*** signed8bit is now known as signed8bit_Zzz | 00:04 | |
*** salv-orlando has quit IRC | 00:09 | |
*** salv-orlando has joined #openstack-kolla | 00:09 | |
*** keper7 has joined #openstack-kolla | 00:11 | |
*** salv-orlando has quit IRC | 00:14 | |
*** yangyapeng has quit IRC | 00:21 | |
*** vhosakot_ has joined #openstack-kolla | 00:23 | |
*** signed8bit_Zzz has quit IRC | 00:23 | |
*** vhosakot has quit IRC | 00:26 | |
*** athomas has quit IRC | 00:27 | |
*** jtriley has quit IRC | 00:29 | |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla-ansible master: Enable multinode scenarios in zuulv3 https://review.openstack.org/512779 | 00:32 |
---|---|---|
*** jbrooks has joined #openstack-kolla | 00:33 | |
*** dardelean has quit IRC | 00:33 | |
*** masber has quit IRC | 00:33 | |
Jeffrey4l | hrw, hi, no we do not have. but i'd like to add it. | 00:36 |
SamYaple | Jeffrey4l: i thought kolla wasnt going to add lumionous support? | 00:36 |
Jeffrey4l | SamYaple, yes. i am trying to add this. | 00:36 |
SamYaple | ah ok | 00:36 |
*** Pavo has quit IRC | 00:36 | |
*** vhosakot has joined #openstack-kolla | 00:37 | |
*** vhosakot_ has quit IRC | 00:38 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Bump to ceph luminous https://review.openstack.org/525216 | 00:38 |
*** masber has joined #openstack-kolla | 00:41 | |
*** tovin07__ has joined #openstack-kolla | 00:43 | |
*** kiennt26 has joined #openstack-kolla | 00:44 | |
*** ntpttr_laptop has quit IRC | 00:44 | |
*** cloud has joined #openstack-kolla | 00:44 | |
*** cloud is now known as Guest1563 | 00:45 | |
*** vhosakot has quit IRC | 00:46 | |
Guest1563 | I use k8s-kolla. "sudo kolla-ansible genconfig" will fail | 00:46 |
Guest1563 | I use the workaroud "sudo sed -i "s/.*ansible_play_batch.*length.*//" /usr/local/share/kolla-ansible/ansible/site.yml" | 00:47 |
Guest1563 | it can success but I don't know it hurt or not | 00:47 |
*** masber has quit IRC | 00:47 | |
Guest1563 | I follow the installation guide step by step. | 00:47 |
*** duonghq has joined #openstack-kolla | 00:48 | |
Guest1563 | and I hit second issue: "kollakube res create configmap" fails | 00:48 |
*** dardelean has joined #openstack-kolla | 00:49 | |
Guest1563 | but I don't know how to debug it | 00:49 |
*** affie has quit IRC | 00:49 | |
*** klindgren__ has quit IRC | 00:50 | |
*** jtriley has joined #openstack-kolla | 00:50 | |
*** dardelean has quit IRC | 00:53 | |
rwellum | Hi Guest1563 - the dev guide is a little out of date, and theres a patch set to fix it in review. The genconfig command should be: | 00:55 |
rwellum | https://www.irccloud.com/pastebin/tYY0EUOP/ | 00:55 |
*** tvignaud has quit IRC | 00:55 | |
rwellum | If gen config fails, then the second step will not succeed. | 00:56 |
Guest1563 | thanks. rwellum. current my workaoud can pass but "kollakube res create configmap" fails. I will use your way | 00:57 |
Guest1563 | so I need to use your way before "sudo kolla-ansible genconfig"? or just replace it | 00:58 |
rwellum | Replace it. | 00:58 |
rwellum | Guest1563: see: https://review.openstack.org/#/c/524125/2/doc/source/deployment-guide.rst | 00:59 |
rwellum | (In review) | 00:59 |
*** sai_ has quit IRC | 00:59 | |
*** klindgren has joined #openstack-kolla | 01:01 | |
*** zhurong has joined #openstack-kolla | 01:03 | |
*** keper7 has quit IRC | 01:04 | |
Guest1563 | great. thanks, rwellum:) | 01:05 |
rwellum | Sure np | 01:06 |
rwellum | if you have further issues, it's kind of late for me here - but if you come on IRC a bit earlier there are people to help. | 01:06 |
rwellum | We also have playbooks and a script to get you going to if needed. | 01:07 |
*** Pavo has joined #openstack-kolla | 01:07 | |
Jeffrey4l | duonghq, hi ,your idea works. https://review.openstack.org/522318 | 01:09 |
Guest1563 | Ubuntu does not currently work. Use centos | 01:09 |
Guest1563 | current, I use ubuntu. | 01:09 |
*** salv-orlando has joined #openstack-kolla | 01:10 | |
Guest1563 | thanks, rwellum | 01:12 |
*** tvignaud has joined #openstack-kolla | 01:13 | |
*** salv-orlando has quit IRC | 01:15 | |
*** yangyapeng has joined #openstack-kolla | 01:16 | |
*** caowei has joined #openstack-kolla | 01:18 | |
lxdong | what's this issue about http://paste.openstack.org/show/628127/? | 01:21 |
*** signed8bit has joined #openstack-kolla | 01:22 | |
duonghq | Jeffrey4l, nice | 01:23 |
duonghq | thank you | 01:23 |
*** keper7 has joined #openstack-kolla | 01:29 | |
*** Pavo has quit IRC | 01:35 | |
*** keper7 has quit IRC | 01:45 | |
*** jbrooks has quit IRC | 01:46 | |
*** mrunge has quit IRC | 01:56 | |
*** hawi has joined #openstack-kolla | 01:56 | |
*** mrunge has joined #openstack-kolla | 01:59 | |
*** kevzha01 has joined #openstack-kolla | 02:00 | |
*** kevzha01 is now known as kevinz | 02:00 | |
*** zhurong has quit IRC | 02:00 | |
*** salv-orlando has joined #openstack-kolla | 02:11 | |
*** salv-orlando has quit IRC | 02:16 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Bump to ceph luminous https://review.openstack.org/525216 | 02:18 |
*** signed8b_ has joined #openstack-kolla | 02:20 | |
*** pbourke_ has quit IRC | 02:23 | |
*** signed8bit has quit IRC | 02:24 | |
rwellum | Guest1563: I don't agree with that statement about ubuntu which is why I commented on it in the review. | 02:25 |
*** pbourke_ has joined #openstack-kolla | 02:25 | |
*** zhurong has joined #openstack-kolla | 02:30 | |
openstackgerrit | Dai Dang Van proposed openstack/kolla-ansible master: Add note about overwriting policy.json file https://review.openstack.org/511186 | 02:30 |
*** zhangfei has joined #openstack-kolla | 02:35 | |
*** signed8b_ is now known as signed8bit_Zzz | 02:38 | |
*** masber has joined #openstack-kolla | 02:39 | |
*** signed8bit_Zzz has quit IRC | 02:43 | |
*** jtriley has quit IRC | 02:48 | |
*** jtriley has joined #openstack-kolla | 02:59 | |
*** dmsimard has joined #openstack-kolla | 03:04 | |
spsurya | morning guys | 03:07 |
spsurya | sup duonghq | 03:07 |
*** klindgren has quit IRC | 03:09 | |
*** salv-orlando has joined #openstack-kolla | 03:12 | |
duonghq | hi spsurya | 03:13 |
*** lukepatrick has quit IRC | 03:14 | |
spsurya | duonghq: hope you doing well these days | 03:14 |
duonghq | spsurya, you too, do you have plan joining next PTG? | 03:16 |
*** salv-orlando has quit IRC | 03:17 | |
spsurya | same ques i was also about to ask :) | 03:17 |
spsurya | Hope so | 03:17 |
*** unicell has quit IRC | 03:18 | |
duonghq | nice, currently, I have not had any plan yet | 03:18 |
*** lukepatrick has joined #openstack-kolla | 03:18 | |
inc0 | ah damn | 03:25 |
inc0 | https://review.openstack.org/#/c/512779/ | 03:25 |
inc0 | I've seen this issue in my local env - ubuntu seems to dislike ceph - is that because it's already L in cloud archive? | 03:26 |
inc0 | hello everyone btw:) | 03:26 |
*** threestrands has joined #openstack-kolla | 03:27 | |
spsurya | hi....inc0 | 03:32 |
lxdong | Is it due to osd_crush_chooseleaf_type = 0 for osds down? | 03:38 |
*** jbrooks has joined #openstack-kolla | 03:39 | |
lxdong | single node ceph situation | 03:51 |
*** dave-mccowan has quit IRC | 04:00 | |
*** klindgren has joined #openstack-kolla | 04:03 | |
lxdong | all osds are down after creating in single node situation with luminous 12.2.2 | 04:04 |
*** klindgren_ has joined #openstack-kolla | 04:06 | |
*** lrensing has joined #openstack-kolla | 04:06 | |
*** coolsvap has joined #openstack-kolla | 04:09 | |
*** klindgren has quit IRC | 04:10 | |
*** janonymous has joined #openstack-kolla | 04:13 | |
*** salv-orlando has joined #openstack-kolla | 04:13 | |
*** salv-orlando has quit IRC | 04:18 | |
*** skramaja has joined #openstack-kolla | 04:18 | |
*** lukepatrick has quit IRC | 04:27 | |
*** zhurong has quit IRC | 04:27 | |
*** lukepatrick has joined #openstack-kolla | 04:35 | |
*** jtriley has quit IRC | 04:36 | |
*** zhurong has joined #openstack-kolla | 04:54 | |
*** jaosorior has joined #openstack-kolla | 04:55 | |
*** lrensing has quit IRC | 04:59 | |
*** gfidente has quit IRC | 05:07 | |
*** heigao_1212 has joined #openstack-kolla | 05:11 | |
*** salv-orlando has joined #openstack-kolla | 05:14 | |
*** heigao_1982 has quit IRC | 05:14 | |
*** heigao_1982 has joined #openstack-kolla | 05:15 | |
*** heigao_1212 has quit IRC | 05:15 | |
*** salv-orlando has quit IRC | 05:18 | |
*** jbrooks has quit IRC | 05:18 | |
*** jtriley has joined #openstack-kolla | 05:20 | |
*** k_mouza has joined #openstack-kolla | 05:22 | |
*** threestrands has quit IRC | 05:25 | |
*** dklyle has joined #openstack-kolla | 05:25 | |
*** k_mouza has quit IRC | 05:27 | |
*** david-lyle has quit IRC | 05:29 | |
*** dklyle has quit IRC | 05:30 | |
*** gfidente has joined #openstack-kolla | 05:34 | |
*** lukepatrick has quit IRC | 05:35 | |
SamYaple | inc0: yea its a known issue. kolla-pike and above ceph is borked | 05:52 |
inc0 | SamYaple: cuz of Luminous? | 05:52 |
SamYaple | inc0: you can potentially do external lumionous to makeit work | 05:52 |
SamYaple | yea | 05:52 |
*** unicell has joined #openstack-kolla | 05:54 | |
*** janki has joined #openstack-kolla | 05:58 | |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla master: Fix deployment of ceph https://review.openstack.org/525470 | 05:59 |
*** unicell1 has joined #openstack-kolla | 05:59 | |
*** unicell has quit IRC | 06:01 | |
*** gfidente has quit IRC | 06:03 | |
*** jtriley has quit IRC | 06:14 | |
*** salv-orlando has joined #openstack-kolla | 06:14 | |
*** salv-orlando has quit IRC | 06:19 | |
*** oanson has quit IRC | 06:20 | |
*** oanson has joined #openstack-kolla | 06:20 | |
openstackgerrit | Kevin Zhao proposed openstack/kolla-kubernetes master: Bump haproxy_exporter version from v0.7.1 to v0.8.0 https://review.openstack.org/525473 | 06:29 |
*** zhangfei has quit IRC | 06:38 | |
*** skramaja is now known as skramaja|afk | 06:38 | |
*** jtriley has joined #openstack-kolla | 06:39 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: ansiblize the init-runonce script https://review.openstack.org/523072 | 06:48 |
*** kiennt26 has quit IRC | 06:48 | |
*** dardelean has joined #openstack-kolla | 06:49 | |
*** zhangfei has joined #openstack-kolla | 06:50 | |
*** dardelean has quit IRC | 06:54 | |
*** egonzalez has joined #openstack-kolla | 07:00 | |
*** jtriley has quit IRC | 07:00 | |
*** cah_link has joined #openstack-kolla | 07:05 | |
*** kiennt26 has joined #openstack-kolla | 07:07 | |
*** salv-orlando has joined #openstack-kolla | 07:15 | |
*** zhurong has quit IRC | 07:16 | |
*** salv-orlando has quit IRC | 07:20 | |
*** salv-orlando has joined #openstack-kolla | 07:26 | |
*** jtriley has joined #openstack-kolla | 07:27 | |
*** jtriley has quit IRC | 07:32 | |
*** pcaruana has joined #openstack-kolla | 07:32 | |
*** jtriley has joined #openstack-kolla | 07:33 | |
openstackgerrit | Kevin Zhao proposed openstack/kolla master: Creates directory /var/run/mysqld on aarch64 when building MariaDB docker image https://review.openstack.org/521479 | 07:33 |
*** zhubingbing has joined #openstack-kolla | 07:34 | |
*** zhubingbing_ has quit IRC | 07:34 | |
*** genek has quit IRC | 07:35 | |
*** genek has joined #openstack-kolla | 07:36 | |
openstackgerrit | Jianghua Wang proposed openstack/kolla-ansible master: XenAPI: support XenAPI in nova configuration https://review.openstack.org/508196 | 07:36 |
openstackgerrit | Merged openstack/kolla-ansible master: fix shell error about tools/init-vpn https://review.openstack.org/517553 | 07:40 |
openstackgerrit | Eduardo Gonzalez proposed openstack/kolla-ansible master: Implement blazar ansible role https://review.openstack.org/454752 | 07:46 |
*** zhubingbing_ has joined #openstack-kolla | 07:46 | |
*** zhubingbing has quit IRC | 07:46 | |
openstackgerrit | Merged openstack/kolla-ansible stable/ocata: Use openstack_service_workers variable for gnocchi metricd workers https://review.openstack.org/525153 | 07:48 |
openstackgerrit | Merged openstack/kolla-ansible master: Fix deploy ceph failed when contoller hava no cluster_interface https://review.openstack.org/507520 | 07:48 |
*** jtriley has quit IRC | 07:49 | |
*** zhurong has joined #openstack-kolla | 07:49 | |
*** jtriley has joined #openstack-kolla | 07:53 | |
*** athomas has joined #openstack-kolla | 07:54 | |
*** codenamelxl has joined #openstack-kolla | 07:55 | |
*** salv-orlando has quit IRC | 07:57 | |
*** salv-orlando has joined #openstack-kolla | 07:57 | |
codenamelxl | hi everyone, i just set up a new openstack on Pike using kolla-ansible. "enable_neutron_dvr" set to yes. My problem is any floating IPs that stay on the network node is unreachable. | 07:58 |
codenamelxl | Floating IPS located on compute node can be reach just fine | 07:58 |
*** tvignaud has quit IRC | 07:58 | |
*** tvignaud has joined #openstack-kolla | 07:59 | |
*** skramaja|afk is now known as skramaja | 08:00 | |
*** jtriley has quit IRC | 08:00 | |
*** egonzalez has quit IRC | 08:01 | |
*** salv-orlando has quit IRC | 08:02 | |
*** jtriley has joined #openstack-kolla | 08:03 | |
*** heigao_1212 has joined #openstack-kolla | 08:04 | |
*** heigao_1982 has quit IRC | 08:04 | |
*** jascott1 has quit IRC | 08:07 | |
*** jascott1 has joined #openstack-kolla | 08:07 | |
*** egonzalez has joined #openstack-kolla | 08:09 | |
*** jtriley has quit IRC | 08:10 | |
*** jaosorior_ has joined #openstack-kolla | 08:10 | |
*** b_bezak has joined #openstack-kolla | 08:11 | |
*** jascott1 has quit IRC | 08:12 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Bump to ceph luminous https://review.openstack.org/525216 | 08:12 |
*** jaosorior has quit IRC | 08:13 | |
openstackgerrit | Chason Chan proposed openstack/kolla master: Add the missing murano package files https://review.openstack.org/522516 | 08:16 |
*** jtriley has joined #openstack-kolla | 08:16 | |
openstackgerrit | Merged openstack/kolla master: Change more repo url to infra mirror https://review.openstack.org/522318 | 08:17 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla-ansible master: Implement ceph-mgr service https://review.openstack.org/525495 | 08:17 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla stable/pike: Change more repo url to infra mirror https://review.openstack.org/525496 | 08:19 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla stable/ocata: Change more repo url to infra mirror https://review.openstack.org/525497 | 08:19 |
*** jtriley has quit IRC | 08:23 | |
*** shardy has joined #openstack-kolla | 08:24 | |
*** genek has quit IRC | 08:26 | |
*** unicell1 has quit IRC | 08:26 | |
*** mdnadeem has joined #openstack-kolla | 08:27 | |
*** jtriley has joined #openstack-kolla | 08:27 | |
codenamelxl | hi everyone, i just set up a new openstack on Pike using kolla-ansible. "enable_neutron_dvr" set to yes. My problem is any floating IPs that stay on the network node is unreachable. Floating IP located on compute node can be reach just fine | 08:29 |
lxdong | I ran into this issue:http://paste.openstack.org/show/628135/ can not create instance in single external ceph node cluster. | 08:30 |
*** heigao_1212 has quit IRC | 08:31 | |
*** jtriley has quit IRC | 08:33 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla master: Bump to ceph luminous https://review.openstack.org/525216 | 08:33 |
*** gkadam has joined #openstack-kolla | 08:38 | |
*** jtriley has joined #openstack-kolla | 08:38 | |
chason | egonzalez Hi, any work need us to do here? https://bugs.launchpad.net/kolla/+bug/1736206 | 08:39 |
openstack | Launchpad bug 1736206 in kolla "Deployer is unable to set MariaDB memory usage" [Wishlist,Triaged] | 08:40 |
*** pcaruana has quit IRC | 08:40 | |
egonzalez | chason, you can add a var into mariadb defaults | 08:41 |
chason | egonzalez OK, I will do it. | 08:42 |
chason | egonzalez Thanks! :D | 08:42 |
*** jtriley has quit IRC | 08:45 | |
*** jaosorior_ is now known as jaosorior | 08:47 | |
*** salv-orlando has joined #openstack-kolla | 08:47 | |
hrw | morning | 08:48 |
hrw | https://review.openstack.org/#/c/521479 - can someone look? | 08:48 |
*** jtriley has joined #openstack-kolla | 08:49 | |
*** k_mouza has joined #openstack-kolla | 08:53 | |
*** jtriley has quit IRC | 08:55 | |
*** skramaja is now known as skramaja|lunch | 08:55 | |
*** salv-orlando has quit IRC | 08:57 | |
*** salv-orlando has joined #openstack-kolla | 08:57 | |
*** magicboiz has quit IRC | 09:01 | |
*** salv-orlando has quit IRC | 09:02 | |
*** jtriley has joined #openstack-kolla | 09:03 | |
*** jtriley has quit IRC | 09:10 | |
*** zhangfei has quit IRC | 09:12 | |
*** magicboiz has joined #openstack-kolla | 09:12 | |
*** dardelean has joined #openstack-kolla | 09:15 | |
*** jtriley has joined #openstack-kolla | 09:16 | |
*** magicboiz has quit IRC | 09:17 | |
*** magicboiz has joined #openstack-kolla | 09:17 | |
*** ktibi has joined #openstack-kolla | 09:20 | |
*** jtriley has quit IRC | 09:22 | |
*** Zophar has joined #openstack-kolla | 09:28 | |
*** jtriley has joined #openstack-kolla | 09:29 | |
*** k_mouza has quit IRC | 09:29 | |
*** k_mouza_ has joined #openstack-kolla | 09:29 | |
*** heigao has joined #openstack-kolla | 09:29 | |
*** affie has joined #openstack-kolla | 09:30 | |
*** zhangfei has joined #openstack-kolla | 09:30 | |
*** k_mouza_ has quit IRC | 09:30 | |
*** k_mouza has joined #openstack-kolla | 09:31 | |
*** k_mouza has quit IRC | 09:32 | |
*** affie_ has joined #openstack-kolla | 09:32 | |
*** k_mouza has joined #openstack-kolla | 09:32 | |
ktibi | morning all ;) | 09:33 |
dardelean | hi | 09:33 |
*** skramaja|lunch is now known as skramaja | 09:34 | |
*** jtriley has quit IRC | 09:35 | |
spsurya | hrw: done | 09:35 |
hrw | thanks | 09:35 |
*** affie has quit IRC | 09:35 | |
*** gfidente has joined #openstack-kolla | 09:37 | |
*** gfidente has quit IRC | 09:37 | |
*** gfidente has joined #openstack-kolla | 09:37 | |
*** jtriley has joined #openstack-kolla | 09:37 | |
chason | o/ | 09:39 |
chason | https://review.openstack.org/#/c/512576/ | 09:39 |
chason | can someone review this patch? :) | 09:39 |
ktibi | I have issue with telegraf and kolla. Who use telegraf here ? | 09:43 |
ktibi | the bug : https://github.com/influxdata/telegraf/issues/1352 | 09:43 |
*** jtriley has quit IRC | 09:44 | |
ktibi | bug can fix on centos with service proc-sys-fs-binfmt_misc.automount restart | 09:46 |
openstackgerrit | Merged openstack/kolla master: Creates directory /var/run/mysqld on aarch64 when building MariaDB docker image https://review.openstack.org/521479 | 09:51 |
*** owalsh_ is now known as owalsh | 09:52 | |
*** heigao has quit IRC | 09:52 | |
*** jtriley has joined #openstack-kolla | 09:52 | |
*** bjolo has quit IRC | 09:54 | |
openstackgerrit | Martin André proposed openstack/kolla master: Validate Dockerfile indentation in pep8 https://review.openstack.org/525516 | 09:55 |
mandre | egonzalez: ^^ you'll love it I'm sure | 09:59 |
*** gfidente has quit IRC | 10:00 | |
*** jtriley has quit IRC | 10:07 | |
*** gfidente has joined #openstack-kolla | 10:09 | |
*** gfidente has quit IRC | 10:09 | |
*** gfidente has joined #openstack-kolla | 10:09 | |
codenamelxl | I feel like centralized floating IP on DVR mode deployed with Kolla is broken. | 10:10 |
codenamelxl | Any floating IPs that stay on the network node is unreachable. Floating IP located on compute node can be reach just fine. | 10:11 |
codenamelxl | Or am i doing something wrong? | 10:11 |
*** sambetts|afk is now known as sambetts | 10:14 | |
*** jtriley has joined #openstack-kolla | 10:17 | |
*** kevinz has quit IRC | 10:17 | |
egonzalez | codenamelxl, in dvr, floating IPs should not be at network hosts, only in compute | 10:20 |
*** kbaegis has joined #openstack-kolla | 10:20 | |
*** genek has joined #openstack-kolla | 10:21 | |
kolla-slack | <aslancimbom100> Morning/evening | 10:21 |
codenamelxl | It will be on network node if i create a centralized router or distributed router with this use case https://bugs.launchpad.net/neutron/+bug/1583694 | 10:21 |
openstack | Launchpad bug 1583694 in neutron "[RFE] DVR support for Allowed_address_pair port that are bound to multiple ACTIVE VM ports" [Wishlist,Fix released] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) | 10:21 |
*** kiennt26 has quit IRC | 10:22 | |
kolla-slack | <aslancimbom100> Anyone know if I have to disable local chronyd to use enable_chrony? | 10:22 |
*** jtriley has quit IRC | 10:23 | |
codenamelxl | @egonzalez So basically, if i use DVR, Octavia won't work since the floating IP of Octavia will be on the network node. But it's not only Octavia, any floating IPs located on network node is unreachable | 10:26 |
*** genek has quit IRC | 10:26 | |
*** k_mouza_ has joined #openstack-kolla | 10:29 | |
*** jtriley has joined #openstack-kolla | 10:30 | |
*** tovin07__ has quit IRC | 10:30 | |
zhangfei | Hi, need help review https://review.openstack.org/#/c/513353/, thanks | 10:30 |
*** k_mouza has quit IRC | 10:33 | |
*** k_mouza_ has quit IRC | 10:34 | |
*** k_mouza has joined #openstack-kolla | 10:35 | |
*** jtriley has quit IRC | 10:37 | |
*** jtriley has joined #openstack-kolla | 10:38 | |
*** jtriley has quit IRC | 10:43 | |
*** daidv has quit IRC | 10:50 | |
*** jtriley has joined #openstack-kolla | 10:51 | |
*** codenamelxl has quit IRC | 10:53 | |
*** jtriley has quit IRC | 10:58 | |
*** kbaegis has quit IRC | 10:58 | |
*** serlex has joined #openstack-kolla | 11:00 | |
*** kbaegis has joined #openstack-kolla | 11:03 | |
*** daidv has joined #openstack-kolla | 11:05 | |
*** jtriley has joined #openstack-kolla | 11:05 | |
*** zhangfei has quit IRC | 11:06 | |
*** jascott1 has joined #openstack-kolla | 11:10 | |
*** jtriley has quit IRC | 11:11 | |
*** jascott1 has quit IRC | 11:14 | |
*** jtriley has joined #openstack-kolla | 11:15 | |
*** mgoddard has joined #openstack-kolla | 11:18 | |
*** caowei has quit IRC | 11:21 | |
*** jtriley has quit IRC | 11:22 | |
*** jtriley has joined #openstack-kolla | 11:29 | |
*** k_mouza_ has joined #openstack-kolla | 11:31 | |
*** mwynne has quit IRC | 11:33 | |
*** k_mouza has quit IRC | 11:34 | |
*** jtriley has quit IRC | 11:36 | |
*** janki has quit IRC | 11:39 | |
*** mwynne has joined #openstack-kolla | 11:41 | |
*** jtriley has joined #openstack-kolla | 11:43 | |
*** jtriley has quit IRC | 11:50 | |
*** mwynne has quit IRC | 11:54 | |
*** mwynne has joined #openstack-kolla | 11:55 | |
*** jtriley has joined #openstack-kolla | 11:59 | |
*** kevinz has joined #openstack-kolla | 12:02 | |
*** kbaegis has quit IRC | 12:03 | |
*** kbaegis has joined #openstack-kolla | 12:06 | |
*** jtriley has quit IRC | 12:06 | |
*** kbaegis has quit IRC | 12:08 | |
*** serlex has quit IRC | 12:14 | |
*** jtriley has joined #openstack-kolla | 12:15 | |
*** k_mouza_ has quit IRC | 12:17 | |
*** k_mouza has joined #openstack-kolla | 12:21 | |
*** jtriley has quit IRC | 12:36 | |
*** skramaja has quit IRC | 12:36 | |
*** zhurong has quit IRC | 12:36 | |
*** jtriley has joined #openstack-kolla | 12:41 | |
*** zhurong has joined #openstack-kolla | 12:45 | |
*** jtriley has quit IRC | 12:47 | |
*** david-lyle has joined #openstack-kolla | 12:48 | |
*** jtriley has joined #openstack-kolla | 12:51 | |
*** gfidente has quit IRC | 12:54 | |
*** gfidente has joined #openstack-kolla | 12:56 | |
*** gfidente has quit IRC | 12:56 | |
*** gfidente has joined #openstack-kolla | 12:56 | |
openstackgerrit | Merged openstack/kolla-ansible master: Removed -e Docker flag in ovsdpdk task https://review.openstack.org/523498 | 12:57 |
*** jtriley has quit IRC | 12:57 | |
*** david-lyle has quit IRC | 13:01 | |
*** zhurong has quit IRC | 13:02 | |
*** keper7 has joined #openstack-kolla | 13:02 | |
*** zhurong has joined #openstack-kolla | 13:02 | |
openstackgerrit | Merged openstack/kolla-ansible master: Add content about cluster_network interface https://review.openstack.org/514544 | 13:05 |
*** jtriley has joined #openstack-kolla | 13:06 | |
*** rhallisey has joined #openstack-kolla | 13:10 | |
*** jtriley has quit IRC | 13:12 | |
Jeffrey4l | inc0, egonzalez could you review this https://review.openstack.org/525497 | 13:17 |
Jeffrey4l | spsurya, ^^ | 13:17 |
*** jtriley has joined #openstack-kolla | 13:22 | |
*** signed8bit has joined #openstack-kolla | 13:24 | |
*** genek has joined #openstack-kolla | 13:25 | |
*** zhurong has quit IRC | 13:27 | |
*** jtriley has quit IRC | 13:28 | |
*** jtriley has joined #openstack-kolla | 13:30 | |
*** yangyapeng has quit IRC | 13:33 | |
*** yangyapeng has joined #openstack-kolla | 13:34 | |
*** kevzha01_ has joined #openstack-kolla | 13:34 | |
*** jtriley has quit IRC | 13:36 | |
*** kevinz has quit IRC | 13:37 | |
*** yangyapeng has quit IRC | 13:38 | |
openstackgerrit | Pierre Hanselmann proposed openstack/kolla-ansible master: Fix dhcp_agent configuration for vmware DVS https://review.openstack.org/525606 | 13:40 |
*** janki has joined #openstack-kolla | 13:40 | |
openstackgerrit | Chason Chan proposed openstack/kolla-ansible master: Add vitrage ansible role https://review.openstack.org/432608 | 13:42 |
*** jtriley has joined #openstack-kolla | 13:44 | |
ktibi | Jeffrey4l Do you think we can expose ceph osd socket for monitoring the cluster with by example calamari ?? | 13:48 |
Jeffrey4l | ktibi, hrm, how? and how about use ceph-mgr, which i am trying to implement. | 13:49 |
*** jtriley has quit IRC | 13:51 | |
dmsimard | Jeffrey4l: I haven't kept up with https://review.openstack.org/#/c/521450/ very much | 13:53 |
dmsimard | Jeffrey4l: does it yield any improvements in performance after all ? | 13:53 |
Jeffrey4l | dmsimard, yes. at least some package is cached by squid | 13:54 |
*** signed8bit has quit IRC | 13:54 | |
Jeffrey4l | http://logs.openstack.org/50/521450/40/check/kolla-build-centos-source/7ea6506/logs/system_log/squid/access.txt.gz | 13:55 |
Jeffrey4l | dmsimard, here is the squid access logs. | 13:55 |
*** signed8bit has joined #openstack-kolla | 13:55 | |
*** jtriley has joined #openstack-kolla | 13:55 | |
dmsimard | Jeffrey4l: I was hoping for something with the squid implementation, maybe you can tell me if it'd work or not -- it's what I had discussed with inc0 | 13:56 |
Jeffrey4l | there are lots of package download multi time in kolla. with squid, the downloaded package is get very fast. | 13:56 |
Jeffrey4l | it works | 13:56 |
dmsimard | Jeffrey4l: some of the kolla images have 30+ layers | 13:56 |
Jeffrey4l | yes. this is another story ;( | 13:57 |
Jeffrey4l | or another issue. | 13:57 |
dmsimard | When doing a "docker push", the docker client does a "GET" for every layer of an image and the registry returns a 200 OK if found (client will not push) or 404 Not found (client will push) | 13:57 |
dmsimard | Those lookups are very expensive when doing a docker push of 125+ images | 13:57 |
Jeffrey4l | is this will be a bottle neck? | 13:57 |
dmsimard | Yeah, RDO builds those 125+ images and pushes them to a registry sitting in the same network and it's definitely a large part of the push bottleneck | 13:58 |
dmsimard | Tailing the logs, you can just see the Docker client hammering the registry with image layer lookups | 13:58 |
*** dave-mcc_ has joined #openstack-kolla | 13:58 | |
Jeffrey4l | we can implement --squash in "docker build" | 13:58 |
dmsimard | Not a good option, at least for us, it would heavily increase the storage usage | 13:58 |
dmsimard | because it kills the savings we get from sharing layers | 13:58 |
Jeffrey4l | dmsimard, no. just use one layer for single image. | 13:58 |
dmsimard | so I was hoping we'd be able to cache the 200 OK's from the registry | 13:58 |
Jeffrey4l | it won't increase the size. | 13:59 |
dmsimard | It would increase the size *on disk* for the registry | 13:59 |
dmsimard | because you have 125 large images with one layer each, instead of 125 images which share a bunch of layers | 13:59 |
Jeffrey4l | hrm. i am not sure whether it will squash parent image. if so, this won't be a good idea. | 14:00 |
Jeffrey4l | so do you have any idea for this? dmsimard | 14:00 |
* Jeffrey4l is checking the squash parameter. | 14:01 | |
*** jtriley has quit IRC | 14:01 | |
ktibi | Jeffrey4l We just need to add a volume on /var/run/ceph/ | 14:02 |
*** mchlumsky has joined #openstack-kolla | 14:02 | |
ktibi | like haproxy : https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/haproxy/defaults/main.yml#L14 | 14:02 |
*** mmethot|away is now known as mmethot | 14:02 | |
Jeffrey4l | ktibi, i think this is OK. | 14:02 |
*** jtriley has joined #openstack-kolla | 14:03 | |
*** mchlumsky has quit IRC | 14:04 | |
*** mchlumsky has joined #openstack-kolla | 14:06 | |
openstackgerrit | Merged openstack/kolla master: Remove duplicated statement of parameter choices of config options https://review.openstack.org/514608 | 14:06 |
*** jtriley has quit IRC | 14:09 | |
openstackgerrit | Merged openstack/kolla master: Update swift for replication issues https://review.openstack.org/522838 | 14:09 |
*** jascott1 has joined #openstack-kolla | 14:12 | |
DosBarbos | ******************** | 14:13 |
DosBarbos | Hello all, | 14:14 |
DosBarbos | how can I fix this problem? anyone knows? https://bugs.launchpad.net/kolla-kubernetes/+bug/1666022 | 14:14 |
openstack | Launchpad bug 1666022 in kolla-kubernetes "kolla-kubernetes run command abnormal" [Critical,Invalid] | 14:14 |
DosBarbos | output of command kolla-kubernetes and kollakube is the same :( | 14:16 |
*** jascott1 has quit IRC | 14:16 | |
DosBarbos | i try to use the bugfix from this link https://review.openstack.org/#/c/447356/8/doc/source/deployment-guide.rst@41 ( git fetch git://git.openstack.org/openstack/kolla-kubernetes refs/changes/40/439740/17 && git cherry-pick FETCH_HEAD ) but it didnt help | 14:17 |
*** dardelean has quit IRC | 14:18 | |
DosBarbos | anyone? where are kolla-kube masters? :) | 14:19 |
DosBarbos | please ... | 14:19 |
dmsimard | Jeffrey4l: have you tested the squid implementation when doing pushes ? | 14:21 |
dmsimard | like for example we shouldn't cache 404's but we should cache 200's | 14:22 |
Jeffrey4l | dmsimard, push image won't use squid. | 14:22 |
dmsimard | Jeffrey4l: why not ? | 14:22 |
Jeffrey4l | but i can add this to check what will happen | 14:22 |
Jeffrey4l | squid is transport proxy. and should export http_proxy explicity. | 14:22 |
dmsimard | Jeffrey4l: ah so it's set a transparent right now, not http ? | 14:22 |
Jeffrey4l | it is a http proxy. and a transparent http proxy. | 14:23 |
Jeffrey4l | it is a http proxy. not a transparent http proxy. | 14:24 |
*** signed8bit is now known as signed8bit_Zzz | 14:30 | |
ktibi | Jeffrey4l on fresh install, I have cinder-volume error with ceph ==> Error connecting to ceph cluster.: TimedOut: error connecting to the cluster | 14:31 |
DosBarbos | вщ нщг рфму фтн швуф фищге ьш вуздщньуте зкщидуь, | 14:33 |
DosBarbos | sorry | 14:33 |
DosBarbos | do you have any idea about my problem | 14:33 |
DosBarbos | ? | 14:33 |
*** signed8bit_Zzz is now known as signed8bit | 14:36 | |
openstackgerrit | Paul Bourke (pbourke) proposed openstack/kolla-ansible master: Add config necessary to use Barbican with Murano https://review.openstack.org/525631 | 14:36 |
DosBarbos | jeffrey41 dmsimard ? | 14:36 |
*** kevzha01_ has quit IRC | 14:40 | |
*** magicboiz has quit IRC | 14:40 | |
*** jtriley has joined #openstack-kolla | 14:41 | |
rwellum | DosBarbos: hi | 14:42 |
rwellum | I have DosBarbos - I've never seen this before - not sure what it is, where it came from? | 14:43 |
*** lrensing has joined #openstack-kolla | 14:44 | |
*** jaosorior has quit IRC | 14:46 | |
rwellum | DosBarbos: if you're having issues, please try: https://github.com/RichWellum/k8s - the tool ko.py will launch an AIO kolla-kubernetes cluster. You can then see if your environment is good or not. | 14:46 |
DosBarbos | Hello rwellum, it is after installations stap (sudo pip install -U kolla-ansible/ kolla-kubernetes/) regarding the official documentations https://docs.openstack.org/kolla-kubernetes/latest/deployment-guide.html | 14:46 |
DosBarbos | thx a lot - will chk | 14:47 |
DosBarbos | seems it is your own deployment way | 14:48 |
rwellum | DosBarbos: yes you don't run kolla-kubernetes directly as per your bug written. Not that I'm aware of anyway. | 14:48 |
keper7 | hey I'm trying to get both ceph and kolla-k8s on the same hardware. The docs says the kolla-k8s ceph code isn't ready. Is the current best practice to make a kolla-ansible inventory with just the storage nodes populated and install ceph that way? | 14:48 |
rwellum | DosBarbos: it's basically the deployment guide written in python | 14:49 |
rwellum | DosBarbos: if you run in --demo mode it will explain each step. The deployment guide is a little out of date and theres a PS currently in review. | 14:50 |
DosBarbos | as i understand I can deploy it on openstack VM. | 14:50 |
rwellum | keper7: some users have deployed ceph in an AIO environment with kolla-k8s. I haven't yet. | 14:50 |
rwellum | DosBarbos: yes I have done that many times. | 14:50 |
rwellum | DosBarbos: networking can be a little interesting, but you can certainly get a viable kolla-k8s cluster up and running. | 14:51 |
DosBarbos | thx a lot again! i will try to use your easiest way to deploy | 14:51 |
*** goldyfruit has joined #openstack-kolla | 14:52 | |
DosBarbos | ok-i need POC to check "what is that". After that i will try to expand cluster | 14:53 |
keper7 | rwellum: I'm trying to do a 3-node deployment currently, for HA. How do you currently deploy ceph? | 14:53 |
rwellum | DosBarbos: if you want pike instead of ocata then you need to do: 'ko.py eth1 eth2 -iv pike -dr rwellum' - as the pike images are not pushed officially to dockerhub. | 14:53 |
DosBarbos | what is your opinion: This solutions is ready to prod? I mean kolla-kubernates | 14:53 |
rwellum | keper7: I haven't used ceph yet (just use a basic LVM backend). But there is a playbook in review (https://review.openstack.org/#/c/520847/) which has ceph to multiple hosts you can look at. | 14:55 |
*** lpetrut has joined #openstack-kolla | 14:55 | |
rwellum | keper7: also the author can help when he's online (jascott1) | 14:55 |
rwellum | DosBarbos: no it is not production ready; release one is planned for Rocky. | 14:55 |
*** manheim has joined #openstack-kolla | 14:57 | |
*** manheim is now known as dardelean | 14:57 | |
dardelean | is it mandatory to use fqdn when using TLS? | 14:58 |
keper7 | rwellum: Regarding production ready-ness, is kolla-ansible production-ready? | 14:58 |
rwellum | keper7: yes production ready, very stable, at least 9% of deployments in the World use it AFAIK. | 15:00 |
*** lrensing has quit IRC | 15:00 | |
*** lrensing has joined #openstack-kolla | 15:03 | |
*** zhubingbing has joined #openstack-kolla | 15:05 | |
zhubingbing | hi guys | 15:06 |
*** shardy is now known as shardy_afk | 15:08 | |
DosBarbos | hi | 15:09 |
*** lrensing_ has joined #openstack-kolla | 15:11 | |
*** janki has quit IRC | 15:13 | |
*** david-lyle has joined #openstack-kolla | 15:13 | |
*** janki has joined #openstack-kolla | 15:14 | |
*** lrensing has quit IRC | 15:14 | |
*** lrensing_ is now known as lrensing | 15:14 | |
*** janki has quit IRC | 15:14 | |
*** janki has joined #openstack-kolla | 15:15 | |
*** cah_link1 has joined #openstack-kolla | 15:22 | |
*** cah_link has quit IRC | 15:24 | |
*** cah_link1 is now known as cah_link | 15:24 | |
*** vhosakot has joined #openstack-kolla | 15:28 | |
*** cduchesne has quit IRC | 15:31 | |
*** cah_link has quit IRC | 15:31 | |
*** cah_link has joined #openstack-kolla | 15:32 | |
*** DosBarbos has quit IRC | 15:32 | |
*** hongbin has joined #openstack-kolla | 15:33 | |
openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla stable/pike: Creates directory /var/run/mysqld on aarch64 when building MariaDB docker image https://review.openstack.org/525663 | 15:36 |
*** cduchesne has joined #openstack-kolla | 15:36 | |
*** david-lyle has quit IRC | 15:39 | |
hrw | this may be one of last stable/pike backports | 15:39 |
*** shardy_afk is now known as shardy | 15:42 | |
*** cah_link1 has joined #openstack-kolla | 15:44 | |
*** cah_link has quit IRC | 15:45 | |
*** cah_link1 is now known as cah_link | 15:45 | |
*** goldyfruit has quit IRC | 15:47 | |
*** cah_link has quit IRC | 15:48 | |
*** openstackgerrit has quit IRC | 15:48 | |
*** cah_link has joined #openstack-kolla | 15:48 | |
*** goldyfruit has joined #openstack-kolla | 15:49 | |
*** dardelean has quit IRC | 15:50 | |
*** dardelean has joined #openstack-kolla | 15:51 | |
*** dardelean has quit IRC | 15:56 | |
*** cah_link has quit IRC | 15:57 | |
*** lrensing_ has joined #openstack-kolla | 15:58 | |
*** lrensing_ has quit IRC | 15:59 | |
*** lrensing has quit IRC | 16:00 | |
*** egonzalez has quit IRC | 16:00 | |
*** codenamelxl has joined #openstack-kolla | 16:02 | |
*** zhubingbing has quit IRC | 16:05 | |
*** coolsvap has quit IRC | 16:08 | |
*** diablo_rojo has joined #openstack-kolla | 16:11 | |
*** athomas has quit IRC | 16:12 | |
*** k_mouza_ has joined #openstack-kolla | 16:14 | |
*** jascott1 has joined #openstack-kolla | 16:14 | |
*** k_mouza_ has quit IRC | 16:15 | |
*** k_mouza has quit IRC | 16:16 | |
*** k_mouza has joined #openstack-kolla | 16:16 | |
*** jascott1 has quit IRC | 16:18 | |
diablo_rojo | inc0, is it cool to go ahead with the kolla-k8 like today? Any certain time you would prefer? | 16:20 |
kolla-slack | <inc0> diablo_rojo I'm at opnfv plugfest today, tomorrow tho.. | 16:21 |
rwellum | What's the ask diablo_rojo ? Regarding storybook or something else? | 16:22 |
Zara | hi! I promised folk some kolla questions a while ago! Here's a question! | 16:29 |
Zara | so aiui kolla needs pretty specific config in dockerfiles and ansible roles to deploy things. | 16:29 |
Zara | And also aiui it's not really possible to use kolla to build and deploy just one container in isolation | 16:29 |
Zara | for development purposes, various openstack services have to come along for the ride. | 16:29 |
Zara | I'm wondering if anyone knows of any tools to reuse dockerfiles used with kolla for local development. | 16:29 |
Zara | Since I think it's possible to use the same ansible roles (just have the variables expand to different | 16:29 |
Zara | values) but I'm not sure what's out there for parsing the jinja-templated dockerfiles that kolla | 16:29 |
Zara | expects. | 16:29 |
Zara | I'm not sure if this makes sense, I'm still working out the problem atm and I'm pretty new to | 16:29 |
Zara | everything. :) | 16:29 |
*** dardelean has joined #openstack-kolla | 16:30 | |
*** dardelean has quit IRC | 16:34 | |
*** lukepatrick has joined #openstack-kolla | 16:39 | |
diablo_rojo | rwellum, yeah just asking if its cool to do the migration today, but sounds like we are waiting till tomorrow. | 16:41 |
kfox1111 | morning. | 16:47 |
*** jascott1 has joined #openstack-kolla | 16:51 | |
*** jbrooks has joined #openstack-kolla | 16:57 | |
rwellum | diablo_rojo: maybe I mis-estimated the impact, I saw it as more as an enablement so I can create a storyboard - than a full migration. | 16:58 |
*** rmart04 has joined #openstack-kolla | 16:59 | |
*** rmart04 has quit IRC | 17:00 | |
rwellum | Zara: You can build a specific image in kolla np. Deploying it to an existing cluster is harder. But there's a new dev mode in kolla-ansible that might work for you. | 17:00 |
*** janki has quit IRC | 17:02 | |
Zara | hm, okay. if it helps, the workflow that I want is 'local dev has some dockerfiles and an ansible role to build an image and deploy a container based on it. those can then be slotted into the rest of the infra for building and deploying all desired services with kolla.' atm I'm using docker-compose locally to build and deploy things and then pretty much copying the instructions with modifications for | 17:06 |
Zara | kolla (config etc) so that kolla can use them. | 17:06 |
Zara | (btw, it might help users in the future if the history of the kolla project were noted somewhere; it took me a while to realise that 'kolla-ansible' used to be within 'kolla' and is now a separate project, and 'kolla' just refers to the docker part of kolla (except for things like 'openstack-kolla' which refers to both kolla and kolla-ansible).) | 17:13 |
Zara | (orrr if I've gotten that wrong then maybe that proves it! :D) | 17:15 |
*** mdnadeem has quit IRC | 17:20 | |
*** pcaruana has joined #openstack-kolla | 17:22 | |
*** lpetrut has quit IRC | 17:23 | |
inc0 | whats aiui?:) | 17:26 |
inc0 | Zara: | 17:26 |
*** pcaruana has quit IRC | 17:26 | |
Zara | oh sorry, 'as I understand it' | 17:27 |
inc0 | ah | 17:27 |
Zara | I'm pretty new to everything so it's very possible I've missed things | 17:27 |
*** pcaruana has joined #openstack-kolla | 17:27 | |
Zara | so I'm qualifying every single statement with that, haha | 17:27 |
*** ntpttr_laptop has joined #openstack-kolla | 17:28 | |
inc0 | I'm not sure if I understand your question, do you want to build single container? and then deploy single container? | 17:28 |
Zara | yeah, with the idea that that we could then reuse the build and deploy instructions in a setup where we had lots of containers, including those for openstack services | 17:29 |
inc0 | well just building is fine | 17:30 |
*** b_bezak has quit IRC | 17:30 | |
inc0 | you can really build however and whatever you want, even single img | 17:30 |
*** gkadam has quit IRC | 17:30 | |
codenamelxl | Hi everyone, it looks like that Kolla with DVR on fails on Floating IP located on network node | 17:30 |
codenamelxl | FIP on compute node is OK | 17:30 |
inc0 | that's how DVR work tho | 17:31 |
inc0 | fip lands on cpu | 17:31 |
inc0 | (afair, not expert on dvr) | 17:31 |
codenamelxl | not with centralized router and VIP like Octavia. For those use case, the FIP will be located on network node | 17:31 |
codenamelxl | FIP on network node can't be reach | 17:32 |
inc0 | hmm...interesting, can you confirm that you have running l3 agent on network node? | 17:32 |
codenamelxl | Yes, it's running | 17:33 |
codenamelxl | i can ping the FIP from SNAT/centralized router namespace | 17:34 |
inc0 | hmm...I think our inventory is tailored for dvr on cpu | 17:34 |
inc0 | so where is it terminated? | 17:34 |
codenamelxl | sorry, can you be clearer? | 17:35 |
inc0 | well if you can ping fip from router, it has to be somewhere right? | 17:35 |
codenamelxl | it reach the VM/VIP correctly | 17:35 |
inc0 | ahh but not octavia? | 17:36 |
codenamelxl | No, it's floating IP on network node in general | 17:36 |
inc0 | correct me if I'm wrong, but octavia lb lives inside vm? | 17:36 |
codenamelxl | yes | 17:37 |
*** affie_ has quit IRC | 17:37 | |
kfox1111 | octavia is weird though. | 17:37 |
kfox1111 | amphora live as vm's in the cloud, | 17:37 |
codenamelxl | but the VIP FIP is on network node | 17:37 |
kfox1111 | but use keepalived for service if Iremember correctly. | 17:37 |
kfox1111 | but they have fip that binds to the vip. | 17:37 |
kfox1111 | so the fip lives on the network node, so it is fixed, while the vip moves around | 17:37 |
*** openstackgerrit has joined #openstack-kolla | 17:38 | |
openstackgerrit | Merged openstack/kolla stable/pike: Move a release note generated in wrong place https://review.openstack.org/509103 | 17:38 |
kfox1111 | (not how I would have chosen to implement it. :/ ) | 17:38 |
inc0 | right, that seems to be conflicting with what DVRs are supposed to be right? | 17:38 |
kfox1111 | kind of, yeah. | 17:38 |
kfox1111 | it use to be flat out broken when dvr was enabled, | 17:39 |
kfox1111 | so a step up at least. :/ | 17:39 |
Zara | inc0: thanks, I need to go back over the problem because I think I'm not expressing it clearly (or possibly just confused haha) :) | 17:39 |
codenamelxl | https://bugs.launchpad.net/neutron/+bug/1583694 | 17:39 |
openstack | Launchpad bug 1583694 in neutron "[RFE] DVR support for Allowed_address_pair port that are bound to multiple ACTIVE VM ports" [Wishlist,Fix released] - Assigned to Swaminathan Vasudevan (swaminathan-vasudevan) | 17:39 |
kfox1111 | I kind of gave up on octavia. :( | 17:39 |
*** signed8bit is now known as signed8bit_Zzz | 17:39 | |
kfox1111 | yeah. that issue. | 17:39 |
codenamelxl | And this fixed in neutron make the FIP binded to VIP stay on the network node | 17:39 |
kfox1111 | I waited a long time for them to fix it. | 17:39 |
kfox1111 | yeah. that should fix it. just I dont know of too many people who have tested that. | 17:40 |
kfox1111 | does regular network node traffic work? | 17:41 |
kfox1111 | ie, vm without a fip accessing data from outside the cloud? | 17:41 |
codenamelxl | yes, inside to outside is OK, the other way is not | 17:43 |
codenamelxl | oh wait, let me check that again | 17:45 |
*** sambetts is now known as sambetts|afk | 17:45 | |
rwellum | inc0: I think Zara is talking more about a dev environment? | 17:46 |
inc0 | well, we do it differently tho | 17:46 |
inc0 | also it's still wip | 17:46 |
*** k_mouza_ has joined #openstack-kolla | 17:48 | |
*** jmccarthy has joined #openstack-kolla | 17:51 | |
codenamelxl | kfox1111 let me check that in 15mins | 17:51 |
*** k_mouza has quit IRC | 17:52 | |
*** k_mouza_ has quit IRC | 17:53 | |
Zara | okay I think my issue was that anything built with kolla had to be deployed with kolla, but there was no way to deploy something in isolation via kolla without other openstack services being configured. so we can't use the results of kolla build to develop locally. | 17:57 |
keper7 | hey all, so i've given up on kolla-k8s for now since it's not prod ready and gone with kolla-ansible. I'm seeing the cinder API container repeatedly restarting with error "Permission denied: AH00058: Error retrieving pid file /run/httpd/httpd.pid" though. Looks like its this problem https://bugs.launchpad.net/kolla-ansible/+bug/1728916. Does anyone know a workaround? | 17:57 |
openstack | Launchpad bug 1728916 in kolla-ansible "centos-source-fluentd:4.0.0 and cinder api keep restarting" [Undecided,Confirmed] | 17:57 |
*** pcaruana has quit IRC | 17:58 | |
*** itlinux has joined #openstack-kolla | 17:58 | |
keper7 | I see the comment saying "please use the same version of kolla and kolla-ansible", but I just used the quickstart instructions for those so I'm not sure what to do differently | 18:01 |
*** diablo_rojo has quit IRC | 18:02 | |
inc0 | keper7: that's strange, never seen this issue before (httpd.pid perms) | 18:03 |
inc0 | did you build your containers? | 18:03 |
*** codenamelxl has quit IRC | 18:03 | |
*** signed8bit_Zzz is now known as signed8bit | 18:03 | |
*** mgoddard has quit IRC | 18:04 | |
*** gfidente is now known as gfidente|afk | 18:06 | |
*** ntpttr_laptop has quit IRC | 18:07 | |
jmccarthy | If a neutron plugin has no limits applied on an item in it's requirements.txt (i.e. https://github.com/openstack/networking-vpp/blob/master/requirements.txt) and it's installed with: pip --no-cache-dir install --upgrade -c requirements/upper-constraints.txt /plugins/*; it can override the upper-constraint ? (oslo.serialization) | 18:07 |
keper7 | no I pulled them. I do have some that I built but I didn't use them this time | 18:07 |
*** codenamelxl has joined #openstack-kolla | 18:07 | |
*** jbrooks has quit IRC | 18:07 | |
*** dardelean has joined #openstack-kolla | 18:08 | |
inc0 | ok, if you pulled 4.0.0 they're *very* old | 18:08 |
dardelean | hi | 18:08 |
keper7 | I have `openstack_release: "4.0.0"`, and the rest of the container/build settings left as default in my globals | 18:08 |
keper7 | oh | 18:08 |
inc0 | fortunately we have fresh and more supported ones | 18:08 |
keper7 | where? the newest i see is on docker hub. I can use the ones i just put up on jfrog but it was my first time trying that so i dont know if they're good or not | 18:08 |
inc0 | if you want pike containers just type "pike" in openstack_releease | 18:08 |
keper7 | oh neat. Thanks, I'll wipe the nodes and try again with that | 18:09 |
inc0 | caveat, only source builds are there | 18:09 |
codenamelxl | I just checked, VM can reach other VM through their IP, can't reach public IP | 18:09 |
inc0 | so centos-source or ubuntu-source | 18:09 |
keper7 | openstack_release: "pike", kolla_install_type: "source", kolla_base_distro: "ubuntu" should work? | 18:09 |
inc0 | https://hub.docker.com/r/kolla/ubuntu-source-glance-api/tags/ | 18:10 |
inc0 | yes | 18:10 |
inc0 | do kolla-ansible destroy --include-images before | 18:10 |
inc0 | just to be sure | 18:10 |
dardelean | if I want to use TLS, is it mandatory to provide fqdn? | 18:10 |
inc0 | yes dardelean | 18:11 |
keper7 | oh neat i can do that instead of re-deploying my base-os? (ubuntu) | 18:11 |
dardelean | I see | 18:11 |
inc0 | keper7: yes | 18:11 |
keper7 | great! | 18:11 |
dardelean | planning to write a blog post on deploying kolla in HA, I know I had some issues with that | 18:12 |
inc0 | well HA is default | 18:12 |
inc0 | let me know what issues you have | 18:13 |
keper7 | dardelean: what didnt work as HA? | 18:13 |
inc0 | codenamelxl: seems like router in general is fleaky, check neutron logs (/var/lib/docker/volumes/kolla_logs/_data/neutron) | 18:14 |
dardelean | it's not that it didn't work, just some stuff that we're not documente, e.g. was expecting glance to be present on al the control nodes, turned out that if I use file backend it does not shace | 18:14 |
dardelean | which is normal, it just wasn't obvious for me at the time | 18:14 |
inc0 | right, if you use file backend to glance you'll have issues | 18:14 |
dardelean | also I had some issues with TLS at that time on stable/pike, did not manage to solve them at the time | 18:15 |
codenamelxl | inc0: OK, i'm tailing log on l3 agent of network node | 18:15 |
dardelean | anyways, just want to create some traction for kolla with an HA article, ppl tent to like that | 18:17 |
inc0 | dardelean: yup, I love that too:) | 18:17 |
Zara | (so if kolla-build isn't an option to make an image that can be deployed locally without configuring other openstack services, and we wanted to reuse the dockerfiles we were using with kolla-build, we'd need another tool that could parse jinja-templated dockerfiles and build them.) | 18:18 |
inc0 | kolla | 18:18 |
inc0 | kolla-ansible doesn't make images | 18:18 |
inc0 | kolla itself make images and they totally can and are deployed outside kolla-ansible | 18:18 |
*** jmccarthy has left #openstack-kolla | 18:19 | |
inc0 | kolla-build is tool that does that:) | 18:19 |
Zara | hm, perhaps that's the step I was missing, as far as I was aware things built by kolla-build could only be deployed with kolla-ansible | 18:19 |
keper7 | inc0: `kolla-ansible destroy --include-images -i multinode --yes-i-really-really-mean-it` failed. Several pages of red... Lots of "reference does not exist". It did remove all images and containers from my nodes though, so I think thats ok right? | 18:19 |
inc0 | should be fine | 18:20 |
keper7 | thanks | 18:20 |
inc0 | Zara: right, they're just images, but default CMD is kinda specific with kolla | 18:20 |
inc0 | you can either override default CMD or make use of it (it is useful;)) | 18:21 |
codenamelxl | inc0: OK so what do i expect from the log :) I'm kinda new on this | 18:21 |
inc0 | not sure...we need to debug what's wrong:) | 18:21 |
inc0 | Zara: this file is a key https://github.com/openstack/kolla-ansible/blob/master/ansible/roles/cinder/templates/cinder-api.json.j2 | 18:22 |
codenamelxl | hold on, let's me switch my IRC client | 18:22 |
inc0 | I don't have much time now, but I'll try to write some docs to use it | 18:22 |
Zara | np, my memory's hazy and I'm ill anyway so planning to vanish shortly :) I just remember a while back I tried `kolla-build foo`, then effectively `docker run foo` and it said the container didn't have some config and it failed to start. | 18:23 |
inc0 | right, you'll need few ENVs | 18:24 |
inc0 | but you can totally do it assuming you provide them | 18:24 |
Zara | okay, hm, is there a list anywhere? | 18:24 |
*** lpetrut has joined #openstack-kolla | 18:25 | |
inc0 | so if you override CMD, you just do what you specify | 18:25 |
inc0 | so if you run docker run -v my-nova-conf:/etc/nova nova-api nova-api | 18:25 |
inc0 | that should be fine | 18:25 |
*** codenamelxl_ has joined #openstack-kolla | 18:26 | |
codenamelxl_ | INFO neutron.agent.common.ovs_lib [req-3ddb49e3-38da-423f-8cb7-f7e726ae1941 - - - - -] Port 372ac4ed-6fcc-4224-bbe8-be30768dbe38 not present in bridge br-int | 18:27 |
Zara | this is already very useful, thanks :) are there any examples out there already anywhere? | 18:27 |
codenamelxl_ | ok, this is the most suspicous line | 18:27 |
codenamelxl_ | that port does not show up when i do a port list | 18:27 |
*** codenamelxl has quit IRC | 18:27 | |
inc0 | Zara: well you can definetly reverse engineer tripleo (they're using just containers) but well | 18:28 |
inc0 | reverse engineering tripleo is not easierst thing | 18:28 |
Zara | haha | 18:28 |
Zara | I'll probably try to set something up tomorrow since it'll be easier to show where I'm getting stuck with the env vars when I can point to actual error messages. | 18:32 |
Zara | thanks for all your help so far :) | 18:33 |
rwellum | inc0: we have pike images on dockerhub now. :) | 18:35 |
rwellum | inc0: Somehow missed this | 18:36 |
inc0 | yeah just got merged | 18:41 |
inc0 | Zara: tomorrow I'll also have more time (I'm on conference now) so I can walk you through it | 18:41 |
rwellum | inc0: +1 :) | 18:41 |
rwellum | Zara: I deploy kolla images outside of kolla-ansible fwiw | 18:41 |
dardelean | rwellum where are the pike images? cannot see the tag on the images on kollas dockerhub | 18:42 |
inc0 | dardelean: they're ther | 18:42 |
rwellum | dardelean: https://hub.docker.com/r/kolla/ubuntu-source-glance-api/tags/ | 18:42 |
inc0 | just source images - tag is pike | 18:42 |
dardelean | hmm, interesting | 18:43 |
*** alexasap has joined #openstack-kolla | 18:43 | |
dardelean | ubuntu source collectd does not have pike | 18:43 |
*** mgoddard has joined #openstack-kolla | 18:43 | |
*** alexasap has left #openstack-kolla | 18:43 | |
dardelean | https://hub.docker.com/r/kolla/ubuntu-source-collectd/tags/ | 18:43 |
*** alexasap has joined #openstack-kolla | 18:43 | |
alexasap | Hello | 18:43 |
alexasap | please help asap | 18:43 |
alexasap | https://gist.github.com/anonymous/4f42bf260e2506c9645f531b5063e0fb | 18:43 |
alexasap | installed openstack kolla with ODL | 18:44 |
alexasap | no traffic for DHCP interface and ODL carbon crashing | 18:44 |
codenamelxl_ | i'm confused as where to report this FIP problem to kolla or neutron now since i can confirm that it's easy to reproduce | 18:45 |
alexasap | sorry? | 18:47 |
alexasap | any idea how to fix that (at least I want to just boot VMs and get DHCP address) | 18:48 |
rwellum | dardelean: could be because it's not pushed yet? inc0 said it was just merged. | 18:49 |
inc0 | no it's there | 18:49 |
alexasap | it doesnt work for the private virtual and physical networks too | 18:49 |
alexasap | ie. ODL in Kolla is broken | 18:50 |
inc0 | but I don't think collectd is in profiles we psh | 18:50 |
dardelean | i see, indeed for keystone there are new images | 18:50 |
*** kbaegis has joined #openstack-kolla | 18:52 | |
kbaegis | Hi all. troubleshooting a config issue with kolla-ansible | 18:52 |
inc0 | alexasap: not sure what to make out of these logs, I'm not expert at odl | 18:53 |
kbaegis | specifically /etc/kolla/config/{cinder,glance,nova} with ceph external | 18:54 |
kbaegis | Are there requirements for the permissions? I need to use the deploy user or the files aren't detected/deployed? | 18:54 |
kbaegis | specifically, do I need to chown -R root:root or kolla:kolla for /etc/kolla/config/? | 18:55 |
*** codenamelxl_ has quit IRC | 18:55 | |
kbaegis | And are there any requirements for rwx? | 18:55 |
*** david-lyle has joined #openstack-kolla | 18:55 | |
*** shardy has quit IRC | 18:58 | |
alexasap | 2017-12-05 18:58:31,523 | ERROR | pool-44-thread-1 | OutboundQueueProviderImpl | 314 - org.opendaylight.openflowplugin.impl - 0.4.1.Carbon | No queue present, failing request 2017-12-05 18:58:31,523 | WARN | pool-44-thread-1 | RpcContextImpl | 314 - org.opendaylight.openflowplugin.impl - 0.4.1.Carbon | Xid cannot be reserved for new RequestContext, node:openflow:53244719937371 | 18:58 |
alexasap | any idea? | 18:58 |
inc0 | you might want to check with ODL people about that | 19:00 |
*** david-lyle has quit IRC | 19:01 | |
dardelean | hmm, i can see that "tunnel_interface" can be used for guest traffic | 19:06 |
dardelean | but what if I want it on non-tunnel network | 19:07 |
dardelean | for example my neutron-agents on the compute nodes don't support vxlan | 19:07 |
dardelean | till now I created a br-data by hand | 19:07 |
dardelean | is there a better way in your opinion? | 19:08 |
inc0 | well depends on what kind of neutron net you want | 19:09 |
inc0 | if it's vlan, you'll have to do some fiddling | 19:09 |
inc0 | not much, but some | 19:09 |
kbaegis | I'm seeing "VolumeBackendAPIException: Bad or unexpected response from the storage volume backend API: Error connecting to ceph cluster." | 19:10 |
*** TxGirlGeek has joined #openstack-kolla | 19:10 | |
kbaegis | in cinder_volume logs | 19:10 |
dardelean | vlan or flat | 19:10 |
*** janonymous has quit IRC | 19:10 | |
inc0 | kbaegis: can you connect to ceph cluster?:) | 19:10 |
inc0 | dardelean: flat should work if your br-ex on cpu nodes is connected | 19:10 |
inc0 | and you spawn vm in it | 19:11 |
kbaegis | Sure. Ceph seems solid. osdmap e661: 24 osds: 24 up, 24 in | 19:11 |
dardelean | this is what I do https://paste.ubuntu.com/26120370/ | 19:11 |
inc0 | vlan - you'll need to do conf override for ml2_conf.ini where you'll specify physnet, vlan ranges and whatnot | 19:11 |
inc0 | like you would normally | 19:11 |
dardelean | eth2 being my dedicated guest network | 19:11 |
dardelean | works | 19:12 |
inc0 | kbaegis: hmm...even from node that runs cinder volume? | 19:12 |
dardelean | but I am thinking if I can do it better | 19:12 |
inc0 | then if you just specify neutron_external_interface: "eth2" | 19:12 |
inc0 | create flat network | 19:12 |
inc0 | and then spawn vm in it | 19:12 |
inc0 | that should work | 19:12 |
dardelean | I don't get this, why neutron external interface? that (br-ex) is only on the control nodes | 19:13 |
dardelean | i use tenant networks with dedicated guest network | 19:14 |
*** TxGirlGeek has quit IRC | 19:14 | |
kbaegis | inc0: Yeah. Ceph is healthy. It's the config/notes on ceph-external that appear to have some gaps maybe? | 19:15 |
inc0 | well, only on network nodes technically | 19:15 |
kbaegis | inc0: I'm attempting to get ceph to connect from the container directly | 19:15 |
*** TxGirlGeek has joined #openstack-kolla | 19:16 | |
dardelean | control = network nodes in my case | 19:16 |
kbaegis | I'm guessing I misconfigured something | 19:16 |
inc0 | should be fine, we use net=host in kolla-ansible everywhere | 19:16 |
inc0 | dardelean: well technically network nodes are ones to connect to external network | 19:16 |
inc0 | vxlans provides horizontal traffit | 19:17 |
inc0 | I think you should be able to spawn vm in flat network (haven't tried that myself) | 19:17 |
dardelean | yeah, east-west traffic | 19:18 |
dardelean | while br-ex is north-south | 19:18 |
kbaegis | inc0: So /etc/kolla/config/cinder/ contains ceph.client.cinder.keyring ceph.conf cinder-volume.conf & cinder-backup.conf | 19:18 |
inc0 | yeah | 19:18 |
dardelean | it would be nice if I could create a east-west network without vxlan | 19:18 |
dardelean | basically creating br-data | 19:18 |
kbaegis | I see none of those in the cinder_volume container | 19:19 |
dardelean | on all nodes, cpu and controllers | 19:19 |
dardelean | br-tun is create afaik if tunnel_traffic interfaceis used | 19:19 |
kbaegis | inc0: /etc/kolla/cinder-volume/ contains only ceph cinder and config.json | 19:20 |
kbaegis | no keys | 19:20 |
inc0 | kbaegis: interesting, it should be there in /etc/ceph | 19:20 |
inc0 | ahh | 19:20 |
kbaegis | only ceph.conf and rbdmap | 19:20 |
kbaegis | No keys in the container | 19:20 |
inc0 | well | 19:21 |
inc0 | so keyring hasn't been copied to /etc/ceph in container? | 19:21 |
kbaegis | correct. Nor is it in the appropriate folder for kolla-ansible | 19:21 |
kbaegis | for the deploy node | 19:22 |
kbaegis | If I copy it manually will it work? cp /etc/kolla/config/cinder/ceph.*.keyring /etc/kolla/cinder-volume/ | 19:22 |
inc0 | check if it's present in /var/lib/kolla/config_files inside container | 19:23 |
kbaegis | only 3 exist | 19:23 |
kbaegis | no keyrings | 19:23 |
inc0 | try to copy it manually and restart cinder-volume container | 19:23 |
inc0 | yeah nothing would copy ceph confs there, seems like a bug | 19:25 |
kbaegis | docker cp /etc/kolla/config/cinder/ceph.client.cinder.keyring cinder_volume:/etc/ceph/ ->>Driver post RPC initialization completed successfully | 19:25 |
*** jbrooks has joined #openstack-kolla | 19:25 | |
kbaegis | Well I'm glad it wasn't just me :) | 19:26 |
*** alexasap has quit IRC | 19:29 | |
kbaegis | Hmm | 19:29 |
kbaegis | Well will inserting the file into /etc/kolla/cinder-volumes at least get it under /var/lib/kolla/config_files? | 19:30 |
kbaegis | inc0? | 19:30 |
inc0 | yes | 19:31 |
kbaegis | Awesome. Makes a workaround easier until that's fixed | 19:31 |
inc0 | however that has to be done on every cinder-volume ndoes | 19:31 |
kbaegis | I can't just do it on the deployment node and then do a reconfigure? | 19:31 |
inc0 | no:( | 19:32 |
kbaegis | *sigh | 19:32 |
kbaegis | inc0: Thanks | 19:32 |
kbaegis | inc0: Nice to have a sanity check on this stuff :) | 19:32 |
inc0 | but if you copy it to all nodes it should be safe to reconfigure or whatnot | 19:32 |
*** lukepatrick has quit IRC | 19:33 | |
kbaegis | Let's see- the other component that really needs this is nova-compute | 19:36 |
kbaegis | so cinder_volume and nova_compute, or is it nova_libvirt? | 19:36 |
inc0 | copy to both to be safe;) | 19:37 |
inc0 | btw ansible -m copy -i your-inv -a "src=your-keyring-file dest=/etc/kolla/nova-compute/" storage compute | 19:38 |
inc0 | ansible ad-hoc commands are dope | 19:38 |
kbaegis | nice, tyvm | 19:40 |
kbaegis | better than my: for i in $(cat /etc/hosts|grep bumblebee|cut -d\ -f2);do scp /etc/kolla/config/nova/*.keyring $i:/etc/ceph/;done | 19:40 |
inc0 | yeah I strongly recommend getting familiar with this, I use ansible -m shell -a "whatever oneliner" all routnely | 19:41 |
inc0 | -m script -a "src=/path/to/some/shell.sh" is also mighty useful | 19:42 |
kbaegis | ERROR! Extraneous options or arguments | 19:44 |
kbaegis | ansible -m copy -i ~/bb -a "src=/etc/kolla/config/nova/ceph.client.cinder.keyring dest=/etc/kolla/nova-compute/" storage compute | 19:44 |
kbaegis | storage and compute need a flag presumably? Can't seem to find it | 19:45 |
kbaegis | Okay, got all the files squared away | 19:49 |
kbaegis | Now kolla-ansible reconfigure will work? | 19:49 |
*** lukepatrick has joined #openstack-kolla | 19:51 | |
*** jbrooks has quit IRC | 19:57 | |
*** jbrooks has joined #openstack-kolla | 19:58 | |
*** wojdec1 has joined #openstack-kolla | 20:04 | |
*** gkadam has joined #openstack-kolla | 20:05 | |
*** jrist has quit IRC | 20:05 | |
*** lpetrut has quit IRC | 20:11 | |
*** mgoddard has quit IRC | 20:13 | |
kbaegis | Well you're right about ansible. I should invest | 20:15 |
kbaegis | inc0: But here's my hideous yet functional one-liner- or i in $(cat tmp);do ssh $i docker exec -it -u root glance_api bash -c 'cp /var/lib/kolla/config_files/*.keyring /etc/ceph/'&& docker restart glance_api;docker exec -it -u root cinder_volume bash -c 'cp /var/lib/kolla/config_files/*.keyring /etc/ceph/'&& docker restart cinder_volume;docker exec -it -u root nova_libvirt bash -c 'cp /var/lib/kolla/config_files/*.keyring /etc/c | 20:15 |
kbaegis | eph/'&& docker restart nova_libvirt;docker exec -it -u root nova_compute bash -c 'cp /var/lib/kolla/config_files/*.keyring /etc/ceph/'&& docker restart nova_compute;do | 20:15 |
kbaegis | *f | 20:16 |
*** ktibi has quit IRC | 20:17 | |
*** yangyapeng has joined #openstack-kolla | 20:20 | |
*** yangyapeng has quit IRC | 20:24 | |
*** b_bezak has joined #openstack-kolla | 20:31 | |
*** wojdec has joined #openstack-kolla | 20:31 | |
*** rwsu has quit IRC | 20:31 | |
*** k_mouza has joined #openstack-kolla | 20:31 | |
*** rwsu has joined #openstack-kolla | 20:31 | |
*** wojdec1 has quit IRC | 20:33 | |
*** jbrooks has quit IRC | 20:34 | |
*** wojdec has quit IRC | 20:35 | |
*** k_mouza has quit IRC | 20:35 | |
*** b_bezak has quit IRC | 20:35 | |
*** wojdec has joined #openstack-kolla | 20:36 | |
*** jbrooks has joined #openstack-kolla | 20:36 | |
wojdec | Hi Folks… I'm trying to locally build a customized container using the kolla tools/build.py with packages installed from an external repo. The repo's yum settings are "COPY" 'ed in the dockerfile for the container. On running the build however I get an error saying that the yum repo couldn't be resolved, but resolution works fine from the host running the build. Any suggestions on how to enable external repos, without a http | 20:39 |
wojdec | accessibe repo file? | 20:39 |
*** k_mouza has joined #openstack-kolla | 20:43 | |
*** unicell has joined #openstack-kolla | 20:48 | |
*** k_mouza has quit IRC | 20:54 | |
*** kelv_ has joined #openstack-kolla | 20:55 | |
kbaegis | wojdec: I'd check your docker config | 21:09 |
kbaegis | likely a dns config issue (--dns=…) | 21:09 |
*** Pavo has joined #openstack-kolla | 21:10 | |
inc0 | also try creating fake dockerfile with just COPY your repo configs and do docker run -it << my fake image>> bash | 21:11 |
inc0 | and see if it works | 21:11 |
*** kbaegis has quit IRC | 21:14 | |
*** k_mouza has joined #openstack-kolla | 21:15 | |
*** jbrooks has quit IRC | 21:22 | |
wojdec | kbaegis: I'm using the same pattern as the other dockerfiles.j2 files… not passing DNS anywhere, nor is it mentioned in the kolla docs as needed… Is it? | 21:27 |
*** gfidente|afk has quit IRC | 21:29 | |
inc0 | wojdec: docker engine might mess with that, that's why I suggested just running clean centos7 and see if that works manually | 21:30 |
wojdec | ok. will try that next | 21:31 |
*** dardelea_ has joined #openstack-kolla | 21:31 | |
inc0 | that will help figuring out if problem is kolla or docker itself | 21:32 |
*** dardelea_ has quit IRC | 21:35 | |
*** k_mouza has quit IRC | 21:36 | |
*** mchlumsky has quit IRC | 21:37 | |
*** gkadam has quit IRC | 21:38 | |
*** itlinux has quit IRC | 21:39 | |
*** itlinux has joined #openstack-kolla | 21:44 | |
*** k_mouza has joined #openstack-kolla | 21:45 | |
*** jrist has joined #openstack-kolla | 21:46 | |
*** lrensing has joined #openstack-kolla | 21:50 | |
*** k_mouza has quit IRC | 21:56 | |
*** kbaegis has joined #openstack-kolla | 21:58 | |
*** rhallisey has quit IRC | 21:59 | |
*** zshi has quit IRC | 22:03 | |
*** threestrands has joined #openstack-kolla | 22:05 | |
*** threestrands has quit IRC | 22:05 | |
*** threestrands has joined #openstack-kolla | 22:05 | |
wojdec | seems like it's a docker issue, or more specifically how docker hooks up into my host that's connected to a VPN... | 22:07 |
wojdec | external repos are reachable. internal ones not | 22:07 |
*** jtriley has quit IRC | 22:08 | |
wojdec | inc0: solved it by flipping the source of the package to be off our external repo than the intranet one | 22:08 |
*** dave-mcc_ is now known as dave-mccowan | 22:15 | |
rwellum | inc0: these pike images have the tag 'pike'? Not 5.0.0 etc? | 22:18 |
*** kbaegis has quit IRC | 22:19 | |
*** kbaegis has joined #openstack-kolla | 22:19 | |
*** wojdec has quit IRC | 22:31 | |
*** jascott1 has quit IRC | 22:31 | |
*** jascott1 has joined #openstack-kolla | 22:32 | |
dardelean | rwellum did you do a presentation @summit? | 22:32 |
rwellum | dardelean: yes - combined with vhosakot | 22:33 |
dardelean | can you provide the video? :D | 22:36 |
*** jascott1 has quit IRC | 22:36 | |
dardelean | searched for it some time ago but could not find it | 22:36 |
rwellum | dardelean: https://www.youtube.com/watch?v=14WjpNC7HKI | 22:41 |
dardelean | great, thanks rwellum :D | 22:42 |
rwellum | It all went smoothly, except I was using his laptop, and my demo was pre-recorded, and it failed to run on his laptop. Had to log in, 2-factor authentication etc. | 22:42 |
dardelean | you should do a blog post on this stuff | 22:42 |
dardelean | i mean the kolla-k8s script you have | 22:42 |
dardelean | it would be easier for ppl to find and use it | 22:43 |
rwellum | dardelean: I do have slides up: https://docs.google.com/presentation/d/1Qlqgqwbbjd-eDUCeOmj2Ugpbbke7W99qxc7KPp_BxnE/edit?usp=sharing | 22:43 |
rwellum | This is what I presented | 22:44 |
dardelean | i see, but still | 22:44 |
dardelean | if you don't do it, I may do it :)) | 22:44 |
rwellum | Could do thing is it's just another deployer basically. I'd rather the focus was on Justin's ansible playbook. | 22:45 |
dardelean | what is that? | 22:45 |
rwellum | ko.py was/is great for me to play with the technology - but not really the way to go in this data-driven environment. :) | 22:45 |
dardelean | oh, ofc | 22:45 |
dardelean | but a nice place to start | 22:45 |
rwellum | Yeah I hope so. It's gotten a lot of people unstuck and using it for the first time. | 22:46 |
rwellum | https://review.openstack.org/#/c/520847/ | 22:46 |
rwellum | dardelean: ^^ that's his playbook | 22:47 |
dardelean | hmm, what does this do? | 22:47 |
dardelean | I mean, it's something like your script? | 22:47 |
rwellum | Multinode kolla-kubernetes deployment playbook with ceph supported. AFAIK. | 22:47 |
dardelean | ah | 22:47 |
rwellum | Yeah but done with ansible. | 22:47 |
rwellum | It's currently ocata only - my tool does pike | 22:48 |
rwellum | as well as ocata | 22:48 |
*** kbaegis1 has joined #openstack-kolla | 22:51 | |
*** kbaegis has quit IRC | 22:52 | |
rwellum | inc0: not sure if you're still around, but definitely I don't see all the images for pike. For example compare: https://hub.docker.com/r/rwellum/centos-source-kubernetes-entrypoint/tags/ with https://hub.docker.com/r/kolla/centos-source-kubernetes-entrypoint/tags/ | 22:52 |
*** signed8b_ has joined #openstack-kolla | 22:54 | |
*** affie_ has joined #openstack-kolla | 22:56 | |
rwellum | dardelean: sorry I missed your presentation btw - meant to attend and say hello. | 22:57 |
*** signed8bit has quit IRC | 22:58 | |
*** kbaegis1 has quit IRC | 22:59 | |
*** kbaegis has joined #openstack-kolla | 23:00 | |
*** kbaegis has quit IRC | 23:03 | |
dardelean | rwellum my presentation? I was not @summit | 23:05 |
*** kbaegis has joined #openstack-kolla | 23:05 | |
rwellum | dardelean: bad with names - sorry! | 23:06 |
dardelean | :D | 23:06 |
rwellum | Overall dardelean the summit was great, but it was a long way to travel! And very much driven by companies - not as useful as the PTG's imo. | 23:10 |
dardelean | rwellum, I get you | 23:11 |
*** lrensing has quit IRC | 23:16 | |
kbaegis | inc0: Even with the hack/workaround for ceph-external, I can't create volumes with cinder-volume | 23:17 |
kbaegis | inc0: Might have been a keyring issue as I get an issue with rbd, but not with ceph —user cinder —cluster ceph lspools | 23:17 |
kbaegis | rbd returns an Operation not permitted or something | 23:18 |
*** goldyfruit has quit IRC | 23:21 | |
*** dave-mccowan has quit IRC | 23:23 | |
*** kbaegis has quit IRC | 23:26 | |
*** kbaegis has joined #openstack-kolla | 23:26 | |
*** kbaegis has quit IRC | 23:26 | |
*** signed8b_ is now known as signed8bit_Zzz | 23:27 | |
*** genek has quit IRC | 23:34 | |
*** Pavo has quit IRC | 23:35 | |
*** lrensing has joined #openstack-kolla | 23:40 | |
*** TxGirlGeek has quit IRC | 23:43 | |
inc0 | rwellum: yeah we limited images we build, probably can remove this limit | 23:45 |
rwellum | inc0: +1 then I can test pike, well outside my own images that is. | 23:47 |
*** lukepatrick has quit IRC | 23:47 | |
*** Pavo has joined #openstack-kolla | 23:48 | |
*** Pavo has quit IRC | 23:49 | |
*** Pavo has joined #openstack-kolla | 23:50 | |
*** yangyapeng has joined #openstack-kolla | 23:50 | |
*** dardelean has quit IRC | 23:53 | |
*** yangyapeng has quit IRC | 23:55 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!