*** sdake has quit IRC | 00:01 | |
*** vkmc has quit IRC | 00:02 | |
*** mattgo_ has joined #openstack-kolla | 00:02 | |
*** k_mouza has joined #openstack-kolla | 00:06 | |
*** sdake has joined #openstack-kolla | 00:06 | |
*** sdake has quit IRC | 00:06 | |
*** sdake has joined #openstack-kolla | 00:06 | |
*** mattgo has quit IRC | 00:06 | |
*** k_mouza has quit IRC | 00:06 | |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Refactor haproxy config (split by service) https://review.openstack.org/576361 | 00:06 |
---|---|---|
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Allow disabling insecure API endpoints https://review.openstack.org/548407 | 00:06 |
*** vkmc has joined #openstack-kolla | 00:06 | |
*** hawi_ has quit IRC | 00:06 | |
*** calbers_ has quit IRC | 00:06 | |
*** tch has quit IRC | 00:06 | |
*** dtk has quit IRC | 00:06 | |
*** mtudosoiu has quit IRC | 00:06 | |
*** dosaboy has quit IRC | 00:06 | |
*** gmann has quit IRC | 00:06 | |
*** zioproto has quit IRC | 00:06 | |
*** eandersson has quit IRC | 00:06 | |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Default redis_master_password in connection_string https://review.openstack.org/584161 | 00:10 |
*** mrostecki[m] has quit IRC | 00:10 | |
*** retr0h has quit IRC | 00:11 | |
*** bachp has quit IRC | 00:11 | |
*** mkrai has quit IRC | 00:11 | |
*** jgu has quit IRC | 00:11 | |
*** hawi_ has joined #openstack-kolla | 00:15 | |
*** calbers_ has joined #openstack-kolla | 00:15 | |
*** tch has joined #openstack-kolla | 00:15 | |
*** dtk has joined #openstack-kolla | 00:15 | |
*** mtudosoiu has joined #openstack-kolla | 00:15 | |
*** dosaboy has joined #openstack-kolla | 00:15 | |
*** gmann has joined #openstack-kolla | 00:15 | |
*** zioproto has joined #openstack-kolla | 00:15 | |
*** eandersson has joined #openstack-kolla | 00:15 | |
*** mliima has quit IRC | 00:15 | |
*** threestrands has joined #openstack-kolla | 00:23 | |
*** threestrands has quit IRC | 00:23 | |
*** threestrands has joined #openstack-kolla | 00:23 | |
openstackgerrit | Clint 'SpamapS' Byrum proposed openstack/kolla-ansible master: Add safe shutdown systemd jobs to tools https://review.openstack.org/584168 | 00:25 |
*** livelace2 has quit IRC | 00:28 | |
*** livelace2 has joined #openstack-kolla | 00:28 | |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Allow disabling insecure API endpoints https://review.openstack.org/548407 | 00:42 |
*** caoyuan has joined #openstack-kolla | 00:43 | |
*** mrostecki[m] has joined #openstack-kolla | 00:51 | |
*** retr0h[m] has joined #openstack-kolla | 00:59 | |
*** madhuri[m] has joined #openstack-kolla | 00:59 | |
*** PascalBach[m] has joined #openstack-kolla | 00:59 | |
*** dave-mccowan has quit IRC | 00:59 | |
*** ipsecguy_ has joined #openstack-kolla | 01:05 | |
*** ipsecguy has quit IRC | 01:07 | |
*** harlowja has quit IRC | 01:09 | |
*** imcsk8_ has quit IRC | 01:09 | |
*** imcsk8 has joined #openstack-kolla | 01:10 | |
*** caoyuan has quit IRC | 01:14 | |
*** liuzz has joined #openstack-kolla | 01:14 | |
*** phuoc_ has quit IRC | 01:27 | |
openstackgerrit | Min Sun proposed openstack/kolla-ansible master: Support nova dev mode in kolla https://review.openstack.org/584188 | 01:31 |
*** klindgren_ has quit IRC | 01:31 | |
*** hongbin_ has joined #openstack-kolla | 01:32 | |
openstackgerrit | Merged openstack/kolla stable/ocata: Fix table formatting issue in README.rst https://review.openstack.org/584111 | 01:32 |
*** mattgo_ has quit IRC | 01:33 | |
*** dave-mccowan has joined #openstack-kolla | 01:38 | |
*** fan_guiju has joined #openstack-kolla | 01:46 | |
*** dave-mccowan has quit IRC | 01:52 | |
*** Bhujay has joined #openstack-kolla | 01:52 | |
*** ipsecguy_ has quit IRC | 01:55 | |
*** itlinux has quit IRC | 01:56 | |
*** jgu has joined #openstack-kolla | 01:59 | |
openstackgerrit | Min Sun proposed openstack/kolla-ansible master: Support checkout dedicated version from git with dev mode https://review.openstack.org/584193 | 02:02 |
*** DongHM has joined #openstack-kolla | 02:02 | |
*** ipsecguy has joined #openstack-kolla | 02:03 | |
*** zhubingbing__ has quit IRC | 02:03 | |
*** caoyuan has joined #openstack-kolla | 02:05 | |
*** zhubingbing has joined #openstack-kolla | 02:05 | |
*** phuoc has joined #openstack-kolla | 02:07 | |
*** vpc has quit IRC | 02:29 | |
*** Bhujay has quit IRC | 02:30 | |
*** adrianreza has joined #openstack-kolla | 02:44 | |
*** unicell has quit IRC | 02:44 | |
*** threestrands has quit IRC | 02:54 | |
*** unicell has joined #openstack-kolla | 03:05 | |
*** unicell has quit IRC | 03:05 | |
*** skramaja has joined #openstack-kolla | 03:09 | |
*** threestrands has joined #openstack-kolla | 03:10 | |
*** threestrands has quit IRC | 03:10 | |
*** threestrands has joined #openstack-kolla | 03:10 | |
*** skramaja_ has joined #openstack-kolla | 03:14 | |
*** skramaja has quit IRC | 03:14 | |
*** skramaja_ is now known as skramaja | 03:29 | |
*** jgu has quit IRC | 03:36 | |
*** mdnadeem has joined #openstack-kolla | 04:05 | |
*** jgu has joined #openstack-kolla | 04:11 | |
*** Bhujay has joined #openstack-kolla | 04:14 | |
*** jgu_ has joined #openstack-kolla | 04:15 | |
*** jgu has quit IRC | 04:16 | |
*** jgu has joined #openstack-kolla | 04:18 | |
*** jgu_ has quit IRC | 04:21 | |
*** hongbin_ has quit IRC | 04:22 | |
*** harlowja has joined #openstack-kolla | 04:30 | |
*** unicell has joined #openstack-kolla | 04:32 | |
*** mdnadeem has quit IRC | 04:39 | |
*** jgu has quit IRC | 04:39 | |
*** threestrands has quit IRC | 04:40 | |
*** jgu has joined #openstack-kolla | 04:43 | |
*** jgu_ has joined #openstack-kolla | 04:46 | |
*** jgu has quit IRC | 04:48 | |
*** jgu has joined #openstack-kolla | 04:48 | |
*** jgu_ has quit IRC | 04:51 | |
*** harlowja has quit IRC | 04:52 | |
*** jgu_ has joined #openstack-kolla | 04:54 | |
*** jgu has quit IRC | 04:57 | |
*** jgu_ has quit IRC | 04:59 | |
*** jgu has joined #openstack-kolla | 05:04 | |
*** jgu has quit IRC | 05:09 | |
*** pcaruana has joined #openstack-kolla | 05:09 | |
*** jgu has joined #openstack-kolla | 05:13 | |
*** jgu_ has joined #openstack-kolla | 05:14 | |
*** jgu has quit IRC | 05:17 | |
*** jgu has joined #openstack-kolla | 05:21 | |
*** jgu_ has quit IRC | 05:23 | |
*** unicell1 has joined #openstack-kolla | 05:28 | |
*** jgu has quit IRC | 05:30 | |
*** unicell has quit IRC | 05:30 | |
*** jgu has joined #openstack-kolla | 05:34 | |
*** jgu has quit IRC | 05:43 | |
*** mdnadeem has joined #openstack-kolla | 05:47 | |
*** dmellado has quit IRC | 05:48 | |
*** cah_link has joined #openstack-kolla | 05:56 | |
*** lvdombrkr has joined #openstack-kolla | 06:22 | |
*** hjensas has joined #openstack-kolla | 06:25 | |
*** hjensas has quit IRC | 06:25 | |
*** hjensas has joined #openstack-kolla | 06:25 | |
*** dmellado has joined #openstack-kolla | 06:29 | |
*** janki has joined #openstack-kolla | 06:35 | |
*** egonzalez has joined #openstack-kolla | 06:44 | |
*** zhangfei has joined #openstack-kolla | 06:48 | |
*** lvdombrkr89 has joined #openstack-kolla | 06:49 | |
openstackgerrit | Chandan Kumar proposed openstack/kolla master: Added tempest user for tempest container https://review.openstack.org/583940 | 06:50 |
*** lvdombrkr has quit IRC | 06:51 | |
*** goutham has joined #openstack-kolla | 07:13 | |
*** DongHM has quit IRC | 07:15 | |
*** conz has joined #openstack-kolla | 07:20 | |
*** hjensas has quit IRC | 07:22 | |
*** peereb has joined #openstack-kolla | 07:26 | |
*** unicell1 has quit IRC | 07:28 | |
*** ducnv has quit IRC | 07:29 | |
*** hamzy_ has joined #openstack-kolla | 07:31 | |
*** hamzy has quit IRC | 07:31 | |
*** wirehead_ has quit IRC | 07:31 | |
*** Bhujay has quit IRC | 07:32 | |
*** vkmc has quit IRC | 07:32 | |
*** Bhujay has joined #openstack-kolla | 07:33 | |
*** caoyuan has quit IRC | 07:36 | |
*** caoyuan has joined #openstack-kolla | 07:36 | |
*** gkadam has joined #openstack-kolla | 07:37 | |
*** caoyuan has quit IRC | 07:45 | |
openstackgerrit | Lakshmi Prasanna Goutham Pratapa proposed openstack/kolla-ansible master: Apply Resource Constraints to Services. https://review.openstack.org/583268 | 07:45 |
*** caoyuan has joined #openstack-kolla | 07:45 | |
*** jchhatbar has joined #openstack-kolla | 07:46 | |
*** janki has quit IRC | 07:48 | |
*** hamzy_ has quit IRC | 07:50 | |
*** jchhatbar has quit IRC | 07:56 | |
*** jchhatbar has joined #openstack-kolla | 07:57 | |
*** rmart04 has joined #openstack-kolla | 07:58 | |
*** gfidente has joined #openstack-kolla | 08:00 | |
*** gfidente has quit IRC | 08:00 | |
*** gfidente has joined #openstack-kolla | 08:00 | |
*** wirehead_ has joined #openstack-kolla | 08:01 | |
*** jchhatbar has quit IRC | 08:08 | |
*** jchhatbar has joined #openstack-kolla | 08:08 | |
*** shardy has joined #openstack-kolla | 08:10 | |
openstackgerrit | Matthias Runge proposed openstack/kolla stable/queens: collectd: add collectd-write_kafka https://review.openstack.org/584268 | 08:11 |
*** mgoddard has joined #openstack-kolla | 08:12 | |
*** k_mouza has joined #openstack-kolla | 08:12 | |
*** dougsz has joined #openstack-kolla | 08:13 | |
*** jchhatbar has quit IRC | 08:13 | |
*** jchhatbar has joined #openstack-kolla | 08:14 | |
*** zhangfei has quit IRC | 08:14 | |
openstackgerrit | Matthias Runge proposed openstack/kolla stable/pike: collectd: add collectd-write_kafka https://review.openstack.org/584271 | 08:15 |
*** k_mouza has quit IRC | 08:16 | |
*** k_mouza_ has joined #openstack-kolla | 08:16 | |
*** athomas has joined #openstack-kolla | 08:17 | |
*** ktibi has joined #openstack-kolla | 08:18 | |
*** zhangfei has joined #openstack-kolla | 08:26 | |
*** jchhatbar has quit IRC | 08:30 | |
*** jchhatbar has joined #openstack-kolla | 08:30 | |
openstackgerrit | Merged openstack/kolla-ansible master: dev mode: Add support for trove https://review.openstack.org/535221 | 08:43 |
openstackgerrit | Merged openstack/kolla-ansible master: remove uid, gid from rsyncd.conf for swift user https://review.openstack.org/583417 | 08:43 |
*** hjensas has joined #openstack-kolla | 08:50 | |
*** hjensas has quit IRC | 08:50 | |
*** hjensas has joined #openstack-kolla | 08:50 | |
openstackgerrit | Min Sun proposed openstack/kolla-ansible master: Support checkout dedicated version from git with dev mode https://review.openstack.org/584193 | 08:51 |
*** pbourke has quit IRC | 08:53 | |
*** pbourke has joined #openstack-kolla | 08:53 | |
*** mattgo_ has joined #openstack-kolla | 08:53 | |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Allow externally managed swift as glance backend https://review.openstack.org/559201 | 08:55 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Less visible passwords please https://review.openstack.org/549858 | 08:56 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Add ability to have RabbitMQ do its own SSL https://review.openstack.org/571825 | 08:56 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Add support for Cells v2 rabbitmq servers https://review.openstack.org/569219 | 08:56 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Allow disabling keepalived for external LBs https://review.openstack.org/545757 | 08:57 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Add sensu role https://review.openstack.org/488367 | 08:57 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Add some volumes to make omsa work in sensu https://review.openstack.org/579695 | 08:57 |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Allow deploying ocata/pike nova on current k-a https://review.openstack.org/583833 | 08:58 |
*** k_mouza_ has quit IRC | 08:59 | |
*** Guest13260 has joined #openstack-kolla | 08:59 | |
openstackgerrit | Adam Harwell proposed openstack/kolla-ansible master: Add safe shutdown systemd jobs to tools https://review.openstack.org/584168 | 09:00 |
*** k_mouza has joined #openstack-kolla | 09:00 | |
*** Guest13260 is now known as w-miller | 09:00 | |
*** liuzz has quit IRC | 09:08 | |
w-miller | hjensas: i’m currently looking at supporting networking-baremetal in kolla-ansible. do you know what the best way to get the python-networking-baremetal package into the neutron-server image would be? | 09:09 |
openstackgerrit | Kevin Tibi proposed openstack/kolla-ansible master: Disable TLS 1.1 on haproxy https://review.openstack.org/583904 | 09:11 |
*** liuzz has joined #openstack-kolla | 09:11 | |
*** caoyuan has quit IRC | 09:12 | |
*** liuzz has quit IRC | 09:13 | |
*** caoyuan has joined #openstack-kolla | 09:13 | |
*** liuzz has joined #openstack-kolla | 09:13 | |
hjensas | w-miller: In tripleo we use kolla's package append feature when building the image. | 09:16 |
hjensas | w-miller: https://docs.openstack.org/kolla/latest/admin/image-building.html#package-customisation | 09:16 |
hjensas | w-miller: https://github.com/openstack/tripleo-common/blob/master/container-images/tripleo_kolla_template_overrides.j2#L63-L66 | 09:16 |
*** mgoddard has quit IRC | 09:17 | |
rm_work | Jeffrey4l: do you think there is any chance you would be able to review the haproxy refactor in time for the freeze? (assuming it is a "feature"? I'm not sure) https://review.openstack.org/#/c/576361/ | 09:17 |
rm_work | egonzalez: ^^ also | 09:18 |
w-miller | hjensas: great, thanks | 09:19 |
Jeffrey4l | rm_work, sure checking | 09:20 |
rm_work | Jeffrey4l: i mean, it's not exactly a minor review T_T | 09:21 |
Jeffrey4l | yeah, but i got your idea, nice patch :) | 09:21 |
rm_work | thanks! we would love to get this in ASAP | 09:21 |
rm_work | the number of other patches that edit the haproxy config is very high, so constant rebasing... T_T | 09:22 |
*** caoyuan has quit IRC | 09:22 | |
rm_work | I have tested it with at least the core set of services and it works as expected for me in our test environment (tempest tests all pass) | 09:23 |
Jeffrey4l | okay | 09:26 |
*** caoyuan_ has joined #openstack-kolla | 09:27 | |
*** jbadiapa has quit IRC | 09:28 | |
rm_work | if you do spot any potential issues, let me know, i will look into it. I know Mark had one concern with the way the configs were done, but i don't want to mix in yet ANOTHER refactor at the same time... and I believe he said it wasn't a blocker for him | 09:28 |
Jeffrey4l | rm_work, i am a little concern about the haproxy restart when adding new service. | 09:29 |
Jeffrey4l | hrm old implementation requires restart haproxy container too. ignore ^ | 09:30 |
rm_work | yeah, I *can* fix that though, as a follow-up | 09:31 |
Jeffrey4l | rm_work, any idea for this? | 09:31 |
*** hamzy_ has joined #openstack-kolla | 09:31 | |
rm_work | it's possible to cause haproxy to do a soft-reload, which is much lighter, if i reach into the container | 09:31 |
rm_work | yes, i happen to come from working on Octavia LBaaS :P so I am very familiar with haproxy, lol | 09:31 |
rm_work | though, there is one problem, which is the way the configs are loaded individually, it can only do that on a CHANGE, not on an addition of a new service | 09:32 |
rm_work | so there will be a little more complex logic around it | 09:32 |
Jeffrey4l | rm_work, one issue is we need restart the container to copy new added haproxy configuration from /var/lib/kolla to /etc/haproxy | 09:33 |
rm_work | yeah, so one way to avoid that is to *just not move them* | 09:33 |
rm_work | originally I was running just by reading all the configs from their mountpoint in /var/lib/kolla | 09:33 |
Jeffrey4l | but this will break COPY_ONCE feature ;) | 09:33 |
rm_work | i honestly am not 100% sure why we bother to copy them in, and don't just mount them directly... | 09:33 |
rm_work | yeah, I don't really understand the point of that though | 09:34 |
rm_work | but I am not a docker expert... | 09:34 |
Jeffrey4l | this is a historical issue, it is talked that remove COPY_ONCE | 09:34 |
Jeffrey4l | but it is blocked. | 09:34 |
Jeffrey4l | anyway, we could solve this later. | 09:35 |
rm_work | personally, i would rather just mount the files in the final destination :P | 09:35 |
Jeffrey4l | btw, have you tested multi node env? | 09:35 |
rm_work | yes, but not extensively | 09:35 |
rm_work | i know it works at least for the core services | 09:36 |
Jeffrey4l | rm_work, yeah, that's an option. and more directly | 09:36 |
Jeffrey4l | great | 09:36 |
rm_work | my test environment has separate HV / map / cap | 09:36 |
Jeffrey4l | what does the HV/map/cap mean? | 09:37 |
rm_work | ah i guess that is internal terminology | 09:37 |
*** w-miller_ has joined #openstack-kolla | 09:40 | |
rm_work | HV is hypervisor, CAP is Cell App Server, MAP is Meta App Servers | 09:40 |
rm_work | it's just the way we sort services | 09:41 |
*** w-miller has quit IRC | 09:42 | |
*** w-miller_ is now known as w-miller | 09:42 | |
*** mrostecki[m] has quit IRC | 09:42 | |
*** madhuri[m] has quit IRC | 09:42 | |
*** PascalBach[m] has quit IRC | 09:43 | |
*** retr0h[m] has quit IRC | 09:43 | |
*** jbadiapa has joined #openstack-kolla | 09:43 | |
Jeffrey4l | good to know, thanks. | 09:44 |
*** mrostecki[m] has joined #openstack-kolla | 09:49 | |
*** vkmc has joined #openstack-kolla | 09:49 | |
*** liuzz has quit IRC | 09:56 | |
*** hjensas has quit IRC | 10:00 | |
*** tone_zrt has quit IRC | 10:01 | |
*** madhuri[m] has joined #openstack-kolla | 10:02 | |
*** PascalBach[m] has joined #openstack-kolla | 10:02 | |
*** retr0h[m] has joined #openstack-kolla | 10:02 | |
Jeffrey4l | rm_work, have you ever thought to use import_role module? | 10:04 |
Jeffrey4l | https://docs.ansible.com/ansible/2.4/import_role_module.html | 10:04 |
Jeffrey4l | then we can use import_role: haproxy | 10:04 |
*** mgoddard has joined #openstack-kolla | 10:04 | |
rm_work | I think that is what Mark mentioned? | 10:05 |
Jeffrey4l | tasks_from: render_haproxy.cfg | 10:05 |
Jeffrey4l | rm_work, checked the comments. | 10:05 |
Jeffrey4l | yes | 10:05 |
rm_work | yeah, so he suggested that, but said it wasn't a blocker for him | 10:06 |
rm_work | and I agree, I love the idea, I always try to keep my code DRY, but I don't know that I want to introduce a new style of thing right when doing this | 10:06 |
rm_work | but, if you both really want, I can try to make that work | 10:06 |
*** jbadiapa is now known as jbadiapa|away | 10:06 | |
rm_work | i guess it would cut out a lot of lines... | 10:07 |
Jeffrey4l | yes. | 10:07 |
Jeffrey4l | and haproxy_common role can be removed tool | 10:07 |
Jeffrey4l | too* | 10:07 |
Jeffrey4l | the disadvantage is we have to pin ansible > 2.4 | 10:08 |
Jeffrey4l | but i think it is time to bump the min version for ansible | 10:08 |
rm_work | I'm not sure if we can remove the haproxy-common role just with that change... | 10:09 |
rm_work | there are some other reasons for that IIRC | 10:09 |
*** mdnadeem has quit IRC | 10:09 | |
Jeffrey4l | i think the main reason is handlers. | 10:10 |
rm_work | yeah, i think that was a big one :P | 10:10 |
Jeffrey4l | include_role / import_role could trigger the header :) | 10:11 |
rm_work | hmmm | 10:11 |
Jeffrey4l | rm_work, check the tasks_from parameter in include_role | 10:11 |
Jeffrey4l | mark still use a different role "haproxy_service" | 10:12 |
Jeffrey4l | but if you use tasks_from parameter, you cloud use the same haproxy. | 10:12 |
rm_work | hmm | 10:12 |
rm_work | yeah, that could work | 10:13 |
rm_work | so i have an important question -- do you think this refactor counts as a *feature* (and thus blocked by the upcoming feature freeze)? | 10:14 |
rm_work | or could it merge after | 10:14 |
*** mdnadeem has joined #openstack-kolla | 10:15 | |
Jeffrey4l | rm_work, could you make this happen in your current patch in the next fewer days? | 10:15 |
Jeffrey4l | even though i do not wanna to block the current implementation. | 10:15 |
Jeffrey4l | my only concern is the same with yours. | 10:16 |
Jeffrey4l | if you cloud not make it, we could apply a FFE. | 10:16 |
rm_work | I could TRY today, but I don't know if I can do much more over the weekend... | 10:16 |
Jeffrey4l | because it will be still like a big change ( lots of code will be change) in the refactor. | 10:16 |
rm_work | yes, that is true | 10:17 |
*** k_mouza has quit IRC | 10:17 | |
*** k_mouza has joined #openstack-kolla | 10:17 | |
Jeffrey4l | i could take part of the works too during the weekend;) | 10:18 |
mgoddard | rm_work: you are Adam Harwell? | 10:18 |
rm_work | ah! yes, hey Mark | 10:18 |
mgoddard | rm_work: hi :) | 10:19 |
Jeffrey4l | and the finally tag should be tagged at later of next week. so we still have fewer days | 10:19 |
rm_work | we're discussing making the include_role change you mentioned | 10:19 |
rm_work | Jeffrey also wanted to see that | 10:19 |
rm_work | I | 10:19 |
mgoddard | yeah, just saw my name go by. I'd be happy to see it as a follow up | 10:19 |
rm_work | I would also love to see this code a little more DRY T_T | 10:19 |
mgoddard | if you have something working, perhaps better not to rock the boat at this stage | 10:19 |
mgoddard | maybe it is something that could be discussed more generally at the PTG as a goal for Stein? | 10:20 |
rm_work | eh, I think we might be able to get just THAT refactor in quickly as a followup maybe | 10:20 |
rm_work | since it is much less major | 10:20 |
rm_work | but i'd like to get to it before i lose focus on this :P | 10:20 |
mgoddard | like you say, it's not the current norm in kolla, so might be good to establish a good pattern first | 10:21 |
mgoddard | and that could be done separately | 10:21 |
mgoddard | I'll add it to the PTG topic proposals | 10:21 |
rm_work | OK, I will be there | 10:21 |
rm_work | hopefully Clint also | 10:22 |
rm_work | I took over most of my team's patches since I prefer to work upstream anyway and it let them focus more on other stuff internally, but there is still a lot of interest in kolla-ansible contributions | 10:23 |
mgoddard | great. I'll be between kolla & ironic, but will try to be present for this session | 10:23 |
mgoddard | yeah I noticed you have adopted a few :) | 10:23 |
rm_work | 12. T_T | 10:23 |
mgoddard | it's not always easy bridging upstream & downstream | 10:23 |
rm_work | Any movement we can make on some of those would be great BTW :) but this one is my priority | 10:23 |
mgoddard | I was at CERN yesterday, trying to soak up their cells v2 experiences | 10:24 |
rm_work | ah yeah, we are also doing cellsv2 | 10:24 |
rm_work | and learning also from CERN :) | 10:24 |
mgoddard | was thinking of drafting a spec, perhaps you could help? | 10:24 |
mgoddard | you guys are from GoDaddy, right? | 10:24 |
rm_work | yeah | 10:24 |
mgoddard | how long have you been using cells v2 in production? | 10:24 |
rm_work | probably any of Clint/Kris/Josh would be better for working on a spec, but I can get their attention for it | 10:25 |
rm_work | only on our newer clouds, so not very long yet | 10:25 |
rm_work | we are coming from a very hacked up cells v1 | 10:25 |
rm_work | maybe a few months | 10:25 |
Jeffrey4l | for cell v2, i have a question, should neutron have a cell solution? is the mq in neutron a bottleneck? | 10:25 |
rm_work | I think the RMQ in ANYWHERE is a bottleneck ;P | 10:26 |
rm_work | but I don't know how well neutron supports cells yet... | 10:26 |
Jeffrey4l | compare nova with neturon. | 10:26 |
rm_work | though we do have a patch for helping with RMQ cells support... | 10:26 |
Jeffrey4l | cool | 10:26 |
rm_work | https://review.openstack.org/#/c/569219/ | 10:26 |
rm_work | that is one that could use reviews too :) | 10:27 |
rm_work | simpler ones that would be nice to get in also, if they look ok: | 10:28 |
rm_work | https://review.openstack.org/#/c/583833/ | 10:28 |
rm_work | https://review.openstack.org/#/c/542506/ | 10:28 |
rm_work | https://review.openstack.org/#/c/559201/ | 10:28 |
rm_work | I am carrying 13 patches internally right now, would love to reduce that number | 10:29 |
*** rmart04 has quit IRC | 10:32 | |
*** mdnadeem has quit IRC | 10:32 | |
mgoddard | rm_work: kolla has always been our most patched code, but we do eventually get things upstream | 10:33 |
rm_work | yeah, i switched us to running kolla-ansible/master, and upstream-first | 10:34 |
rm_work | so it isn't too bad | 10:34 |
mgoddard | runnning master, interesting | 10:34 |
mgoddard | upstream first is good when possible | 10:34 |
rm_work | i just `cat patch-list | xargs -n1 git review -x` :) | 10:34 |
rm_work | and that gets me the patches | 10:35 |
rm_work | yeah, we're working on getting to where we run everything on master | 10:36 |
rm_work | I have been running Octavia on master for a year now (though I'm core on that, so it's easy for me) | 10:36 |
mgoddard | that's certainly an honourable goal | 10:37 |
rm_work | yeah, soon we'll see if we're insane or not | 10:37 |
*** egonzalez has quit IRC | 10:37 | |
rm_work | I'm thinking maybe we're just all mad-men | 10:37 |
*** caoyuan_ has quit IRC | 10:40 | |
rm_work | Jeffrey4l: ahh, interesting... i thought it was common to cross-deploy | 10:45 |
rm_work | like, i thought that was the whole point of that variable :P | 10:45 |
rm_work | for the record, with that patch I am able to deploy Pike kolla, with master ;) | 10:46 |
Jeffrey4l | rm_work, no cross deploy is hard, and we still even have to jobs to test and ensure that | 10:46 |
hrw | morning | 10:46 |
hrw | docker-utils repo (https://github.com/hrw/docker-utils one) got script to remove images from docker.hub | 10:46 |
Jeffrey4l | rm_work, may be but still hard for kolla | 10:46 |
rm_work | can you think of an alternative that would allow this to work? | 10:46 |
rm_work | like, could we include both vars for a little longer? | 10:46 |
rm_work | or, am I going to be stuck with a downstream-only patch for a while :( | 10:47 |
Jeffrey4l | hrw, skopeo is also a good tool to do such thing :) | 10:47 |
hrw | Jeffrey4l: sure is. | 10:47 |
hrw | Jeffrey4l: just do not like to install yet another tool when script is all I need | 10:48 |
Jeffrey4l | rm_work, no. i see lots of person mix use kolla and kolla-ansible which cause issue, even a small different in z stream. | 10:48 |
rm_work | hmmm | 10:48 |
Jeffrey4l | rm_work for configuration issue, i think you'd better to use /etc/kolla/config feature to solve | 10:48 |
rm_work | so why does this variable exist? :P | 10:49 |
rm_work | `openstack_release` | 10:49 |
rm_work | it seems to have some actual effect on the workflow | 10:49 |
Jeffrey4l | rm_work, hrm, it is only use for image_tag number | 10:49 |
rm_work | right | 10:50 |
rm_work | so if you tell it to use Pike images | 10:50 |
rm_work | then it will install pike :) | 10:50 |
mgoddard | I have proposed discussion of cross-release support for the PTG. We are currently doing this a little bit, but it is not tested and will not work in all cases | 10:50 |
rm_work | but, in that case nova.conf will not work, lol | 10:50 |
Jeffrey4l | but when you have multi env, you are hardly to keep only one tag pike. | 10:50 |
mgoddard | there are some assumptions in kolla ansible about the contents of the kolla images | 10:51 |
rm_work | ok, so should we consider this one blocked until then? | 10:51 |
mgoddard | we need to do more for backwards compatibility | 10:51 |
Jeffrey4l | so in my company, we have 5.1 5.2 5.3 tag name | 10:51 |
rm_work | hmmm | 10:51 |
rm_work | ok | 10:51 |
rm_work | I was just reading the docs | 10:51 |
rm_work | https://docs.openstack.org/kolla-ansible/latest/user/quickstart.html#kolla-globals-yml | 10:52 |
rm_work | where it actually gave the example of `openstack_release: "pike"` :P | 10:52 |
Jeffrey4l | *To use DockerHub images*, the default image tag has to be overridden. Images are tagged with release names | 10:52 |
rm_work | that is how I thought it was used. but, if it can also be tag numbers like you say, then this approach won't work :( | 10:53 |
Jeffrey4l | for configuration change, i recommend you to check https://docs.openstack.org/kolla-ansible/latest/admin/advanced-configuration.html#openstack-service-configuration-in-kolla | 10:54 |
Jeffrey4l | which could solve your current issue :) | 10:54 |
rm_work | ok, I can -Workflow it and say it is pending until PTG when we can discuss cross-deploys | 10:54 |
Jeffrey4l | okay | 10:54 |
rm_work | interesting | 10:54 |
rm_work | yes, this may also work | 10:54 |
rm_work | i'll look at doing some of this | 10:54 |
Jeffrey4l | and you could also use jinja2 template in the /etc/kolla/config files | 10:55 |
rm_work | yeah | 10:55 |
rm_work | but then they stop getting updated as kolla updates | 10:55 |
rm_work | which is usually bad | 10:55 |
*** shardy has quit IRC | 10:56 | |
Jeffrey4l | this is custom configuration, | 10:56 |
Jeffrey4l | just treat it as globals.yml | 10:56 |
Jeffrey4l | it won't and shouln't change automatically when upgrade kolla. | 10:56 |
rm_work | well, i'll give this a shot | 10:56 |
openstackgerrit | Merged openstack/kolla-ansible master: Allow externally managed swift as glance backend https://review.openstack.org/559201 | 10:57 |
openstackgerrit | Merged openstack/kolla-ansible master: Allow disabling nova-ssh https://review.openstack.org/542506 | 10:58 |
rm_work | ah, thanks for those two! :) | 10:58 |
rm_work | yeah i think this config override is a better option for us if it works like i think it does | 11:01 |
rm_work | ok I need to go for a bit (it's 4AM here) | 11:01 |
rm_work | get some sleep before my standup ;) | 11:01 |
*** mdnadeem has joined #openstack-kolla | 11:03 | |
*** hjensas has joined #openstack-kolla | 11:04 | |
*** hjensas has quit IRC | 11:04 | |
*** hjensas has joined #openstack-kolla | 11:04 | |
*** fan_guiju has quit IRC | 11:06 | |
*** wolverineav has joined #openstack-kolla | 11:16 | |
*** wolverineav has quit IRC | 11:24 | |
*** wolverineav has joined #openstack-kolla | 11:25 | |
*** cah_link has quit IRC | 11:31 | |
*** lvdombrkr has joined #openstack-kolla | 11:39 | |
*** lvdombrkr89 has quit IRC | 11:40 | |
*** lvdombrkr89 has joined #openstack-kolla | 11:41 | |
*** mdnadeem has quit IRC | 11:42 | |
*** lvdombrkr has quit IRC | 11:44 | |
*** mdnadeem has joined #openstack-kolla | 11:45 | |
*** mdnadeem_ has joined #openstack-kolla | 11:47 | |
*** mdnadeem has quit IRC | 11:48 | |
*** xinliang has quit IRC | 11:55 | |
*** k_mouza_ has joined #openstack-kolla | 11:55 | |
*** xinliang has joined #openstack-kolla | 11:56 | |
*** k_mouza has quit IRC | 11:57 | |
*** k_mouza_ has quit IRC | 11:58 | |
*** jbadiapa|away is now known as jbadiapa | 11:58 | |
*** zhangfei has quit IRC | 12:00 | |
*** caoyuan has joined #openstack-kolla | 12:44 | |
*** edleafe is now known as figleaf | 12:46 | |
mandre | Jeffrey4l: hey, I'll likely not have the time to brush up the httpd base image before m3 | 13:07 |
mandre | I've finished the initial patch, but there are many things left to do: add the ability to modify the hierarchy of the images (i.e. customize the FROM and the extend_start chain) so we can easily base an custom image on httpd-base | 13:09 |
*** fan_guiju has joined #openstack-kolla | 13:09 | |
mandre | and add macro to remove the httpd packages from the images where there is no reason for it to be present | 13:10 |
mandre | Jeffrey4l: BTW, the total image size is 15.25GB when using httpd-base vs 16.03GB now | 13:12 |
*** fan_guiju has quit IRC | 13:14 | |
*** k_mouza has joined #openstack-kolla | 13:17 | |
*** k_mouza has quit IRC | 13:17 | |
*** rmart04 has joined #openstack-kolla | 13:24 | |
*** skramaja has quit IRC | 13:26 | |
drwahl | it seems as though our kolla-ansible deploy is setting up only a single glance-api container. is that normal? is there a reason it's not doing it on all control nodes? | 13:27 |
*** zhangfei has joined #openstack-kolla | 13:39 | |
*** TheJulia is now known as needssleep | 13:40 | |
*** EmilienM is now known as EvilienM | 13:41 | |
*** caoyuan_ has joined #openstack-kolla | 13:43 | |
openstackgerrit | Will Miller proposed openstack/kolla-ansible stable/queens: Add networking-baremetal configuration https://review.openstack.org/584376 | 13:44 |
*** caoyuan has quit IRC | 13:47 | |
Jeffrey4l | mandre, roger. | 13:47 |
*** jchhatbar has quit IRC | 13:49 | |
ktibi | Jeffrey4l, can you review https://review.openstack.org/#/c/583904/ and https://review.openstack.org/#/c/580338/ plz ;) | 13:52 |
openstackgerrit | Will Miller proposed openstack/kolla-ansible master: Add networking-baremetal configuration https://review.openstack.org/584379 | 13:52 |
*** jtriley has joined #openstack-kolla | 14:01 | |
*** mjturek has joined #openstack-kolla | 14:03 | |
*** peereb has quit IRC | 14:05 | |
*** jgu has joined #openstack-kolla | 14:09 | |
*** jgu_ has joined #openstack-kolla | 14:10 | |
*** mjturek has quit IRC | 14:10 | |
openstackgerrit | Will Miller proposed openstack/kolla master: Add networking-baremetal package to neutron-base https://review.openstack.org/584386 | 14:10 |
openstackgerrit | Merged openstack/kolla-ansible master: Fix monasca use of "action" -> "kolla_action" https://review.openstack.org/583893 | 14:11 |
*** mdnadeem_ has quit IRC | 14:12 | |
*** jgu has quit IRC | 14:13 | |
*** mjturek has joined #openstack-kolla | 14:14 | |
*** beagles is now known as beagles_afk | 14:33 | |
*** mjturek has quit IRC | 14:35 | |
*** Bhujay has quit IRC | 14:42 | |
*** ktibi has quit IRC | 14:45 | |
*** mjturek has joined #openstack-kolla | 14:48 | |
*** mjturek has quit IRC | 14:50 | |
*** lvdombrkr89 has quit IRC | 14:55 | |
*** zhangfei has quit IRC | 15:01 | |
*** rmart04 has quit IRC | 15:03 | |
*** mjturek has joined #openstack-kolla | 15:05 | |
*** mjturek has quit IRC | 15:07 | |
*** mjturek has joined #openstack-kolla | 15:10 | |
*** dave-mccowan has joined #openstack-kolla | 15:19 | |
*** dave-mcc_ has joined #openstack-kolla | 15:23 | |
*** wolverineav has quit IRC | 15:23 | |
*** goutham has quit IRC | 15:24 | |
*** dave-mccowan has quit IRC | 15:25 | |
*** trozet has quit IRC | 15:27 | |
*** wolverineav has joined #openstack-kolla | 15:28 | |
*** mattgo_ has quit IRC | 15:32 | |
*** mattgo has joined #openstack-kolla | 15:32 | |
*** mattgo has quit IRC | 15:32 | |
*** fan_guiju has joined #openstack-kolla | 15:34 | |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: [WIP] Move to the latest rabbitmq https://review.openstack.org/584426 | 15:35 |
*** dpascheka has joined #openstack-kolla | 15:35 | |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: [WIP] Update rabbitmq to use new conf & clustering https://review.openstack.org/584427 | 15:38 |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: [WIP] Move to the latest rabbitmq https://review.openstack.org/584426 | 15:39 |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla master: [WIP] Move to the latest rabbitmq https://review.openstack.org/584426 | 15:40 |
openstackgerrit | Doug Szumski proposed openstack/kolla-ansible master: Support setting rp_filter mode https://review.openstack.org/584429 | 15:44 |
*** caoyuan_ has quit IRC | 16:07 | |
*** klindgren has joined #openstack-kolla | 16:11 | |
openstackgerrit | Murali Annamneni proposed openstack/kolla-cli master: Fix bug in updating inventory file from an existing inventory https://review.openstack.org/584436 | 16:12 |
*** caoyuan has joined #openstack-kolla | 16:13 | |
*** caoyuan has quit IRC | 16:14 | |
*** klindgren has quit IRC | 16:16 | |
*** klindgren has joined #openstack-kolla | 16:19 | |
*** gkadam has quit IRC | 16:29 | |
*** fan_guiju has quit IRC | 16:31 | |
*** mgoddard has quit IRC | 16:34 | |
*** klindgren_ has joined #openstack-kolla | 16:46 | |
*** klindgren has quit IRC | 16:49 | |
*** w-miller has left #openstack-kolla | 16:55 | |
*** dougsz has quit IRC | 16:55 | |
*** conz has quit IRC | 17:15 | |
*** rmart04 has joined #openstack-kolla | 18:47 | |
*** rmart04 has quit IRC | 18:49 | |
*** athomas has quit IRC | 18:50 | |
*** gfidente has quit IRC | 18:52 | |
klindgren_ | Hello - trying to work on getting kolla/kolla nasible working with multiple nova cells | 19:36 |
klindgren_ | The issue that I am running into is that it appears that only nova-api container is built with an extended_start.sh file that handles the bootstrap cases | 19:38 |
klindgren_ | The issue is that in a multiple cell setup. The cells only need to run nova-conductor. | 19:38 |
klindgren_ | So as it currently sits its not really possible to bootstrap more than one cell in kolla/kolla-ansible. | 19:39 |
klindgren_ | I am waiting to copy and/or move the bootstrapping of nova from the nova-api container to the nova-conductor container. Since conductors run at every level of a nova-cell setup. | 19:40 |
klindgren_ | IE place: https://github.secureserver.net/openstack/kolla/blob/master/docker/nova/nova-api/extend_start.sh#L1-L21 into https://github.secureserver.net/openstack/kolla/tree/master/docker/nova/nova-conductor own extended_start.sh that also contains: https://github.secureserver.net/openstack/kolla/blob/master/docker/nova/nova-base/extend_start.sh#L3-L10 | 19:41 |
klindgren_ | But one thing I am not actually sure about is why can't the extended_start.sh in the base container have the bootstrap steps? | 19:42 |
*** jtriley has quit IRC | 19:45 | |
*** trozet has joined #openstack-kolla | 19:46 | |
klindgren_ | Actually never mind - I figured out a way to do it by adding a few more overrides in kolla-ansible. | 19:54 |
*** hjensas has quit IRC | 20:04 | |
*** jtriley has joined #openstack-kolla | 20:11 | |
*** jtriley has quit IRC | 20:18 | |
*** mjturek has quit IRC | 20:37 | |
*** munimeha1 has joined #openstack-kolla | 21:13 | |
*** PascalBach[m] has quit IRC | 21:16 | |
*** mrostecki[m] has quit IRC | 21:16 | |
*** retr0h[m] has quit IRC | 21:17 | |
*** madhuri[m] has quit IRC | 21:17 | |
*** jgu_ has quit IRC | 21:17 | |
*** munimeha1 has quit IRC | 21:24 | |
openstackgerrit | Clint 'SpamapS' Byrum proposed openstack/kolla-ansible master: Add safe shutdown systemd jobs to tools https://review.openstack.org/584168 | 21:39 |
*** figleaf is now known as edleafe | 21:41 | |
*** jgu has joined #openstack-kolla | 21:50 | |
*** mrostecki[m] has joined #openstack-kolla | 22:05 | |
*** PascalBach[m] has joined #openstack-kolla | 22:19 | |
*** retr0h[m] has joined #openstack-kolla | 22:19 | |
*** madhuri[m] has joined #openstack-kolla | 22:19 | |
*** itlinux has joined #openstack-kolla | 22:49 | |
*** itlinux has quit IRC | 23:08 | |
*** wolverineav has quit IRC | 23:58 | |
*** wolverineav has joined #openstack-kolla | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!