one some... | 16:46 |
sdake | ok gotta jet | 16:46 |
kfox1111 | yeah. | 16:46 |
kfox1111 | l8r | 16:46 |
SatyaP | while deploying dvr I am using a different external interface for compute. to use a different interface i added below line to kolla multi node file "compute01 neutron_external_interface=eno3" in [compute] section | 16:47 |
*** absubram has quit IRC | 16:47 | |
*** kbaegis has joined #openstack-kolla | 16:47 | |
SatyaP | but my change is not reflecting in compute node | 16:47 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible: Optimize reconfiguration for tacker https://review.openstack.org/422036 | 16:47 |
SatyaP | sdake or inc0 any suggestions? | 16:47 |
kfox1111 | SamYaple: ping | 16:47 |
SamYaple | kfox1111: pick a channel man! | 16:48 |
kfox1111 | sorry. didn't relize you were here when I poked you in the other channel. | 16:48 |
SamYaple | :) | 16:48 |
kfox1111 | so, you have a good handle on how multistr works in openstack's ini files, right? | 16:48 |
*** absubram has joined #openstack-kolla | 16:49 | |
SatyaP | SamYaple any suggestions for my issue | 16:50 |
kfox1111 | I was kind of curious if there was a subset of yaml that could map to openstack's ini files. | 16:50 |
kfox1111 | like, for a multistr ini line, you might be able to use a yaml list? | 16:50 |
kfox1111 | foo: | 16:50 |
kfox1111 | - a | 16:50 |
kfox1111 | - b | 16:50 |
kfox1111 | would be: | 16:50 |
kfox1111 | foo=a | 16:50 |
kfox1111 | foo=b | 16:50 |
kfox1111 | ? | 16:50 |
openstackgerrit | caoyuan proposed openstack/kolla-ansible: Optimize reconfiguration for heat https://review.openstack.org/421184 | 16:51 |
kfox1111 | think that would be possible? | 16:51 |
SamYaple | kfox1111: not without breaking yaml, no | 16:52 |
SamYaple | kfox1111: just as ini is broken for openstack | 16:52 |
kfox1111 | is there things in ini not representable in yaml? | 16:52 |
SamYaple | no there isnt! however, openstack is not using ini anymore thanks to multistr | 16:52 |
kfox1111 | thats what I ment... | 16:53 |
SamYaple | yep | 16:53 |
kfox1111 | is there anything in openstack's ini format that are not somehow representable in yaml? | 16:53 |
openstackgerrit | Merged openstack/kolla-ansible: Fix the group check for glance-registry https://review.openstack.org/421899 | 16:53 |
SamYaple | kfox1111: yes. multistr | 16:53 |
kfox1111 | if we represented multistr as a yaml list of strings, would that solve that particular problem? | 16:53 |
kfox1111 | are there others? | 16:53 |
SamYaple | no, becuase openstack options sometimes use comma list strings (foo=opt1,opt2,opt3) and those _aren't_ multistr | 16:54 |
*** kristian__ has joined #openstack-kolla | 16:54 | |
*** david-lyle has quit IRC | 16:55 | |
SamYaple | its hackish with a custom parser/renderer either way | 16:55 |
kfox1111 | you could represent them as just foo: opt1,opt2,opt3 | 16:55 |
*** kristian__ has quit IRC | 16:55 | |
*** absubram has quit IRC | 16:55 | |
SamYaple | which is how lists in yaml are represented in ini | 16:55 |
*** kristian__ has joined #openstack-kolla | 16:56 | |
SamYaple | im just saying, you are writing a custom renderer and it wont work like everyone expects | 16:56 |
*** kristian__ has quit IRC | 16:56 | |
kfox1111 | openstack ini doesnt either. so I'm kind of ok with that. :) | 16:56 |
*** kristian__ has joined #openstack-kolla | 16:57 | |
*** salv-orlando has joined #openstack-kolla | 16:57 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Fix gate failure https://review.openstack.org/422734 | 16:57 |
*** jascott1 has joined #openstack-kolla | 16:57 | |
*** jistr is now known as jistr|afk | 16:57 | |
kfox1111 | if there was a way to map a subset of yaml to a the openstack ini file format, back and forth, it might make for some interesting configuration options in helm. | 16:57 |
jascott1 | kfox1111 anything weird in the ini that prevents that? isnt it just 'classic' ini? | 16:58 |
SamYaple | jascott1: openstack breaks ini | 16:58 |
kfox1111 | jascott1: they support something called multistr | 16:58 |
*** lamt has joined #openstack-kolla | 16:58 | |
jascott1 | ah of course | 16:58 |
kfox1111 | which is pretty improper for ini. :/ | 16:59 |
kfox1111 | ini has no concept of lists though, | 16:59 |
kfox1111 | so if we mapped that to a yaml list, it could be maped back and forth. | 16:59 |
kfox1111 | are there any other gotchas then multistr? | 16:59 |
*** absubram has joined #openstack-kolla | 16:59 | |
satyar | Hi jeffrey4l_ | 16:59 |
satyar | i downgraded the docker-py to 1.8.1 | 17:00 |
SamYaple | kfox1111: mostly just not all options are multistr, and not all multistr options must be multistr | 17:00 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix the gate failure https://review.openstack.org/422344 | 17:00 |
satyar | i am nomore getting that issue | 17:00 |
Jeffrey4l_ | satyar, nice | 17:00 |
SamYaple | kfox1111: you can do some multistr as comma _and_ multiple matching keys | 17:00 |
satyar | bit its in line no 95 of kolla_toolbox.py | 17:00 |
SamYaple | kfox1111: however, you can't mix them | 17:00 |
satyar | it shows keyerror 'state' | 17:01 |
kfox1111 | is it really a multistr if it has comma's? | 17:01 |
satyar | where the script is checking the kolla toolbox is running | 17:01 |
kfox1111 | or is it just a python str that it splits? | 17:01 |
*** kristian__ has quit IRC | 17:01 | |
satyar | jeffrey4l_ checked out the kolla_toolbox is running | 17:02 |
Jeffrey4l_ | hrm. let me check. what your docker-engine version? | 17:02 |
SamYaple | kfox1111: the structure that oslo.config produces on the other side is the same | 17:02 |
SamYaple | docker-py is just called docker now | 17:02 |
kfox1111 | conversion from ini -> yaml would just look for multiple of the same key, and make that a list, and the reverse would convert lists to multistr's? | 17:02 |
satyar | 1.10 | 17:02 |
satyar | api_version is 1.22 | 17:02 |
satyar | 1.10.3 | 17:02 |
Jeffrey4l_ | ok. | 17:02 |
Jeffrey4l_ | chekcing. | 17:02 |
kfox1111 | SamYaple: can it be either comma or multi then? | 17:03 |
kfox1111 | if you switched one to the other it would always still work? | 17:03 |
kfox1111 | or at least, always from comma to multi? | 17:03 |
satyar | http://paste.openstack.org/show/595666/ | 17:03 |
SamYaple | kfox1111: unless it has changed, its comma _or_ multi, not a combination. but this depends on the option | 17:03 |
kfox1111 | per option, or per key in file? | 17:03 |
SamYaple | per option | 17:03 |
kfox1111 | so in the code, it says its multi, but always comma? | 17:04 |
SamYaple | strings are parsed in project side code | 17:04 |
SamYaple | multistr is oslo.config side | 17:04 |
kfox1111 | the comma variant should be proper ini. so if it just gets sucked into a yaml string, that should still be fine? | 17:04 |
SamYaple | sounds right | 17:05 |
kfox1111 | so its only the multistr version we would need to worry about. | 17:05 |
*** sayantani01 has quit IRC | 17:05 | |
*** lamt has quit IRC | 17:05 | |
SamYaple | kfox1111: you know Jeffrey4l_ is the smart one on this subject right? | 17:05 |
kbaegis | Who's working on Tacker ooc? | 17:05 |
SamYaple | just throwing that out tehre | 17:05 |
kfox1111 | so if we mapped that to a string list, it shoudl be directly mapable back and forth... | 17:05 |
kfox1111 | Jeffrey4l_: do you have any thoughts on the topic? :) | 17:06 |
*** lamt has joined #openstack-kolla | 17:06 | |
Jeffrey4l_ | reading scrollback | 17:06 |
*** yangzb09 has quit IRC | 17:06 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 17:07 |
Jeffrey4l_ | kfox1111, kfox1111> I was kind of curious if there was a subset of yaml that could map to openstack's ini files.? <-- is this the issue? why you need this? | 17:08 |
*** absubram has quit IRC | 17:08 | |
egonzalez | kbaegis: sup | 17:08 |
Jeffrey4l_ | for python ini parser, and oslo.config ini parser | 17:08 |
*** jemcevoy has joined #openstack-kolla | 17:08 | |
Jeffrey4l_ | foo = a | 17:09 |
kfox1111 | Jeffrey4l_: yeah. | 17:09 |
Jeffrey4l_ | b | 17:09 |
Jeffrey4l_ | should work | 17:09 |
kfox1111 | basically, helm has really good support for dealing with yaml. | 17:09 |
kfox1111 | exploring ideas for how we could do unopinionated openstack config. | 17:09 |
*** SatyaP has quit IRC | 17:09 | |
Jeffrey4l_ | ok. maybe it is possible. | 17:10 |
kfox1111 | if we took, say, nova's generated ini file, | 17:10 |
kfox1111 | converted it to yaml, and stuck it in a helm package, | 17:10 |
kbaegis | egonzalez: Hey, thanks. :) How is tacker looking on Kolla? Is the UI piece implemented? | 17:10 |
kfox1111 | it would be really easy for a user using helm to tweak it. | 17:10 |
jemcevoy | What can cause this error when starting the first heka container? raise errors.APIError(e, response, explanation=explanation)\\nAPIError: 500 Server Error: Internal Server Error (\"{\"message\":\"maximum retry count cannot be used with restart policy \\'unless-stopped\\'\"}\")\\n'"} | 17:10 |
kfox1111 | then we could convert it back to a proper openstack ini format I think during deploy? | 17:10 |
egonzalez | kbaegis: not yet https://review.openstack.org/#/c/407822/ | 17:10 |
Jeffrey4l_ | you want convert yaml to ini or make oslo.config to support yaml file? | 17:10 |
kfox1111 | woudl requrie handling multistr for sure. not sure what other gotcha's that aproach woudl have. | 17:10 |
egonzalez | jemcevoy: downgrade to docker-engine 1.12.6, 1.13 is broken atm | 17:11 |
kfox1111 | the former. | 17:11 |
*** kristian__ has joined #openstack-kolla | 17:11 | |
kfox1111 | map back/forth between openstack ini format, and a given yaml subset. | 17:11 |
Jeffrey4l_ | the former one is not matter with oslo-generator, right? | 17:11 |
*** kristian__ has quit IRC | 17:11 | |
kfox1111 | oslo generator creates the example configs? | 17:11 |
jemcevoy | Thanks egonzalez | 17:12 |
Jeffrey4l_ | egonzalez, my patch failed with some selinux issue. wired. i disabled selinux, and trying again. | 17:12 |
Jeffrey4l_ | kfox1111, during deployment, the example config is not needed. | 17:12 |
Jeffrey4l_ | because the example config is totally commented. | 17:12 |
kfox1111 | ah. | 17:13 |
kfox1111 | hmm... might still be interesting to take that commented file and convert it to a commented yaml format. | 17:14 |
kfox1111 | but not critical. | 17:14 |
egonzalez | kbaegis: just as suggestion, noticed that autogenerated password does not work in with tacker because reach max size in DB, opened a bug report for that. If face any issue creating default vif, use a shorter password | 17:14 |
*** unicell has joined #openstack-kolla | 17:14 | |
Jeffrey4l_ | ok. so you wanna to the latter one: make oslo.config to support yaml ( at least for generate ) | 17:14 |
satyar | jeffrey4l_ http://paste.openstack.org/show/595668/ | 17:15 |
egonzalez | kbaegis: https://bugs.launchpad.net/tacker/+bug/1657508 | 17:15 |
openstack | Launchpad bug 1657508 in tacker "Data too long for column 'password'" [Medium,In progress] - Assigned to yong sheng gong (gongysh) | 17:15 |
satyar | jeffrey4l_ http://paste.openstack.org/show/595669/ | 17:15 |
Jeffrey4l_ | satyar, try this python -c 'import docker;print(docker.Client().containers()[0])' and tell me the output. | 17:16 |
kfox1111 | k. | 17:16 |
sbezverk | kfox1111: https://review.openstack.org/#/c/422229/ | 17:16 |
*** egonzalez has quit IRC | 17:17 | |
Jeffrey4l_ | kfox1111, if so, i have to say that need change the oslo.config's code ;( | 17:17 |
kfox1111 | Jeffrey4l_: I think we can live without that. | 17:18 |
satyar | jeffrey4l_ http://paste.openstack.org/show/595671/ | 17:18 |
Jeffrey4l_ | kfox1111, so you just want convert yaml to ini in k8s? ( i think this should possible. ) | 17:18 |
*** david-lyle has joined #openstack-kolla | 17:18 | |
*** david-lyle has quit IRC | 17:18 | |
kfox1111 | Jeffrey4l_: yeah. | 17:19 |
Jeffrey4l_ | the issue is: oslo.config support two type of list value. | 17:19 |
Jeffrey4l_ | ListOpt and MultiOpt | 17:19 |
Jeffrey4l_ | you can not distinguish this two kind in yaml. | 17:20 |
kfox1111 | what is listopt? | 17:20 |
kfox1111 | is that the foo=a,b,c thing? | 17:20 |
Jeffrey4l_ | ListOpt is: foo=value1,value2 MultiOpt is: foot=value1\n\tvalue2.. | 17:21 |
Jeffrey4l_ | ok you can use foo: "value1, valu2" | 17:21 |
kfox1111 | yeah. | 17:21 |
satyar | jeffrey4l_ http://paste.openstack.org/show/595672/ | 17:21 |
Jeffrey4l_ | satyar, docker's api is a shit ;( | 17:22 |
satyar | how to solve this :( | 17:23 |
satyar | its broke my whole deployment :( | 17:23 |
Jeffrey4l_ | one simple solution is upgrade docker engine to 1.12 | 17:24 |
Jeffrey4l_ | another one is fix the kolla_toolbox module. | 17:24 |
Jeffrey4l_ | i will push a fix ^^ | 17:24 |
*** sayantani01 has joined #openstack-kolla | 17:24 | |
*** vhosakot has joined #openstack-kolla | 17:25 | |
*** yangzb09 has joined #openstack-kolla | 17:25 | |
*** jascott1 has quit IRC | 17:26 | |
*** jascott1 has joined #openstack-kolla | 17:26 | |
*** david-lyle has joined #openstack-kolla | 17:28 | |
*** shardy has quit IRC | 17:29 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Rabbitmq service deployment with entrypoint https://review.openstack.org/422229 | 17:30 |
Jeffrey4l_ | satyar, please try this on you node and tell me the output python -c "import docker;c=docker.Client();t=c.containers(filters=dict(name='kolla_toolbox', status='running'));print(t)" | 17:31 |
kfox1111 | sbezverk: looks nice. :) | 17:31 |
*** jrich523_ is now known as jrich523 | 17:32 | |
*** sp__ has joined #openstack-kolla | 17:32 | |
*** yangzb09 has quit IRC | 17:32 | |
sp__ | https://review.openstack.org/#/c/422444/ | 17:34 |
sp__ | Jeffrey4l_: can you please check why the gate is failing this time ? | 17:34 |
sp__ | Jeffrey4l_: everything running all good on my machine | 17:35 |
jascott1 | kfox1111 if we landed conditions and tags in helm, we could move microservices folders back under their respective parent /charts folders? | 17:35 |
Jeffrey4l_ | gate is bust due to docker 1.13 is released yesterday. no idea why. i have a patch to pin the docker-engine version to 1.12.6 which seems work. | 17:35 |
kfox1111 | jascott1: no. | 17:35 |
kfox1111 | because there is not always a 1:1 mapping. | 17:35 |
kfox1111 | mariadb will be deployable under nova/glance/etc | 17:36 |
*** JeroenBo has joined #openstack-kolla | 17:36 | |
*** JeroenBo has left #openstack-kolla | 17:36 | |
*** mkoderer has quit IRC | 17:36 | |
satyar | jeffrey4l_ http://paste.openstack.org/show/595675/ | 17:37 |
sp__ | Jeffrey4l_: thanks | 17:37 |
*** dturner has joined #openstack-kolla | 17:37 | |
Jeffrey4l_ | satyar, thanks. trying to fix the issue. | 17:37 |
kfox1111 | there is no tree structure I think. its more of a graph. | 17:37 |
kfox1111 | well, it may be representable as a tree with shared branches, but most file systems don't like that. ;) | 17:37 |
*** matrohon has quit IRC | 17:38 | |
satyar | Thanks :) | 17:38 |
jascott1 | heheh yeah I have been looking into the relative path for charts idea | 17:38 |
*** vhosakot has quit IRC | 17:38 | |
kfox1111 | yeah. local path would solve it better I think. | 17:38 |
*** vhosakot has joined #openstack-kolla | 17:38 | |
kfox1111 | still think though that its an optimization for a use case that shouldn't be common. | 17:38 |
kfox1111 | users should be deploying packages, not stright out of a git checkout most of the time. | 17:39 |
kfox1111 | more 'helm native' :) | 17:39 |
jascott1 | :D | 17:39 |
*** Serlex has quit IRC | 17:39 | |
kfox1111 | brb. meeting. | 17:40 |
*** unicell has quit IRC | 17:40 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Fix the kolla_toolbox module do not work on docker engine 1.10 https://review.openstack.org/422781 | 17:43 |
Jeffrey4l_ | satyar, check this https://review.openstack.org/422781 | 17:44 |
Jeffrey4l_ | you can try this patch ;) | 17:44 |
*** sp___ has joined #openstack-kolla | 17:44 | |
*** salv-orlando has quit IRC | 17:45 | |
satyar | yes this will work out :) | 17:45 |
*** adrian_otto has quit IRC | 17:46 | |
*** sdake_ has joined #openstack-kolla | 17:47 | |
*** yingjun has quit IRC | 17:48 | |
Jeffrey4l_ | SamYaple, still around? | 17:48 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: Fix gate failure https://review.openstack.org/422734 | 17:49 |
*** sdake has quit IRC | 17:50 | |
SamYaple | Jeffrey4l_: briefly | 17:51 |
Jeffrey4l_ | SamYaple, will u finish the monasca bp before m3 is released? ( around Jan 25 ) | 17:51 |
SamYaple | Jeffrey4l_: unlikely | 17:52 |
*** markmcclain has joined #openstack-kolla | 17:53 | |
Jeffrey4l_ | ok. | 17:53 |
*** absubram has joined #openstack-kolla | 17:56 | |
kfox1111 | arg.... | 17:56 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Fix the gate failure https://review.openstack.org/422344 | 17:56 |
kfox1111 | kubernetes in the reboot job is really grumpy now... | 17:56 |
kfox1111 | so strange.... | 17:57 |
*** srwilkers__ is now known as srwilkers_brb | 17:58 | |
kfox1111 | maybe you can't timeout a sudo... | 17:59 |
sbezverk | kfox1111: worked for me 25 minutes ago.. | 17:59 |
sbezverk | I guess it really depends on a cloud provider | 17:59 |
kfox1111 | sbezverk: yeah. all the other jobs are working. but something changed in the reboot gate. the last day and a half has been nothing but hanging in k8s deploy. :/ | 18:00 |
kfox1111 | frustrating. cause I think its pretty close to working now. :/ | 18:00 |
kfox1111 | its taking the whole hour to detect its broken and die. so can't run too many jobs. :/ | 18:00 |
*** lrensing has quit IRC | 18:01 | |
kfox1111 | just getting stuck here forever: 2017-01-19 17:25:18.488145 | [apiclient] Created API client, waiting for the control plane to become ready | 18:01 |
*** sp__ has quit IRC | 18:01 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 18:01 |
*** jistr|afk is now known as jistr | 18:01 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible: [TEST][MASTER][KOLLA-ANSIBLE] https://review.openstack.org/422786 | 18:02 |
*** tonanhngo has quit IRC | 18:03 | |
*** tonanhngo has joined #openstack-kolla | 18:03 | |
*** tonanhngo has quit IRC | 18:04 | |
*** athomas has quit IRC | 18:04 | |
*** krtaylor has quit IRC | 18:05 | |
*** yangzb09 has joined #openstack-kolla | 18:05 | |
*** MNghia has joined #openstack-kolla | 18:06 | |
kfox1111 | inc0: do you think this is devicemapper related? http://logs.openstack.org/29/422229/7/experimental/gate-kolla-kubernetes-deploy-centos-binary-2-iscsi-nv/296c926/console.html | 18:06 |
kfox1111 | we see it now and then. | 18:06 |
*** MNghia has quit IRC | 18:06 | |
kfox1111 | the other thought is its a pull race condition. | 18:06 |
inc0 | hmm I dunno | 18:07 |
inc0 | never seen it | 18:07 |
sbezverk | kfox1111: sha it is refering is docker image or layer? | 18:10 |
kfox1111 | guessing thats a layer. | 18:10 |
sbezverk | kfox1111: we are not supposed to change layer, we pull, never push | 18:11 |
kfox1111 | yeah. so its docker itself throwing that. | 18:11 |
*** srwilkers_brb has quit IRC | 18:12 | |
*** gfidente has quit IRC | 18:12 | |
sbezverk | kfox1111: yeah, we should bumup log verbosity for docker daemon to catch it | 18:12 |
kfox1111 | I wonder if we ever see it on ubuntu. | 18:13 |
kfox1111 | maybe a bug in the version shipped with centos. | 18:13 |
openstackgerrit | Merged openstack/kolla-kubernetes: Rabbitmq service deployment with entrypoint https://review.openstack.org/422229 | 18:13 |
sbezverk | kfox1111: will do the same element_name change for memcached and mariadb | 18:15 |
*** adrian_otto has joined #openstack-kolla | 18:17 | |
*** david-lyle has quit IRC | 18:18 | |
*** srwilkers has joined #openstack-kolla | 18:18 | |
*** schwicht has quit IRC | 18:19 | |
*** yangzb09 has quit IRC | 18:21 | |
*** sayantani01 has quit IRC | 18:21 | |
*** sdake has joined #openstack-kolla | 18:24 | |
*** schwicht has joined #openstack-kolla | 18:26 | |
*** vhosakot has quit IRC | 18:28 | |
*** sdake_ has quit IRC | 18:28 | |
*** vhosakot has joined #openstack-kolla | 18:28 | |
kfox1111 | sbezverk: awesome. thanks. :) | 18:30 |
*** lrensing has joined #openstack-kolla | 18:30 | |
openstackgerrit | prameswar proposed openstack/kolla-kubernetes: Installation path added for ubuntu to pathfinder.py https://review.openstack.org/422803 | 18:36 |
*** absubram has quit IRC | 18:36 | |
kfox1111 | ok. got it to at least fail faster now... | 18:38 |
*** prameswar has joined #openstack-kolla | 18:38 | |
kfox1111 | so weird.... | 18:38 |
*** sdake_ has joined #openstack-kolla | 18:40 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Fixing element_name path for mariadb adn memcached https://review.openstack.org/422808 | 18:41 |
*** sdake has quit IRC | 18:42 | |
bmace | kfox1111: starting on some of the file renames. what is exposed / referenced in the infra repository as far as the gate setup / run. i don't want to try to do any file renames in tools that break things. i know it will be obvious once the code is posted but if it is just the setup_gate that is referenced in the infra repo i can change everything else. | 18:47 |
openstackgerrit | prameswar proposed openstack/kolla-kubernetes: Installation path added for ubuntu to pathfinder.py https://review.openstack.org/422803 | 18:51 |
*** vhosakot has quit IRC | 18:51 | |
*** vhosakot has joined #openstack-kolla | 18:52 | |
*** sdake has joined #openstack-kolla | 18:52 | |
*** narasimha_SV has quit IRC | 18:53 | |
*** sdake_ has quit IRC | 18:54 | |
*** krtaylor has joined #openstack-kolla | 18:56 | |
kfox1111 | the renanmes should just be in helm/microservices/*/templates/* | 18:59 |
kfox1111 | those names shouldn't matter at all so you should be free to just rename them. | 18:59 |
*** prameswar has quit IRC | 19:06 | |
jemcevoy | egonzalez: are the images I built last night with docker 1.13.0 ok or do I need to run kolla-build again with 1.12.6? | 19:06 |
*** chopmann has joined #openstack-kolla | 19:07 | |
*** srwilkers has quit IRC | 19:09 | |
*** MasterOfBugs has joined #openstack-kolla | 19:12 | |
*** pramodrj07 has joined #openstack-kolla | 19:12 | |
*** matrohon has joined #openstack-kolla | 19:13 | |
*** sdake has quit IRC | 19:14 | |
sbezverk | kfox1111: https://review.openstack.org/422808 | 19:17 |
*** sdake has joined #openstack-kolla | 19:17 | |
bmace | kfox1111: so continued chaos in script naming in tools / tests? :) | 19:18 |
*** sdake_ has joined #openstack-kolla | 19:20 | |
*** sdake has quit IRC | 19:22 | |
*** absubram has joined #openstack-kolla | 19:25 | |
kfox1111 | oh. you were wanting to update the tools too. | 19:26 |
kfox1111 | hmm... | 19:26 |
kfox1111 | maybe do that in a seperate ps? | 19:26 |
kfox1111 | the helm stuff should cause the most merge conflicts and should be quick to go in. | 19:27 |
kfox1111 | sbezverk: reviewed | 19:27 |
*** adrian_otto has quit IRC | 19:28 | |
kfox1111 | what possibly borke in the gate for kube... how weird... | 19:28 |
sdake_ | kfox1111 i changed project-config | 19:32 |
sdake_ | kfox1111 doubtful it has mergfed yet | 19:32 |
*** adrian_otto has joined #openstack-kolla | 19:33 | |
*** openstackgerrit has quit IRC | 19:33 | |
sdake_ | kfox1111 my gate changes have not yet merged | 19:33 |
kfox1111 | sdake_: got a review? | 19:34 |
*** openstackgerrit has joined #openstack-kolla | 19:34 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 19:34 |
sdake_ | https://review.openstack.org/422597 | 19:35 |
satyar | Hi pbourke | 19:35 |
kfox1111 | ah. cool. | 19:35 |
sdake_ | general idea is gate jobs that pass 80%+ of the time should be promoted to non-voting | 19:35 |
sdake_ | to eliminate thee need for check experimental | 19:35 |
sdake_ | and force the need for recheck in the 20% case | 19:36 |
sdake_ | since the gate jobs i promoted pass all the time unless there is some wierd kind of heisenbug | 19:36 |
kfox1111 | so, that ps removes a job. | 19:36 |
kfox1111 | do we want to do that? | 19:36 |
sdake_ | are you sure about that? | 19:36 |
sdake_ | what job does it remove | 19:37 |
kfox1111 | https://review.openstack.org/#/c/422597/1/zuul/layout.yaml | 19:37 |
sdake_ | there was no intent to remove a job | 19:37 |
kfox1111 | the ubuntu iscsi job was removed. | 19:37 |
sdake_ | damn | 19:37 |
sdake_ | thanks for checking that | 19:37 |
*** msimonin1 has quit IRC | 19:37 | |
kfox1111 | np. | 19:37 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Fixing element_name path for mariadb and memcached https://review.openstack.org/422808 | 19:40 |
*** lrensing has quit IRC | 19:44 | |
*** lrensing has joined #openstack-kolla | 19:45 | |
*** david-lyle has joined #openstack-kolla | 19:48 | |
openstackgerrit | Edmund Rhudy proposed openstack/kolla: Add block for dumb-init installation https://review.openstack.org/422249 | 19:50 |
kfox1111 | david-lyle: hey. :) | 19:50 |
*** tonanhngo has joined #openstack-kolla | 19:50 | |
david-lyle | kfox1111: o/ | 19:51 |
*** vhosakot has quit IRC | 19:54 | |
*** tonanhngo has quit IRC | 19:54 | |
*** tonanhngo has joined #openstack-kolla | 19:55 | |
*** tonanhngo has quit IRC | 19:55 | |
*** vhosakot has joined #openstack-kolla | 19:55 | |
*** tonanhngo has joined #openstack-kolla | 19:55 | |
rhallisey | sbezverk, around? | 19:56 |
jrich523 | so, was just tying to build up kolla again, and just installed docker (via quick start steps) and the "docker daemon" is apparently "going to be depreciated" but... doesnt work lol | 19:57 |
rhallisey | jrich523, it's dockerd now | 19:58 |
jrich523 | it says to use dockerd instead, anyone run in to this? | 19:58 |
jrich523 | right, but, shouldnt that be what ends up in the kolla.conf file there? | 19:58 |
kfox1111 | weird... it was the selinux disabling in the gate master that broke kubernetes... very weird. | 19:58 |
kfox1111 | just moving it to after the kube init fixed it. | 19:59 |
*** vhosakot has quit IRC | 19:59 | |
rhallisey | jrich523, kolla.conf will need to reflect that change | 19:59 |
*** TxGirlGeek has quit IRC | 19:59 | |
sbezverk | rhallisey: hey | 20:00 |
jrich523 | i guess what i mean is, shouldnt that be reflected in the quick start? | 20:00 |
*** TxGirlGeek has joined #openstack-kolla | 20:01 | |
rhallisey | sbezverk, what did you want to discuss about operators? | 20:01 |
rhallisey | I can talk now | 20:01 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 20:01 |
*** sambetts is now known as sambetts|afk | 20:03 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Fixing element_name path for mariadb and memcached https://review.openstack.org/422808 | 20:03 |
sbezverk | rhallisey: it seems I will need to redo a few things due to what was discovered last night. I will ping you when new version is ready.. | 20:04 |
rhallisey | ok | 20:04 |
*** lrensing has quit IRC | 20:06 | |
*** absubram has quit IRC | 20:06 | |
*** TxGirlGeek has quit IRC | 20:07 | |
kfox1111 | sbezverk: looks good. thanks. :) | 20:07 |
*** dave-mccowan has quit IRC | 20:08 | |
*** dave-mccowan has joined #openstack-kolla | 20:09 | |
*** salv-orlando has joined #openstack-kolla | 20:11 | |
*** TxGirlGeek has joined #openstack-kolla | 20:13 | |
sbezverk | kfox1111: novncproxy, does it depend on nova-api service? | 20:14 |
kfox1111 | hmm.. thats a good question. | 20:14 |
kfox1111 | not sure. I wana say yes. but it may just depend on rpc services. | 20:15 |
kfox1111 | might ask in the nova channel to double check. | 20:15 |
sbezverk | kfox1111: if we add this dependency just as a safeguard? it should not break anything. anyway if nova-api is not running novncproxy would be useless | 20:16 |
kfox1111 | I'm ok with that. | 20:16 |
sbezverk | kfox1111: cool, thanks | 20:17 |
*** TxGirlGeek has quit IRC | 20:17 | |
*** TxGirlGeek has joined #openstack-kolla | 20:17 | |
kelv | is there a particular order to add control nodes for multinode deployment? mariadb fails to bootstrap when multiple machines are initially defined. keystone fernet also complains about key repository being present with multi nodes | 20:21 |
*** vhosakot has joined #openstack-kolla | 20:21 | |
*** TxGirlGeek has quit IRC | 20:23 | |
kfox1111 | weird..... | 20:26 |
kfox1111 | http://logs.openstack.org/93/413193/52/experimental/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-reboot-nv/56b3f66/console.html | 20:26 |
kfox1111 | the openstack command existed before the reboot.... | 20:26 |
*** TxGirlGeek has joined #openstack-kolla | 20:27 | |
openstackgerrit | Mark Giles proposed openstack/kolla-kubernetes: Nova-control service deployment https://review.openstack.org/421457 | 20:27 |
*** vhosakot has quit IRC | 20:30 | |
*** vhosakot has joined #openstack-kolla | 20:30 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 20:32 |
*** TxGirlGeek has quit IRC | 20:35 | |
*** TxGirlGeek has joined #openstack-kolla | 20:36 | |
*** adrian_otto has quit IRC | 20:39 | |
*** TxGirlGeek has quit IRC | 20:41 | |
*** vhosakot has quit IRC | 20:41 | |
*** vhosakot has joined #openstack-kolla | 20:42 | |
openstackgerrit | sean mooney proposed openstack/kolla-ansible: This change allow setting kernel commandline args https://review.openstack.org/408876 | 20:53 |
openstackgerrit | sean mooney proposed openstack/kolla-ansible: extends ansible roles to deploy ovsdpdk https://review.openstack.org/408872 | 20:53 |
openstackgerrit | sean mooney proposed openstack/kolla-ansible: adds ovs-dpdkctl tool https://review.openstack.org/408871 | 20:53 |
*** TxGirlGeek has joined #openstack-kolla | 20:53 | |
kbaegis | How do I override the default timeout for mariadb? It's taking longer than the ansible deploy scripts seem to like | 20:55 |
sean-k-mooney | kbaegis: where is this timeing out? | 20:57 |
*** pablo|500| has quit IRC | 20:57 | |
kbaegis | sean-k-mooney: Waiting for the mariadb port I believe. | 20:57 |
*** salv-orlando has quit IRC | 20:57 | |
sean-k-mooney | kbaegis: https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/mariadb/defaults/main.yml#L8 | 20:57 |
sean-k-mooney | kbaegis: if you set database_max_timeout in the global.yml it might help | 20:58 |
kbaegis | tyvm | 20:58 |
sean-k-mooney | kbaegis: is https://github.com/openstack/kolla-ansible/blob/7bb88390d88aebfc257e09e7e055a09fe4121782/ansible/roles/mariadb/tasks/bootstrap_cluster.yml#L2-L21 the task that is failing | 20:59 |
kbaegis | "mariadb : Waiting for MariaDB service to be ready" | 20:59 |
sean-k-mooney | kbaegis: https://github.com/openstack/kolla-ansible/blob/7bb88390d88aebfc257e09e7e055a09fe4121782/ansible/roles/mariadb/tasks/start.yml#L21-L31 | 21:00 |
kbaegis | It's failing all 10 times | 21:00 |
sean-k-mooney | it looks like it is hardcoded | 21:00 |
kbaegis | Gotcha. I can play with that value. Something else may be going on | 21:00 |
*** lrensing has joined #openstack-kolla | 21:02 | |
*** sayantani01 has joined #openstack-kolla | 21:02 | |
kfox1111 | whaa? | 21:04 |
kfox1111 | http://logs.openstack.org/93/413193/53/experimental/gate-kolla-kubernetes-deploy-centos-binary-2-ceph-reboot-nv/96752bf/console.html | 21:04 |
kfox1111 | how can the openstack client disapear, but the keystonerc file remain... | 21:04 |
kbaegis | I set the delay to 60. Should help :) | 21:05 |
kfox1111 | oh.. wait... did I virtualenv that? | 21:05 |
kbaegis | I figure if that doesn't fix it, something else is wrong | 21:05 |
*** unicell has joined #openstack-kolla | 21:05 | |
kfox1111 | yup. virtualenv. nm. I'm just dumb. :) | 21:05 |
sean-k-mooney | kfox1111: is the gate ok in general https://review.openstack.org/#/c/408876/2 seams to be foobared but im pretty sure it has nothing to do with the patch | 21:06 |
kfox1111 | sean-k-mooney: I don't know the status of the kolla-ansible gate. | 21:07 |
kfox1111 | the kolla-kubernetes one seems pretty stable. | 21:07 |
sean-k-mooney | kfox1111: well docker: Error response from daemon: linux init cgroups path: mkdir /docker: not a directory. is a new error for me to see in the gate | 21:07 |
kfox1111 | we're using the distro docker in kolla-kubernetes gate. I think kolla-ansible's using docker inc's? | 21:08 |
kfox1111 | so maybe they did a release? | 21:08 |
*** v1k0d3n has quit IRC | 21:09 | |
sean-k-mooney | actully it looks like it my fault... | 21:09 |
*** unicell has quit IRC | 21:10 | |
sean-k-mooney | one of the other ptaches in the chain has a syntax error though i think there are 2 errors at once | 21:10 |
kfox1111 | ah. | 21:10 |
*** msimonin has joined #openstack-kolla | 21:12 | |
sean-k-mooney | it weird though casue ubuntu source tells me about the syntax error in the yml file and centos has the docker cgroups issue. anyway ill deal with it tomorow | 21:13 |
*** msimonin1 has joined #openstack-kolla | 21:14 | |
*** msimonin has quit IRC | 21:14 | |
*** absubram has joined #openstack-kolla | 21:15 | |
kbaegis | Well, that didn't help. Looks like it's not able to validate that the port is getting exposed | 21:15 |
kfox1111 | sean-k-mooney: yeah, thats weird. | 21:15 |
sean-k-mooney | i actully think the centos and oraclelinux gate are broken http://logs.openstack.org/71/408871/3/check/gate-kolla-ansible-dsvm-deploy-oraclelinux-binary-centos-7-nv/09dfc48/console.html#_2017-01-19_21_00_20_931883 | 21:16 |
*** satyar has quit IRC | 21:18 | |
sdake_ | sean-k-mooney i htink someon ewas working on that | 21:19 |
openstackgerrit | Merged openstack/kolla-kubernetes: Fixing element_name path for mariadb and memcached https://review.openstack.org/422808 | 21:19 |
sean-k-mooney | sdake_: ok was not sure if i was the first person to hit it or not | 21:19 |
sdake_ | sean-k-mooney someone asked Jeffrey4l_ about it last night | 21:19 |
sdake_ | sean-k-mooney i assume emea was working on a gate fialure resolution | 21:20 |
sdake_ | but not sure | 21:20 |
*** krtaylor has quit IRC | 21:21 | |
sean-k-mooney | ok well i need to do more local testing of that chain anyway so im not to worried about the gate passing on them as i will like have to rework part of them anyway | 21:21 |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 21:21 |
sean-k-mooney | sdake_: do you know when feature freeze for kolla/kolla-ansible will be for ocata | 21:22 |
sdake_ | sean-k-mooney i do not, although it is generally milestone 3 in the past | 21:23 |
sdake_ | sean-k-mooney which is in the openstack schedule | 21:23 |
sdake_ | i could find it for you on google but have to run to pick up daughter from school | 21:23 |
sean-k-mooney | no worries | 21:23 |
sean-k-mooney | ill find it myself. | 21:23 |
*** nmn has joined #openstack-kolla | 21:24 | |
sean-k-mooney | milestone 3 would be january 26 which would be 1 week from today. the cycle trailing deadline would be ~ march 2nd which is the lates that kolla can release i belive | 21:26 |
*** mgiles has quit IRC | 21:29 | |
*** sayantani01 has quit IRC | 21:30 | |
*** sayantani01 has joined #openstack-kolla | 21:31 | |
kfox1111 | any chance of getting ceph-mds/cephfs containers in by then? | 21:35 |
*** david-lyle has quit IRC | 21:35 | |
*** krtaylor has joined #openstack-kolla | 21:35 | |
*** Jeffrey4l_ has quit IRC | 21:35 | |
*** Jeffrey4l_ has joined #openstack-kolla | 21:36 | |
*** adrian_otto has joined #openstack-kolla | 21:36 | |
*** salv-orlando has joined #openstack-kolla | 21:36 | |
*** jascott1 has quit IRC | 21:36 | |
*** srwilkers has joined #openstack-kolla | 21:38 | |
*** david-lyle has joined #openstack-kolla | 21:39 | |
*** david-lyle has quit IRC | 21:39 | |
*** david-lyle has joined #openstack-kolla | 21:39 | |
*** jascott1 has joined #openstack-kolla | 21:40 | |
*** v1k0d3n has joined #openstack-kolla | 21:46 | |
ScoopIcicle | Our timeline credibly synergizes the policy makers as a consequence of an accelerated cost reduction. The senior support staff drives efficient frontiers. | 21:47 |
*** inc0 has quit IRC | 21:48 | |
*** inc0_ has joined #openstack-kolla | 21:48 | |
*** inc0_ has quit IRC | 21:48 | |
*** inc0 has joined #openstack-kolla | 21:48 | |
*** pbourke has quit IRC | 21:50 | |
*** portdirect has joined #openstack-kolla | 21:52 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 21:53 |
*** fguillot has quit IRC | 21:56 | |
*** JoseMello has quit IRC | 22:05 | |
openstackgerrit | BorneMace proposed openstack/kolla-kubernetes: Updated helm template names to consistent https://review.openstack.org/422889 | 22:05 |
*** dave-mccowan has quit IRC | 22:08 | |
openstackgerrit | BorneMace proposed openstack/kolla-kubernetes: Updated helm template names to consistent https://review.openstack.org/422889 | 22:10 |
*** jtriley has quit IRC | 22:13 | |
inc0 | kfox1111, sbezverk others, so I'm looking at helm repo and I have a question | 22:17 |
inc0 | https://github.com/kubernetes/charts/tree/master/stable/mariadb/templates <- where do I define order of which things are created? | 22:18 |
openstackgerrit | Cornelio Hopmann proposed openstack/kolla-ansible: Fix magnum [DEFAULT] configuration. https://review.openstack.org/422896 | 22:22 |
*** salv-orl_ has joined #openstack-kolla | 22:24 | |
dturner | Jeffrey4l_ still around? | 22:24 |
*** imcsk8 has quit IRC | 22:27 | |
*** salv-orlando has quit IRC | 22:27 | |
*** jheroux has quit IRC | 22:33 | |
*** sdake_ has quit IRC | 22:44 | |
*** nmn has quit IRC | 22:46 | |
kfox1111 | inc0: you don't. | 22:47 |
kfox1111 | helm doesn't support that. :/ | 22:47 |
inc0 | ah fun | 22:47 |
kfox1111 | we've been using helm+kubernetes-entrypoint to try and orchestrate things. | 22:47 |
inc0 | is it just throws everything in templates at k8s and patiently wait till it resolve itself? | 22:47 |
kfox1111 | depneds on the program. | 22:48 |
kfox1111 | k8s will keep relaunching pods that crash. | 22:48 |
inc0 | yeah I'm looking at mariadb | 22:48 |
kfox1111 | it should have a liveness probe to handle programs taht start but lock up if their deps arn't there. | 22:48 |
inc0 | I'll probably do the same - init container with entrypoint -> init container for galera bootstrap -> statefulset for cluster | 22:48 |
kfox1111 | yeah. should work. | 22:49 |
kfox1111 | see the existing mariadb. it should be close to what you want. | 22:49 |
kfox1111 | sbezverk is working on a service package. | 22:49 |
kfox1111 | shoudl look very similar to the just merged rabbit one though, if you want to reference that. | 22:50 |
dturner | Hi- Hitting a docker api version mismatch error in our tests of kolla-ansible trunk. We discovered that the deploy nodes are pinned to Docker engine 1.10.3, but latest docker-py is getting installed, causing mismatch. | 22:50 |
dturner | Is there an "official" list of min versions of these components for "O" release of kolla-ansible yet? | 22:50 |
inc0 | dturner, pinned to 1.10.3? They shouldn't be | 22:53 |
dturner | That's something we do in our provisioning scripts (kickstart) when prepping our baremetal nodes. The thinking was, we didn't want to install a version of docker-engine not compatible with Kolla. | 22:55 |
inc0 | in bootstrap servers we pin to 1.11.* | 22:55 |
inc0 | yeah. but 1.10 is minimal, 1.11 is one we deploy normally | 22:55 |
inc0 | newer should work too, just 1.11 is best tested | 22:55 |
dturner | Right. We should probably start taking advantage of that. We set this up prior to Kolla including the bootstrap goodies. I should have checked there for clues. | 22:56 |
*** adrian_otto has quit IRC | 22:56 | |
dturner | inc0 Thanks for the tip. | 22:56 |
inc0 | any time, let me know if you run into any issues | 22:57 |
*** lamt has quit IRC | 23:00 | |
*** lamt has joined #openstack-kolla | 23:01 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 23:02 |
*** david-lyle has quit IRC | 23:03 | |
*** absubram has quit IRC | 23:05 | |
*** lamt has quit IRC | 23:06 | |
*** absubram has joined #openstack-kolla | 23:09 | |
*** krtaylor has quit IRC | 23:10 | |
jascott1 | kfox1111 wrt to earlier conversation about chart deps, prefer pv/pvc as one toggle or separate? | 23:12 |
kfox1111 | yes? :/ | 23:14 |
kfox1111 | theres a dynamic pvc coming, that will support auto creating pv's. | 23:14 |
kfox1111 | so pv's wont be needed in all cases. | 23:15 |
kfox1111 | its not really useful to ever make a pv without a pvc in a chart though? | 23:15 |
jascott1 | thats what I was trying to ask I think | 23:15 |
bmace | thanks for the reviews on that name change gents. | 23:16 |
kfox1111 | bmace: thx for doing the work. :) | 23:16 |
jascott1 | i can make them separate, can be easily combined later as further example | 23:16 |
bmace | kfox1111: my pleasure. i'm a huge fan of consistency :) | 23:17 |
kfox1111 | +1 for consistency | 23:17 |
openstackgerrit | Merged openstack/kolla-kubernetes: Updated helm template names to consistent https://review.openstack.org/422889 | 23:21 |
*** matrohon has quit IRC | 23:21 | |
*** srwilkers has quit IRC | 23:21 | |
*** sacharya_ has joined #openstack-kolla | 23:23 | |
*** absubram has quit IRC | 23:23 | |
*** sacharya has quit IRC | 23:23 | |
*** vhosakot has quit IRC | 23:23 | |
*** inc0 has quit IRC | 23:26 | |
*** TxGirlGeek has quit IRC | 23:33 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP memcached operator https://review.openstack.org/418073 | 23:38 |
*** lrensing has quit IRC | 23:39 | |
*** schwicht has quit IRC | 23:41 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 23:45 |
*** v1k0d3n has quit IRC | 23:45 | |
*** chopmann has quit IRC | 23:46 | |
*** chopmann has joined #openstack-kolla | 23:46 | |
*** wirehead_ has quit IRC | 23:49 | |
*** PramodJ has joined #openstack-kolla | 23:50 | |
*** wirehead_ has joined #openstack-kolla | 23:50 | |
*** pramodrj07 has quit IRC | 23:53 | |
*** MasterOfBugs has quit IRC | 23:53 | |
openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP: test rebooting in the gate https://review.openstack.org/413193 | 23:55 |
*** schwicht has joined #openstack-kolla | 23:58 | |
*** sayantani01 has quit IRC | 23:59 | |
*** jascott1 has quit IRC | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!