*** BjoernT has joined #openstack-kolla | 00:43 | |
*** BjoernT has quit IRC | 00:49 | |
*** spsurya has joined #openstack-kolla | 01:09 | |
*** redrobot has quit IRC | 02:23 | |
*** Guest90568 has joined #openstack-kolla | 02:29 | |
*** Guest90568 is now known as redrobot | 02:32 | |
*** BjoernT has joined #openstack-kolla | 02:45 | |
*** dave-mccowan has quit IRC | 03:46 | |
*** BjoernT has quit IRC | 04:43 | |
*** dpawlik has joined #openstack-kolla | 05:01 | |
*** skramaja has joined #openstack-kolla | 05:09 | |
*** janki has joined #openstack-kolla | 05:24 | |
*** mrunge has quit IRC | 05:52 | |
*** mrunge_ has joined #openstack-kolla | 05:52 | |
*** mrunge_ is now known as mrunge | 05:56 | |
*** jbadiapa has joined #openstack-kolla | 06:05 | |
*** skramaja_ has joined #openstack-kolla | 06:07 | |
*** skramaja has quit IRC | 06:07 | |
*** threestrands has quit IRC | 06:34 | |
*** threestrands has joined #openstack-kolla | 06:35 | |
*** shyamb has joined #openstack-kolla | 06:55 | |
shyamb | Hi | 06:55 |
---|---|---|
shyamb | I am mounting a nfs share on nova_libvirt container | 06:56 |
shyamb | but container does not have nfs mount helper package | 06:56 |
*** shyamb has quit IRC | 07:00 | |
*** shyam89 has joined #openstack-kolla | 07:00 | |
*** shyam89 has quit IRC | 07:06 | |
*** shyamb has joined #openstack-kolla | 07:06 | |
openstackgerrit | Merged openstack/kolla-ansible master: Remove nova [DEFAULT]firewall_driver option https://review.opendev.org/668139 | 07:07 |
mnasiadka | shyamb: use template override to install it - build your own image | 07:15 |
shyamb | mnasiadka: yes, thanks | 07:16 |
*** threestrands has quit IRC | 07:19 | |
*** lemko has joined #openstack-kolla | 07:21 | |
openstackgerrit | Merged openstack/kolla-ansible master: Use default BGP speaker driver in bgp_dragent.ini https://review.opendev.org/675111 | 07:31 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Use fluentd image labels https://review.opendev.org/676413 | 07:32 |
*** dpawlik has quit IRC | 07:34 | |
*** rpittau|afk is now known as rpittau | 07:35 | |
hrw | morning | 07:41 |
yoctozepto | morning | 07:43 |
*** ivve has joined #openstack-kolla | 07:45 | |
hrw | yoctozepto: commented your comment | 07:46 |
hrw | fluentd-- | 07:46 |
yoctozepto | hrw: can see nothing new :/ | 07:47 |
hrw | yoctozepto: ah, not noticed your comment to my comment to your comment ;D | 07:48 |
yoctozepto | http://www.quickmeme.com/img/19/19df7503a5dd5885a309b94c48c561de832f45726a77dac00b2d518317694b78.jpg | 07:49 |
hrw | ;D | 07:49 |
hrw | https://review.opendev.org/#/c/677063 argh | 07:52 |
hrw | yoctozepto: can you suggest patch which removes that k8s module? so we can get fluentd building again in my patch and then remove k8s stuff in other | 07:53 |
hrw | in case someone needs it | 07:54 |
hrw | I am one step from patch to make tripleo CI non-voting | 07:59 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: CI: make TripleO non voting https://review.opendev.org/677136 | 08:02 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Add a spec for ceph-ansible https://review.opendev.org/544980 | 08:02 |
*** dougsz has joined #openstack-kolla | 08:02 | |
hrw | done | 08:03 |
hrw | dougsz: elo | 08:03 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Add a spec for ceph-ansible https://review.opendev.org/544980 | 08:04 |
mnasiadka | hrw: only one step away? :) | 08:04 |
mnasiadka | morning guys | 08:05 |
hrw | mnasiadka: ;D | 08:05 |
hrw | mnasiadka: tripleo sorts out their issue for several days now. and it blocks whole work in our project | 08:06 |
mnasiadka | hrw: yeah, I know :) | 08:06 |
mnasiadka | hrw: it's funny that a RedHat employee proposes that :) | 08:06 |
hrw | mnasiadka: Red Hat - we have space in name ;d | 08:06 |
stingrayza | morning k-a folks :) happy monday | 08:10 |
*** skramaja has joined #openstack-kolla | 08:12 | |
*** skramaja_ has quit IRC | 08:12 | |
hrw | stingrayza: elo | 08:13 |
hrw | yoctozepto: and that fluentd fix needs to go to stein too | 08:15 |
hrw | and probably to other branches too | 08:15 |
*** shyamb has quit IRC | 08:24 | |
hrw | fuck | 08:27 |
yoctozepto | hrw: all right, fluentd here I come | 08:27 |
hrw | in short: we are fucked when it comes to CI | 08:27 |
hrw | or we join fluentd fix with disabling voting for tripleo or we wait for tripleo to fix their issue | 08:28 |
hrw | #$Tu324789t7389065720! | 08:28 |
mnasiadka | hrw: wonder what the name will be in 3-4 years :D | 08:30 |
hrw | mnasiadka: will see | 08:30 |
mnasiadka | hrw: btw, I think tripleo doesn't use fluentd :) | 08:31 |
hrw | mnasiadka: fluentd broke ubuntu CI jobs. | 08:31 |
mnasiadka | hrw: you mean activesomething plugin for fluent? :) | 08:32 |
hrw | yes | 08:32 |
hrw | s/plugin/gem | 08:32 |
hrw | as it is gem used by one of plugins | 08:32 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla master: Unbreak CI: fluentd and TripleO dependency https://review.opendev.org/677144 | 08:32 |
mnasiadka | hrw: as yoctozepto probably proposed, let's dump k8s plugins in fluentd, we retired kolla-k8s long time ago | 08:32 |
yoctozepto | mnasiadka: yeah, let's :-) | 08:32 |
yoctozepto | same second | 08:32 |
yoctozepto | tripleo has some issue with td-agent atm | 08:33 |
mnasiadka | yoctozepto: the same comes with all those kubernetes related images, we probably should deprecate them as soon as possible :) | 08:33 |
yoctozepto | so they ain't using it but it breaks them | 08:33 |
hrw | yoctozepto: +2+2 | 08:33 |
yoctozepto | ff logic :D | 08:33 |
yoctozepto | mnasiadka: yeah, kubes should be gone | 08:33 |
yoctozepto | I think there is some note about them being obsoleted | 08:33 |
yoctozepto | will have to look | 08:33 |
yoctozepto | nah, there is none | 08:35 |
yoctozepto | but they are not used in k-a at all | 08:35 |
yoctozepto | seems the deployment implementation never really started off | 08:36 |
yoctozepto | or maybe it's in this wicked kolla-kubernetes repo | 08:36 |
yoctozepto | too afraid to check | 08:36 |
yoctozepto | mnasiadka: https://review.opendev.org/677144 | 08:36 |
yoctozepto | https://review.opendev.org/#/q/project:openstack/kolla-kubernetes | 08:37 |
yoctozepto | scary | 08:37 |
yoctozepto | https://review.opendev.org/565221 <- this what I saw :D | 08:38 |
yoctozepto | hrw: hopefully only that unbound kubernetes plugins is causing ruby issues | 08:39 |
hrw | yoctozepto: yes. so far only this one | 08:39 |
hrw | checked | 08:39 |
yoctozepto | gr8 | 08:39 |
yoctozepto | dougsz gave some some fluentd-plugins-to-remove examples | 08:42 |
yoctozepto | https://review.opendev.org/#/c/675614/24/docker/fluentd/Dockerfile.j2@28 | 08:42 |
yoctozepto | some more* | 08:42 |
yoctozepto | hrw, mnasiadka: ^ | 08:42 |
yoctozepto | grok is definitely used atm | 08:43 |
yoctozepto | secure forward is not | 08:43 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible stable/stein: Use default BGP speaker driver in bgp_dragent.ini https://review.opendev.org/677146 | 08:43 |
*** Blinkiz has joined #openstack-kolla | 08:56 | |
Blinkiz | Hello. To automate installation of OpenStack from baremetal, is bifrost recommended? Is that project going strong? Or should I put my time somewhere else? | 08:58 |
Blinkiz | It seems like bifrost deploys ubuntu by default. It says so in documentation. I have gotten the feeling that kolla prefers centos. Right? | 08:59 |
yoctozepto | Blinkiz: kayobe deliverable guys, including our PTL, mgoddard, consider bifrost the right path | 09:00 |
yoctozepto | kayobe is kolla-on-bifrost | 09:00 |
Blinkiz | yoctozepto, thank you for the guidance :) | 09:01 |
yoctozepto | Blinkiz: kolla supports ubuntu and centos equally | 09:01 |
yoctozepto | since some weeks ago kayobe is part of kolla project | 09:02 |
yoctozepto | so this channel is proper place to discuss it too | 09:02 |
Blinkiz | alright. thanks | 09:02 |
yoctozepto | just the relevant folks need to show up | 09:02 |
openstackgerrit | Mark Goddard proposed x/kayobe-config stable/stein: Update .gitreview for Stein https://review.opendev.org/677151 | 09:03 |
openstackgerrit | Mark Goddard proposed x/kayobe-config stable/stein: Update upper constraints for Stein https://review.opendev.org/677152 | 09:03 |
yoctozepto | personally I did not use bifrost due to time constraints | 09:03 |
*** dpawlik has joined #openstack-kolla | 09:03 | |
yoctozepto | the issue was with ipa not picking up my glorious fujitsu systems properly | 09:03 |
yoctozepto | add infiniband driver support to that and you get: let's use what works already | 09:04 |
yoctozepto | :D | 09:04 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla master: Add fluentd_binary and fluentd_version labels https://review.opendev.org/676411 | 09:05 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Use fluentd image labels https://review.opendev.org/676413 | 09:06 |
Blinkiz | yeah. I got you. Mellanox has started to support Ubuntu 18.04 since a few weeks back for ASAP2. Accelerated Switchning and packet processing. So as long as openvswitch also support Ubuntu, then ubuntu is good to go for all my dependencies. My colleagues here at work prefer ubuntu so I may use that instead of centos. | 09:07 |
mnasiadka | yoctozepto: so seems like fluentd needs a bit more love :) | 09:08 |
yoctozepto | Blinkiz: in here ceph runs over infiniband as this is our fastest network, the stock driver in centos had issues so we replace it with upstream's | 09:11 |
yoctozepto | which works pretty well | 09:12 |
Blinkiz | I see.. | 09:12 |
yoctozepto | and the process is already prepared in foreman | 09:12 |
yoctozepto | so bifrost would be no added value atm | 09:12 |
yoctozepto | but added baggage ;p | 09:12 |
yoctozepto | mnasiadka: yeah, like all kolla | 09:13 |
yoctozepto | ubuntu unbroken | 09:13 |
Blinkiz | I have previous wanted to use Foreman but after seeing a youtube video about Foreman I turned the other way. It seems over complicated, almost bloated. | 09:13 |
stingrayza | I think if you have hundreds of hosts running a few set templates, foreman is great. but it takes some time to get set up and is probably overkill in smaller deployments. | 09:16 |
stingrayza | I also don't like the fact that it is mostly puppet based, when I'm trying to ansible the world | 09:17 |
stingrayza | (I know you can extend it to do some ansible stuff, but the core is still puppet) | 09:17 |
Blinkiz | I have a question about Bifrost. On there product page it says that Ubuntu is supported up to 16.04. Is this correct? Where is 18.04 support? | 09:17 |
yoctozepto | stingrayza: the project "kill foreman in UCO" is officially approved by me | 09:17 |
yoctozepto | (UCO is my unit at university) | 09:17 |
*** janki has quit IRC | 09:17 | |
yoctozepto | so don't worry | 09:18 |
yoctozepto | it's there because it works | 09:18 |
stingrayza | XD XD | 09:18 |
yoctozepto | and the first to dump otherwise | 09:18 |
Blinkiz | On Kayobe page it says Centos 7.6 and Ubuntu 16.04 is supported. So why not Ubuntu 18.04? | 09:19 |
Blinkiz | I mean, has development halted? | 09:19 |
yoctozepto | Blinkiz: I read their sources on, like, Saturday? and they have Ubuntu 18.04 in there | 09:19 |
*** tonythomas has joined #openstack-kolla | 09:19 | |
yoctozepto | just no idea how well supported | 09:19 |
hrw | Blinkiz: Ubuntu development? maybe | 09:19 |
yoctozepto | Blink: just try, let them know | 09:20 |
yoctozepto | let us (kayobe) know | 09:20 |
yoctozepto | our power comes from community :-) | 09:20 |
Blinkiz | :) | 09:20 |
openstackgerrit | Mark Goddard proposed x/kayobe-config-dev stable/stein: Update upper constraints for stein https://review.opendev.org/676987 | 09:23 |
openstackgerrit | Mark Goddard proposed x/kayobe-config stable/stein: Update upper constraints for Stein https://review.opendev.org/677152 | 09:23 |
openstackgerrit | Scott Solkhon proposed openstack/kolla master: Add effective and permitted capability to blackbox exporter https://review.opendev.org/677156 | 09:27 |
openstackgerrit | Merged openstack/kolla-ansible master: CI - remove unused required-projects https://review.opendev.org/666634 | 10:03 |
*** shyamb has joined #openstack-kolla | 10:11 | |
*** hamdyk has joined #openstack-kolla | 10:13 | |
*** dougsz has quit IRC | 10:17 | |
openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: CI: Test accessing dashboard https://review.opendev.org/676412 | 10:20 |
*** hjensas has joined #openstack-kolla | 10:24 | |
yoctozepto | I thought Zull would finish checking that unbreak by now but it seems it went crazy and started jobs it shouldn't | 10:28 |
yoctozepto | http://zuul.openstack.org/status | 10:28 |
yoctozepto | it queued all jobs we have | 10:28 |
openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Set default timeout to 60 seconds for docker stop https://review.opendev.org/669897 | 10:29 |
*** Wasaac has quit IRC | 10:30 | |
*** Wasaac has joined #openstack-kolla | 10:31 | |
*** dougsz has joined #openstack-kolla | 10:35 | |
*** shyamb has quit IRC | 10:36 | |
*** skramaja has quit IRC | 10:39 | |
openstackgerrit | Radosław Piliszek proposed openstack/kolla-ansible master: Set default timeout to 60 seconds for docker stop https://review.opendev.org/669897 | 10:50 |
openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: Set default timeout to 60 seconds for docker stop https://review.opendev.org/669897 | 10:54 |
*** shyamb has joined #openstack-kolla | 10:58 | |
*** dave-mccowan has joined #openstack-kolla | 11:06 | |
*** rohit02 has joined #openstack-kolla | 11:08 | |
*** hjensas has quit IRC | 11:08 | |
rohit02 | i am facinf error while configuring multinode kolla-ansible ""msg": "The conditional check 'kolla_external_vip_interface not in ansible_interfaces' failed." | 11:09 |
*** gfidente has joined #openstack-kolla | 11:10 | |
rohit02 | has anyone face this issue earlier | 11:11 |
*** hjensas has joined #openstack-kolla | 11:16 | |
*** dougsz has quit IRC | 11:17 | |
*** kplant has joined #openstack-kolla | 11:20 | |
yoctozepto | rohit02: make sure the interface you specified is present on controller nodes you added to inventory | 11:20 |
yoctozepto | specifically kolla_external_vip_interface | 11:20 |
yoctozepto | it's a variable you should have set in globals.yml | 11:20 |
rohit02 | i have alrady set in global.yml file | 11:31 |
rohit02 | complete error is "Unable to look up a name or access an attribute in template string ({% if kolla_external_vip_interface not in ansible_interfaces %} True {% else %} False {% endif %}).\nMake sure your variable name does not contain invalid characters like '-': argument of type 'StrictUndefined' is not iterable\n\nThe error appears to have been in '/usr/local/share/kolla-ansible/ansible/roles/haproxy/tasks/precheck.yml': line 85, column 3, but | 11:32 |
rohit02 | may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name: Checking the kolla_external_vip_interface is present\n ^ here\n" | 11:32 |
yoctozepto | ah, that makes it a different thingy | 11:36 |
yoctozepto | does your interface have '-' in its name? | 11:37 |
openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: CI: Test accessing dashboard https://review.opendev.org/676412 | 11:42 |
mnasiadka | yoctozepto: can we just have the tripleo non-voting & removing from gating change as a separate one, so it's easy to revert it when they start behaving? | 11:48 |
hrw | mnasiadka: nope | 11:49 |
rohit02 | no '-' in interface name.....interface name is ens192 | 11:49 |
hrw | mnasiadka: we have two independent failures | 11:49 |
mnasiadka | damn | 11:49 |
*** janki has joined #openstack-kolla | 11:50 | |
hrw | mnasiadka: welcome in 'fetch random binary and ship' world | 11:50 |
mnasiadka | hrw: well, blame ruby guys that they don't restrict their gems to check what ruby version you have installed... | 11:52 |
mgoddard | Blinkiz: re kayobe, we support centos on the control plane hosts. Ubuntu is supported only for the host on which you run kayobe. We should update as I'm sure it would work on more recent ubuntu than 16.04 | 11:52 |
mgoddard | Blinkiz: if you're interested in trying it out, let me know | 11:53 |
mgoddard | Blinkiz: this is an easy way to get started in a test environment: https://github.com/stackhpc/a-universe-from-nothing | 11:53 |
hrw | mnasiadka: ? pip does not check either | 11:53 |
hrw | mnasiadka: the only to blame imho are td-agent guys for shipping old ruby | 11:54 |
mnasiadka | hrw: well true | 11:54 |
yoctozepto | hrw, mnasiadka: then blame both | 11:54 |
hrw | yoctozepto: can I blame whoever used td-agent instead of fluentd? | 11:55 |
*** dougsz has joined #openstack-kolla | 11:55 | |
Blinkiz | mgoddard, thank you. :) | 11:55 |
yoctozepto | hrw: or fluentd instead of td-agent :-) | 11:56 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla master: Add fluentd_binary and fluentd_version labels https://review.opendev.org/676411 | 11:56 |
mnasiadka | hrw: as long as OS distro maintainers don't give a shite about Ruby - we need to use td-agent :D | 11:57 |
mnasiadka | at least in CentOS case | 11:57 |
hrw | mnasiadka: Debian and Ubuntu ship Ruby 2.5+ | 11:57 |
mnasiadka | hrw: so why use td-agent there instead of fluentd? :) | 11:57 |
yoctozepto | re: many unnecessary jobs | 11:57 |
yoctozepto | <frickler> yoctozepto: 2019-08-19 10:04:33,230 DEBUG zuul.layout: [e: d383ebbbfc1248a3b59af867de2eaba2] The configuration of job <Job kolla-ansible-ubuntu-source-ironic branches: {MatchAny:{BranchMatcher:master}} source: opendev/base-jobs/zuul.d/jobs.yaml@master#25> is changed by <Change 0x7f185d716908 openstack/kolla 677144,1>; ignoring file matcher | 11:57 |
yoctozepto | <frickler> so it seems that indeed the reconfiguraton triggered by 666634 causes new jobs to be added because the file matcher is ignored in this special situation | 11:58 |
hrw | mnasiadka: Debian uses fluentd only | 11:58 |
hrw | mnasiadka: https://review.opendev.org/676909 waits for working CI | 11:58 |
yoctozepto | guys, are not we overengineering it - just build modern ruby toolkits in those containers | 11:58 |
yoctozepto | that's what they are for anyways | 11:59 |
mnasiadka | hrw: debian x86_64 uses td-agent :) | 11:59 |
hrw | mnasiadka: no, it just was not built so no one noticed ;D | 11:59 |
yoctozepto | scrap "ruby" meant any | 11:59 |
mnasiadka | yoctozepto: well, that's against some common logic in this project :) | 11:59 |
mnasiadka | yoctozepto: building something that distro should deliver :) | 12:00 |
yoctozepto | we should rethink the principles then ;-) | 12:00 |
*** weshay_pto is now known as weshay | 12:00 | |
*** lemko has quit IRC | 12:00 | |
hrw | heh... distro deliver... | 12:00 |
hrw | when hell froze and pigs will fly | 12:01 |
mnasiadka | I'm not saying we should not - just saying what was the approach until now :) | 12:01 |
*** henriqueof has joined #openstack-kolla | 12:01 | |
hrw | mnasiadka: when I joined Kolla it was even stronger | 12:02 |
mnasiadka | hrw: sometimes all those if/elif/endif statements 700 hundred times in Dockerfile.j2 give me headaches :) | 12:02 |
mnasiadka | hrw: especially in fluentd Dockerfile :) | 12:02 |
hrw | mnasiadka: I think that we should be able now to remove all those "sorry, no binaries for this distro/arch/buildtype" things as we have UNBUILDABLE_IMAGES now | 12:03 |
mnasiadka | yeah | 12:03 |
mnasiadka | this is an infinite journey... | 12:03 |
mnasiadka | yoctozepto: the number of jobs ignited by the CI unblocker is crazy ;) | 12:06 |
yoctozepto | mnasiadka: they appeared later due to merge in k-a repo | 12:06 |
yoctozepto | Zuul quirk | 12:06 |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla master: Include iputils package to tempest container https://review.opendev.org/677191 | 12:07 |
yoctozepto | mnasiadka, hrw: I don't know if we shouldn't satisfy AJaeger in a follow-up | 12:11 |
yoctozepto | because it takes time to check all jobs | 12:11 |
yoctozepto | and the tripleo job finishes quickly with that failure | 12:12 |
hrw | yoctozepto: add a comment | 12:12 |
mnasiadka | yoctozepto: just do it, it's like over 3 hours for a CI run :) | 12:13 |
*** vmixor has joined #openstack-kolla | 12:13 | |
yoctozepto | hrw, mnasiadka: guys, you want me to make it run again or not because your signals were unclear :D | 12:13 |
hrw | yoctozepto: add a comment for AJaeger and merge | 12:14 |
yoctozepto | hrw: you mean as in in-gerrit reply? | 12:15 |
hrw | yes | 12:15 |
hrw | 16 files changed, 28 insertions(+), 160 deletions(-) | 12:16 |
hrw | small docker/ cleanup | 12:16 |
hrw | in progress | 12:16 |
hrw | 26 files changed, 71 insertions(+), 285 deletions(-) | 12:20 |
hrw | 45 files changed, 116 insertions(+), 487 deletions(-) | 12:28 |
yoctozepto | geeez, it grows :O | 12:29 |
yoctozepto | hrw, mnasiadka: https://review.opendev.org/677144 | 12:30 |
yoctozepto | Zuul is finishing | 12:30 |
yoctozepto | please slap w+1 when done | 12:30 |
yoctozepto | my comment is there | 12:30 |
yoctozepto | i will be commuting for some time now | 12:30 |
yoctozepto | till later | 12:30 |
hrw | ok | 12:31 |
hrw | +W given | 12:36 |
goldyfruit | o/ | 12:36 |
*** spsurya has quit IRC | 12:43 | |
hrw | 82 files changed, 167 insertions(+), 713 deletions(-) | 12:43 |
hrw | that will be fun to review | 12:47 |
*** shyamb has quit IRC | 12:48 | |
hrw | let push | 12:49 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: Cleanup: drop checks handled by UNBUILDABLE_IMAGES https://review.opendev.org/677194 | 12:50 |
hrw | in reality it is only 9 added lines - rest are indents | 12:55 |
*** dougsz has quit IRC | 12:59 | |
*** dougsz has joined #openstack-kolla | 13:03 | |
openstackgerrit | Merged openstack/kolla master: Unbreak CI: fluentd and TripleO dependency https://review.opendev.org/677144 | 13:18 |
hrw | YES! | 13:19 |
hrw | 6 rechecks sent | 13:19 |
*** janki has quit IRC | 13:20 | |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla stable/stein: Unbreak CI: fluentd and TripleO dependency https://review.opendev.org/677202 | 13:20 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla stable/stein: Unbreak CI: fluentd and TripleO dependency https://review.opendev.org/677202 | 13:23 |
hrw | yoctozepto: tripleo in gate dropped | 13:24 |
*** rohit02 has quit IRC | 13:27 | |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla master: Include iputils package to tempest container https://review.opendev.org/677191 | 13:36 |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla master: Include iputils package to tempest container https://review.opendev.org/677191 | 13:37 |
*** BjoernT has joined #openstack-kolla | 13:52 | |
mgoddard | sweet | 13:54 |
hrw | mgoddard: take a look at "Cleanup: drop checks handled by UNBUILDABLE_IMAGES" https://review.opendev.org/677194 one | 13:55 |
hrw | without even commenting in review - just would like to see first opinion | 13:56 |
mgoddard | hrw: I think it makes sense. Only downside is that it's no longer obvious from the Dockerfile that e.g binary support is lacking | 14:03 |
*** dougsz has quit IRC | 14:05 | |
mnasiadka | mgoddard: then it's the famous support matrix :) | 14:05 |
mgoddard | true | 14:05 |
mnasiadka | I'll start it tomorrow | 14:05 |
Blinkiz | Anyone here know about a script/project/container that can automatically get TLS certificate from lets Encrypt every ~third month? Or is it something I need to code together myself in ansible? | 14:11 |
hrw | Blinkiz: certbot you mean? in cron? | 14:11 |
Blinkiz | And of course work with kolla-Ansible. Meaning giving haproxy the right cert(s) on all control nodes | 14:12 |
Blinkiz | hrw, certbot/cron is on a single machine, kolla-ansible control nodes is many machines. So.. yeah.. asking for guidance if exist any existing tool. | 14:14 |
openstackgerrit | Isaac Prior proposed openstack/kolla-ansible master: Removes monasca_grafana persistent volume https://review.opendev.org/677211 | 14:14 |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla stable/stein: Use openstack-tempest-all for rpm tempest container https://review.opendev.org/677212 | 14:16 |
*** cah_link has joined #openstack-kolla | 14:16 | |
*** dpawlik has quit IRC | 14:20 | |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla stable/stein: Use openstack-tempest-all for rpm tempest container https://review.opendev.org/677212 | 14:20 |
openstackgerrit | Isaac Prior proposed openstack/kolla-ansible master: Removes monasca_grafana persistent volume https://review.opendev.org/677211 | 14:21 |
*** KeithMnemonic has joined #openstack-kolla | 14:25 | |
*** dougsz has joined #openstack-kolla | 14:26 | |
*** cah_link has quit IRC | 14:27 | |
hrw | Blinkiz: and thanks for reminding - cleaned up my letsencrypt certs | 14:29 |
kplant | that'd be a cool integration, kolla and letsencrypt via certbot | 14:30 |
yoctozepto | hrw: thx | 14:34 |
*** hamdyk has quit IRC | 14:34 | |
*** BjoernT_ has joined #openstack-kolla | 14:45 | |
yoctozepto | mgoddard: replied re that proposed bifrost change | 14:47 |
*** BjoernT has quit IRC | 14:47 | |
Blinkiz | Does OpenStack/Kolla have a way to distribute files between control nodes? Am planning my implementation of Let's Encrypt certificate. :) | 14:49 |
Blinkiz | I mean.. not by running Kolla deploy. I mean automatically. | 14:50 |
openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: follow-up on TripleO demotion - no gating https://review.opendev.org/677225 | 14:51 |
yoctozepto | Blinkiz: you can have custom mounted docker volumes pointing to host directory | 14:54 |
yoctozepto | use ansible to deploy your certs | 14:54 |
yoctozepto | remember to signal to haproxy they got changed | 14:54 |
Blinkiz | yoctozepto, yeah, but which control node is master in keepalived? Only that one can get a new cert with ACME http -01challenge response. | 14:55 |
kplant | yoctozepto: could probably `docker restart haproxy` no? | 14:55 |
yoctozepto | kplant: downtime? | 14:55 |
kplant | is an haproxy restart that noticeable? | 14:56 |
kplant | well.. a successful haproxy restart :p | 14:56 |
yoctozepto | Blinkiz: ah, you do it this way, I see, hmm | 14:56 |
kplant | Blinkiz: whichever node has the VIP is the current 'master' | 14:56 |
yoctozepto | kplant: if you have requests coming at all times... | 14:56 |
yoctozepto | so yes | 14:56 |
kplant | you could also snipe VRRP and learn it that way | 14:56 |
yoctozepto | easier to check the VIP | 14:57 |
yoctozepto | ansible playbook run against all the machines | 14:57 |
yoctozepto | getting facts | 14:57 |
yoctozepto | set which has the VIP address | 14:57 |
Blinkiz | yoctozepto, I can for my own domain use DNS verification but customers I do not have control over. Am planning to just ask them to point there subdomain to a A record. With this I can get certificates for customers domains if I use http-01 challenge response | 14:57 |
yoctozepto | delegate certbot to it | 14:57 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: Enable TripleO CI as voting one again https://review.opendev.org/677228 | 14:58 |
yoctozepto | Blinkiz: me and kplant prepared a solution for you :-) | 14:58 |
hrw | so we will not forget to enable it back ;D | 14:58 |
hrw | yoctozepto: it is on top of your one | 14:58 |
Blinkiz | yoctozepto, yeah, it is possible if I run certbot (logic) on a management node outside the cluster. Am thinking about a way to let the cluster solve it | 14:59 |
yoctozepto | hrw: good thinking | 15:00 |
Blinkiz | But the more I think about it the more I see that it needs to be an outside machine solving the renewal of tls certificates. | 15:00 |
Blinkiz | probably my deploy machine where kolla-ansible is running from | 15:00 |
yoctozepto | Blinkiz: nah, it can run from inside the cluster | 15:00 |
hrw | https://review.opendev.org/#/c/674240/ (Debian CI) got green from Zuul. +2 +W please | 15:01 |
yoctozepto | you could orchestrate all crons to check if they are master | 15:01 |
yoctozepto | and only then proceed | 15:01 |
hrw | it's deps are in zuul | 15:01 |
Blinkiz | yoctozepto, okay. What is your suggestion how I get the new certificate to the control nodes that is not currently master? | 15:01 |
yoctozepto | cloudnull: how bad is that tripleo ci issue - I see you, folks, are rechecking and rechecking but it fails on same jobs | 15:01 |
yoctozepto | fyi, we disable tripleo in our gate to unblock ourselves but we are staying vigilant | 15:02 |
cloudnull | its a frustrating one. not related to that change. | 15:02 |
yoctozepto | Blinkiz: rsync | 15:02 |
cloudnull | it may be worth making that job non-voting | 15:02 |
yoctozepto | cloudnull: yeah, I see it's more general | 15:02 |
cloudnull | once we get stable again we can re-enable it | 15:03 |
yoctozepto | cloudnull: ping us please if we miss it | 15:03 |
cloudnull | ++ | 15:03 |
yoctozepto | hrw prepared the tripleo glory restoration patch | 15:03 |
cloudnull | I'll make a note to make that job voting again sometime in the after-now | 15:03 |
yoctozepto | cloudnull: https://review.opendev.org/677228 | 15:03 |
yoctozepto | you can recheck that once it is fixed | 15:04 |
yoctozepto | we care for your comfort | 15:04 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Use fluentd image labels https://review.opendev.org/676413 | 15:04 |
cloudnull | +1 | 15:04 |
yoctozepto | your friendly kolla team | 15:04 |
cloudnull | thank you yoctozepto sorry for the issues | 15:04 |
yoctozepto | no problem, we all make mistakes from time to time | 15:04 |
openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Use fluentd image labels https://review.opendev.org/676413 | 15:07 |
yoctozepto | hrw: did you try building bifrost with that mariadb change | 15:10 |
yoctozepto | ? | 15:10 |
hrw | yoctozepto: not | 15:18 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Copy Nova role as a basis for the Nova cell role https://review.opendev.org/675658 | 15:18 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: [WIP] Factor out nova-cell role https://review.opendev.org/675659 | 15:18 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Copy Nova role as a basis for the Nova cell role https://review.opendev.org/675658 | 15:19 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: [WIP] Factor out nova-cell role https://review.opendev.org/675659 | 15:19 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Minimal support for deploying multiple instances of MariaDB https://review.opendev.org/619756 | 15:22 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Copy Nova role as a basis for the Nova cell role https://review.opendev.org/675658 | 15:22 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: [WIP] Factor out nova-cell role https://review.opendev.org/675659 | 15:22 |
*** factor has quit IRC | 15:32 | |
*** factor has joined #openstack-kolla | 15:32 | |
*** ivve has quit IRC | 15:33 | |
*** factor has quit IRC | 15:40 | |
*** factor has joined #openstack-kolla | 15:41 | |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla master: Include iputils package to tempest container https://review.opendev.org/677191 | 15:43 |
*** shyamb has joined #openstack-kolla | 15:47 | |
openstackgerrit | Merged openstack/kolla master: Fixes wrong local_setting.py file issue for Ubuntu binary deployment. https://review.opendev.org/668409 | 15:50 |
hrw | mnasiadka: can you rebase https://review.opendev.org/#/c/675614 patch? will be easier to probably rebase https://review.opendev.org/#/c/675614/ on top | 16:00 |
hrw | https://review.opendev.org/676909 I meant (at the end) | 16:00 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: debian: add some images to unbuildable list https://review.opendev.org/676905 | 16:01 |
hrw | yoctozepto: your comment accepted ^^ | 16:01 |
mnasiadka | hrw: yeah, will rebase first thing tomorrow morning, I'm rather focused now on https://review.opendev.org/#/c/676413/9 and the dependent change :) | 16:03 |
hrw | sure | 16:03 |
*** gfidente has quit IRC | 16:03 | |
hrw | mnasiadka: I think that after merging all those fluentd changes it will look like new... so much rewriting | 16:04 |
mnasiadka | hrw: might be :) | 16:05 |
*** rpittau is now known as rpittau|afk | 16:07 | |
hrw | if rpm then td-agent unless !x86 elif deb then fluentd unless (ubuntu x86);fi | 16:07 |
*** icarusfactor has joined #openstack-kolla | 16:07 | |
yoctozepto | hrw: like poop sprinkled with perfumes you mean? ;-) | 16:07 |
hrw | yoctozepto: s/sprinkled with/drawn in/ | 16:08 |
yoctozepto | agreed | 16:08 |
hrw | mnasiadka: will steal your fluentd_user stuff to https://review.opendev.org/#/c/676909/ | 16:08 |
vmixor | Hi, I have created some changes to kolla and kolla-ansible to support deploy of Stein on Debian Buster. I used code from master branches and packages from Stein backports (http://buster-stein.debian.net/debian). Now I have deployment in quite operational state, I think. I have been instructed to try push my changes to upstream. | 16:08 |
vmixor | Therefore I would be glad for some advice and/or guidance about next steps I should do. | 16:08 |
*** factor has quit IRC | 16:09 | |
yoctozepto | hrw, vmixor: hereby I link you with the debian bound | 16:09 |
*** henriqueof has quit IRC | 16:09 | |
hrw | vmixor: the problem is that those are Stein packages. so code should go to stable/stein. but that's also new feature while we do not backport features | 16:09 |
hrw | yoctozepto: I asked vmixor to bring it to public channel | 16:10 |
yoctozepto | ah, understood | 16:10 |
yoctozepto | I don't mind backporting to the latest stable as long as we can be *sure* we are not affecting stability | 16:11 |
yoctozepto | I feel deployment projects should not really fully freeze their deployment solution for the latest stable | 16:11 |
hrw | vmixor: send patches to gerrit | 16:12 |
yoctozepto | as master deploying master is probably fencing off many people from use | 16:12 |
mgoddard | I don't think we should get into the habit of backporting features | 16:12 |
yoctozepto | vmixor, hrw: yeah, master will accept them probably | 16:12 |
mgoddard | if you need a new feature on an old branch, maintain a fork | 16:13 |
hrw | mgoddard: agreed | 16:13 |
yoctozepto | ok then, I'm outvoted :D | 16:13 |
yoctozepto | then debian is train+ only | 16:13 |
mgoddard | stable branches should be stable | 16:13 |
hrw | vmixor: I more look at those patches as 'let us see how good Debian/binary target works' | 16:13 |
yoctozepto | hrw: you did not mention binary | 16:14 |
yoctozepto | till now | 16:14 |
hrw | yoctozepto: vmixor has Debian/binary with Stein packages. | 16:14 |
yoctozepto | I see | 16:14 |
yoctozepto | hmm | 16:14 |
vmixor | yoctozepto, hrw: yes, I have used binary packages from Stein backports for Buster | 16:15 |
hrw | yoctozepto: most of work can be merged and kept in Train so will be useful when Train packages arrive | 16:15 |
yoctozepto | vmixor: is debian planning to support train soon? are you building some packages off master branches? | 16:16 |
hrw | yoctozepto: most of packages are around RC or after release | 16:16 |
yoctozepto | we have such packages for centos iirc | 16:16 |
yoctozepto | not sure about ubuntu though | 16:16 |
yoctozepto | I see | 16:16 |
hrw | yoctozepto: ubuntu and centos do builds from git. Debian OpenStack team iirc not | 16:16 |
yoctozepto | but could they? that would make it easier to introduce k-a debian-binary | 16:17 |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: fluentd: there is no 'td-agent' for Debian/x86-64 https://review.opendev.org/676909 | 16:17 |
yoctozepto | so far there don't seem to be any breaking changes between stein and train noticed so it might be only a hypothetical issue | 16:18 |
hrw | yoctozepto: ask on #debian-openstack @oftc? | 16:18 |
yoctozepto | hrw: too lazy ;-) | 16:18 |
*** igordc has joined #openstack-kolla | 16:18 | |
mnasiadka | hrw: sure | 16:20 |
hrw | vmixor: once you send a patch I will review, comment | 16:21 |
vmixor | hrw: ok, I will create patchset for current master | 16:23 |
openstackgerrit | Merged openstack/kolla stable/stein: Unbreak CI: fluentd and TripleO dependency https://review.opendev.org/677202 | 16:30 |
yoctozepto | hrw, mnasiadka, mgoddard: https://review.opendev.org/677225 and be done with this stuff for now (master & stein) | 16:32 |
yoctozepto | (stein just merged) | 16:32 |
*** shyamb has quit IRC | 16:43 | |
hrw | vmixor: thanks | 16:46 |
vmixor | hrw: thank you too. | 16:49 |
vmixor | hrw: one more question: Is threre some blueprint for Stein support for Debian? | 16:50 |
*** jonaspaulo has joined #openstack-kolla | 16:54 | |
openstackgerrit | Merged openstack/kolla master: Add pciutils package to cyborg-agent and zun-compute https://review.opendev.org/676932 | 17:01 |
*** henriqueof has joined #openstack-kolla | 17:17 | |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: [WIP] Factor out nova-cell role https://review.opendev.org/675659 | 17:19 |
*** raukadah has joined #openstack-kolla | 17:24 | |
raukadah | Hello Kolla dev! | 17:24 |
openstackgerrit | James Kirsch proposed openstack/kolla-ansible master: Implement TLS encryption for internal endpoints https://review.opendev.org/663555 | 17:26 |
openstackgerrit | Chandan Kumar (raukadah) proposed openstack/kolla stable/stein: Use openstack-tempest-all for rpm tempest container https://review.opendev.org/677212 | 17:26 |
raukadah | please have a look at these reviews https://review.opendev.org/#/c/677191/ and https://review.opendev.org/#/c/677212/ | 17:26 |
raukadah | thanks@ | 17:26 |
openstackgerrit | Doug Szumski proposed openstack/kolla master: WIP: Support DB bootstrap from Nova Conductor container https://review.opendev.org/677263 | 17:36 |
openstackgerrit | Doug Szumski proposed openstack/kolla master: WIP: Support DB bootstrap from Nova Conductor container https://review.opendev.org/677263 | 17:43 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: [WIP] Factor out nova-cell role https://review.opendev.org/675659 | 17:44 |
openstackgerrit | Merged openstack/kolla-ansible master: Removes monasca_grafana persistent volume https://review.opendev.org/677211 | 17:46 |
*** icarusfactor has quit IRC | 17:53 | |
hrw | vmixor: no | 18:25 |
* hrw out | 18:25 | |
hrw | see you in a week - will be back on 27th (yoctozepto, mnasiadka, mgoddard) | 18:26 |
mnasiadka | hrw: have fun ;) | 18:26 |
*** strigazi has quit IRC | 18:37 | |
*** strigazi has joined #openstack-kolla | 18:38 | |
*** dougsz has quit IRC | 19:03 | |
yoctozepto | hrw: take care ;-) | 19:06 |
*** ivve has joined #openstack-kolla | 19:09 | |
*** dougsz has joined #openstack-kolla | 19:21 | |
*** factor has joined #openstack-kolla | 19:22 | |
*** icarusfactor has joined #openstack-kolla | 19:44 | |
*** kplant has quit IRC | 19:44 | |
*** factor has quit IRC | 19:46 | |
*** dougsz has quit IRC | 20:29 | |
*** pomac has joined #openstack-kolla | 20:30 | |
*** vmixor has quit IRC | 20:38 | |
*** BjoernT_ has quit IRC | 21:36 | |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: Cleanup: drop checks handled by UNBUILDABLE_IMAGES https://review.opendev.org/677194 | 21:49 |
hrw | yoctozepto: if you have time then please rebase https://review.opendev.org/672700 (Docker: daemon.json). your zun changes broke it ;D | 21:49 |
* hrw -> out | 21:49 | |
*** jonaspaulo has quit IRC | 22:19 | |
*** tonythomas has quit IRC | 22:21 | |
*** dave-mccowan has quit IRC | 22:22 | |
*** dave-mccowan has joined #openstack-kolla | 22:25 | |
*** dave-mccowan has quit IRC | 22:32 | |
*** ivve has quit IRC | 22:34 | |
*** KeithMnemonic has quit IRC | 22:49 | |
*** EmilienM|pto has quit IRC | 22:54 | |
*** EmilienM has joined #openstack-kolla | 22:56 | |
*** KeithMnemonic has joined #openstack-kolla | 23:14 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!