openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 00:03 |
---|---|---|
openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 00:03 |
*** bmace has joined #kolla | 00:06 | |
*** vinkman has joined #kolla | 00:06 | |
sbezverk | sdake is it the same depency that magnum has on keystonev3? | 00:07 |
sbezverk | dependecy | 00:07 |
sdake | that is a different issue | 00:07 |
*** dims has joined #kolla | 00:07 | |
sbezverk | sdake in this case keystone dependency looks more serious :-) without keystonev3 it will never work in kolla | 00:09 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Rework heka log copying https://review.openstack.org/294206 | 00:12 |
sdake | we have keystone v3 | 00:15 |
sbezverk | right bug magnum does not support it | 00:17 |
sbezverk | and kolla does not support keystone v2 | 00:17 |
SamYaple | sbezverk: kolla doesnt create v2 enoints by default | 00:21 |
SamYaple | technically they work just fine with kolla | 00:21 |
SamYaple | operator just needs to add the endpoint themselves | 00:21 |
*** weiyu_ has quit IRC | 00:22 | |
sdake | magnum needs porting to keystone v3 rther then putting that on the operators | 00:25 |
sdake | although i am pretty sure magnum requires barbican so it may not matter | 00:25 |
sdake | need m0ar gates | 00:26 |
SamYaple | did that bug say magnum worked with v3 it just wasnted tested sdake ? | 00:27 |
*** vinkman has left #kolla | 00:27 | |
sdake | SamYaple i wasn't casting blame | 00:28 |
SamYaple | i was asking questions about that magnum bug | 00:29 |
SamYaple | ? | 00:29 |
SamYaple | remember> | 00:29 |
sdake | right where i asked if it was tested | 00:29 |
sdake | you said no and i said ok future trouble spot iirc | 00:29 |
*** alisonh has quit IRC | 00:30 | |
openstackgerrit | Andrei-Lucian Șerb proposed openstack/kolla: Attach external NIC to a NAT-Network if on Wi-Fi https://review.openstack.org/294340 | 00:30 |
SamYaple | i dont recall | 00:31 |
*** ssurana has quit IRC | 00:33 | |
*** alisonh has joined #kolla | 00:49 | |
*** Jeffrey4l_ has joined #kolla | 01:01 | |
openstackgerrit | Andrei-Lucian Șerb proposed openstack/kolla: Attach external NIC to a NAT-Network if on Wi-Fi https://review.openstack.org/294340 | 01:10 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Fix 'optional' config.json values https://review.openstack.org/294350 | 01:11 |
*** weiyu_ has joined #kolla | 01:13 | |
*** salv-orlando has joined #kolla | 01:16 | |
openstackgerrit | Andrei-Lucian Șerb proposed openstack/kolla: Attach external NIC to a NAT-Network if on Wi-Fi https://review.openstack.org/294340 | 01:16 |
*** salv-orlando has quit IRC | 01:17 | |
*** vhosakot has joined #kolla | 01:17 | |
*** salv-orlando has joined #kolla | 01:18 | |
SamYaple | whats up with gerrit | 01:18 |
*** salv-orl_ has quit IRC | 01:19 | |
*** dims has quit IRC | 01:21 | |
*** salv-orlando has quit IRC | 01:22 | |
openstackgerrit | Andrei-Lucian Șerb proposed openstack/kolla: Attach external NIC to a NAT-Network if on Wi-Fi https://review.openstack.org/294340 | 01:24 |
*** SiRiuS_ has quit IRC | 01:31 | |
*** achanda has joined #kolla | 01:40 | |
*** rhallisey has quit IRC | 01:42 | |
*** iceyao has joined #kolla | 01:43 | |
*** iceyao has quit IRC | 01:44 | |
*** iceyao has joined #kolla | 01:45 | |
*** vinkman1 has joined #kolla | 01:50 | |
openstackgerrit | Merged openstack/kolla: Reorder operations of mkdir and useradd in toolbox https://review.openstack.org/290280 | 01:58 |
openstackgerrit | Merged openstack/kolla: Change libvirt volume name to libvirtd https://review.openstack.org/293106 | 02:06 |
openstackgerrit | Merged openstack/kolla: Add cron to pull playbooks https://review.openstack.org/294116 | 02:21 |
openstackgerrit | Merged openstack/kolla: Refactor glance config tasks https://review.openstack.org/294185 | 02:29 |
*** vinkman1 has left #kolla | 02:31 | |
*** ssurana has joined #kolla | 02:37 | |
openstackgerrit | Angus Salkeld proposed openstack/kolla-mesos: WIP: custom reload signal on config change https://review.openstack.org/294367 | 02:38 |
*** banix has quit IRC | 02:47 | |
*** ccesario has quit IRC | 02:48 | |
openstackgerrit | Sam Yaple proposed openstack/kolla: Add unmap to cinder for ceph https://review.openstack.org/294372 | 02:49 |
*** ccesario has joined #kolla | 02:50 | |
*** vhosakot has quit IRC | 03:09 | |
*** kjejlly_ has quit IRC | 03:25 | |
*** coolsvap has joined #kolla | 03:32 | |
*** vhosakot has joined #kolla | 03:33 | |
openstackgerrit | Merged openstack/kolla: Allow per node configuration with augments https://review.openstack.org/294098 | 03:45 |
*** iceyao_ has joined #kolla | 03:48 | |
*** Jeffrey4l_ is now known as Jeffrey4l | 03:49 | |
*** tfukushima has joined #kolla | 03:51 | |
*** iceyao has quit IRC | 03:52 | |
*** salv-orlando has joined #kolla | 03:56 | |
*** sdake_ has joined #kolla | 04:08 | |
*** salv-orlando has quit IRC | 04:10 | |
*** sdake has quit IRC | 04:11 | |
openstackgerrit | MD NADEEM proposed openstack/kolla: Correct var name in globals.yml https://review.openstack.org/294383 | 04:16 |
*** sdake_ has quit IRC | 04:18 | |
*** dave-mccowan has quit IRC | 04:21 | |
*** ssurana has quit IRC | 04:22 | |
*** ssurana has joined #kolla | 04:31 | |
*** sdake has joined #kolla | 04:33 | |
*** weiyu_ has quit IRC | 04:34 | |
*** coolsvap has quit IRC | 04:36 | |
*** larsks has quit IRC | 04:54 | |
*** signed8bit has quit IRC | 04:57 | |
*** ccesario_ has quit IRC | 04:57 | |
*** signed8bit has joined #kolla | 05:00 | |
*** larsks has joined #kolla | 05:00 | |
*** achanda has quit IRC | 05:01 | |
*** ssurana has quit IRC | 05:07 | |
*** ssurana has joined #kolla | 05:09 | |
openstackgerrit | Vikram Hosakote proposed openstack/kolla: Install Elasticsearch and Kibana packages using yum/apt-get https://review.openstack.org/291986 | 05:09 |
*** ccesario_ has joined #kolla | 05:13 | |
openstackgerrit | MD NADEEM proposed openstack/kolla: Fix inconsistancy in globals.yml https://review.openstack.org/294392 | 05:14 |
openstackgerrit | MD NADEEM proposed openstack/kolla: Fix inconsistancy in globals.yml https://review.openstack.org/294392 | 05:16 |
*** vinkman has joined #kolla | 05:17 | |
*** ccesario_ has quit IRC | 05:22 | |
*** salv-orlando has joined #kolla | 05:23 | |
*** salv-orlando has quit IRC | 05:34 | |
*** ccesario_ has joined #kolla | 05:34 | |
openstackgerrit | Merged openstack/kolla: Correct var name in globals.yml https://review.openstack.org/294383 | 05:37 |
*** unicell has quit IRC | 05:44 | |
*** unicell has joined #kolla | 05:45 | |
mdnadeem | sdake, SamYaple , Can we tag kolla to assert:supports-upgrade | 05:52 |
mdnadeem | >> http://governance.openstack.org/reference/tags/assert_supports-upgrade.html | 05:52 |
*** akwasnie has joined #kolla | 06:06 | |
vhosakot | akwasnie: Thanks for your comments! I have incorporated all the comments and uploaded new patch set https://review.openstack.org/#/c/291986/.. I have tested the patch set | 06:07 |
*** sdake has quit IRC | 06:09 | |
*** weiyu_ has joined #kolla | 06:11 | |
*** vinkman has quit IRC | 06:11 | |
*** tfukushima has quit IRC | 06:12 | |
*** coolsvap has joined #kolla | 06:12 | |
*** alisonh has left #kolla | 06:13 | |
openstackgerrit | MD NADEEM proposed openstack/kolla: Adding swift config parameters to all.yml https://review.openstack.org/294410 | 06:22 |
openstackgerrit | MD NADEEM proposed openstack/kolla: Fix inconsistancy in globals.yml https://review.openstack.org/294392 | 06:23 |
*** unicell1 has joined #kolla | 06:24 | |
*** Jeffrey4l has left #kolla | 06:24 | |
*** Jeffrey4l has joined #kolla | 06:24 | |
*** unicell has quit IRC | 06:25 | |
*** akwasnie has quit IRC | 06:27 | |
*** vhosakot has quit IRC | 06:29 | |
*** salv-orlando has joined #kolla | 06:33 | |
*** akwasnie has joined #kolla | 06:38 | |
*** ssurana has quit IRC | 06:39 | |
*** salv-orlando has quit IRC | 06:41 | |
*** ssurana has joined #kolla | 06:42 | |
*** tzn has joined #kolla | 06:47 | |
*** tzn has quit IRC | 06:48 | |
*** akwasnie has quit IRC | 06:53 | |
*** coolsvap has quit IRC | 06:53 | |
openstackgerrit | Merged openstack/kolla: Add unmap to cinder for ceph https://review.openstack.org/294372 | 06:53 |
*** coolsvap has joined #kolla | 06:53 | |
*** AndChat|241401 has joined #kolla | 06:55 | |
*** ccesario has quit IRC | 06:55 | |
*** akwasnie has joined #kolla | 07:07 | |
akwasnie | vhosakot: thanks! I will review it today | 07:07 |
openstackgerrit | Merged openstack/kolla-mesos: tox: delete all pyc before running test https://review.openstack.org/294083 | 07:11 |
*** salv-orlando has joined #kolla | 07:14 | |
openstackgerrit | Merged openstack/kolla: Install Elasticsearch and Kibana packages using yum/apt-get https://review.openstack.org/291986 | 07:15 |
*** mikelk has joined #kolla | 07:16 | |
openstackgerrit | Merged openstack/kolla-mesos: Fix order of assertEqual https://review.openstack.org/294100 | 07:16 |
*** salv-orl_ has joined #kolla | 07:16 | |
*** Marga_ has quit IRC | 07:18 | |
*** salv-orlando has quit IRC | 07:19 | |
*** unicell1 has quit IRC | 07:25 | |
*** Marga_ has joined #kolla | 07:32 | |
*** Marga_ has quit IRC | 07:35 | |
*** Marga_ has joined #kolla | 07:35 | |
*** Marga_ has quit IRC | 07:36 | |
*** Marga_ has joined #kolla | 07:36 | |
*** achanda has joined #kolla | 07:44 | |
*** apuimedo has joined #kolla | 07:45 | |
*** coolsvap has quit IRC | 07:46 | |
*** achanda has quit IRC | 07:48 | |
ajafo | o/ | 07:55 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 07:57 |
*** coolsvap has joined #kolla | 07:57 | |
*** stvnoyes has quit IRC | 08:02 | |
*** stvnoyes has joined #kolla | 08:02 | |
*** salv-orl_ has quit IRC | 08:03 | |
*** akwasnie has quit IRC | 08:04 | |
*** jmccarthy has quit IRC | 08:05 | |
*** jmccarthy has joined #kolla | 08:06 | |
*** akwasnie has joined #kolla | 08:08 | |
*** ssurana has quit IRC | 08:15 | |
*** coolsvap has quit IRC | 08:15 | |
*** akwasnie has quit IRC | 08:19 | |
*** coolsvap has joined #kolla | 08:20 | |
*** AndChat|241401 has quit IRC | 08:29 | |
*** ccesario has joined #kolla | 08:30 | |
*** mgoddard has joined #kolla | 08:35 | |
openstackgerrit | Rodolfo Alonso Hernandez proposed openstack/kolla: Fix retrieving api_interface https://review.openstack.org/289241 | 08:40 |
*** asalkeld has quit IRC | 08:41 | |
*** jmccarthy1 has joined #kolla | 09:01 | |
*** mbound has joined #kolla | 09:03 | |
*** coolsvap has quit IRC | 09:04 | |
*** akwasnie has joined #kolla | 09:10 | |
*** SiRiuS_ has joined #kolla | 09:18 | |
*** gfidente has joined #kolla | 09:22 | |
*** salv-orlando has joined #kolla | 09:24 | |
*** AndChat|241401 has joined #kolla | 09:27 | |
*** ccesario has quit IRC | 09:31 | |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 09:38 |
*** dims has joined #kolla | 09:43 | |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment commands https://review.openstack.org/292741 | 09:45 |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment list command https://review.openstack.org/293444 | 09:45 |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement commands list command https://review.openstack.org/293081 | 09:46 |
*** kproskurin has joined #kolla | 09:52 | |
*** mbound has quit IRC | 09:59 | |
*** Marga_ has quit IRC | 10:03 | |
*** kproskurin_ has joined #kolla | 10:07 | |
*** kproskurin has quit IRC | 10:09 | |
*** kproskurin_ is now known as kproskurin | 10:09 | |
*** ccesario has joined #kolla | 10:12 | |
*** pbourke has quit IRC | 10:13 | |
*** pbourke has joined #kolla | 10:13 | |
*** AndChat|241401 has quit IRC | 10:13 | |
*** ccesario has quit IRC | 10:16 | |
*** rstarmer has joined #kolla | 10:23 | |
*** iceyao_ has quit IRC | 10:40 | |
openstackgerrit | Andrei-Lucian Șerb proposed openstack/kolla: Attach external NIC to a NAT-Network if on Wi-Fi https://review.openstack.org/294340 | 10:49 |
*** kproskurin has quit IRC | 10:55 | |
*** mlima has joined #kolla | 11:00 | |
*** akwasnie has quit IRC | 11:00 | |
*** weiyu_ has quit IRC | 11:04 | |
*** rhallisey has joined #kolla | 11:21 | |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 11:28 |
*** weiyu_ has joined #kolla | 11:37 | |
*** jtriley has joined #kolla | 11:44 | |
*** rstarmer has quit IRC | 11:55 | |
*** dwalsh has joined #kolla | 11:57 | |
SiRiuS_ | mandre: I submitted a bug and patch about the issue we discussed the other day (about problems with promiscuous mode on wi-fi) | 11:57 |
SiRiuS_ | mandre: https://bugs.launchpad.net/kolla/+bug/1558766 | 11:57 |
openstack | Launchpad bug 1558766 in kolla "neutron not working inside virtualbox on wi-fi" [Undecided,In progress] - Assigned to Andrei-Lucian Șerb (lucian-serb) | 11:57 |
SiRiuS_ | mandre: https://review.openstack.org/294340 | 11:58 |
*** mbound has joined #kolla | 11:59 | |
*** mlima_ has joined #kolla | 12:01 | |
*** mlima has quit IRC | 12:04 | |
*** dave-mccowan has joined #kolla | 12:05 | |
*** weiyu_ has quit IRC | 12:13 | |
*** weiyu_ has joined #kolla | 12:14 | |
*** mbound has quit IRC | 12:16 | |
*** weiyu_ has quit IRC | 12:18 | |
SamYaple | morning | 12:21 |
ccesario_ | morning | 12:23 |
ccesario_ | SamYaple, http://paste.openstack.org/show/491084/ ... there was changes on mysql deploy!? | 12:24 |
ccesario_ | I just try deploy now (using master branch) | 12:24 |
SamYaple | ccesario_: there was changes to the deploy, yes, but it still deploys. | 12:25 |
SamYaple | was this an existing cluster? | 12:25 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 12:26 |
ccesario_ | no, I did clean up all | 12:27 |
SamYaple | ccesario_: it looks like you may have not clusted up all volumes properly | 12:27 |
ccesario_ | SamYaple, and change my inventory file like your.... | 12:27 |
SamYaple | i would suggest to try cleanup again and double check all voluemes are scrubbed | 12:27 |
SamYaple | may not have cleaned up all volumes* | 12:28 |
ccesario_ | http://paste.openstack.org/show/491085/ | 12:28 |
ccesario_ | ok, I 'll do it | 12:30 |
sbezverk | Good morning | 12:31 |
sbezverk | SamYaple do you have by any chance a contact in swift ? | 12:32 |
SamYaple | sbezverk: i know some people yes, but you can always ask in thier channel yourself | 12:32 |
SamYaple | everyones friendly | 12:32 |
SamYaple | whatdo you heed? | 12:32 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 12:33 |
sbezverk | SamYaple, with very default swift installation I see an authorization issue, by some reason keystone does not see authentication token in requests from swift | 12:34 |
sbezverk | I checked swift config and it looks good to me, but I need validate it by someone intimatly familiar with swift :-) | 12:35 |
openstackgerrit | Michal Rostecki proposed openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 12:36 |
*** diogogmt has quit IRC | 12:37 | |
*** Marga_ has joined #kolla | 12:37 | |
ccesario_ | SamYaple, cleanup done | 12:37 |
ccesario_ | prechecks is ok | 12:37 |
SamYaple | sbezverk: hmmm maybe check with pbourke first | 12:39 |
sbezverk | SamYaple will do, thanks | 12:40 |
*** diogogmt has joined #kolla | 12:42 | |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 12:44 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 12:46 |
ccesario_ | SamYaple, the same error http://sprunge.us/fBLc?log | 12:46 |
*** openstackgerrit has quit IRC | 12:48 | |
*** openstackgerrit has joined #kolla | 12:48 | |
*** diogogmt has quit IRC | 12:59 | |
openstackgerrit | Merged openstack/kolla-mesos: Connect to mariadb with dnsname. https://review.openstack.org/290176 | 13:02 |
mandre | SiRiuS_: congrats on your first patch | 13:03 |
mandre | do not freak out if it doesn't get reviewed right away, it usually takes a few days | 13:03 |
mandre | it will get reviewed eventually | 13:03 |
*** salv-orlando has quit IRC | 13:06 | |
sbezverk | pbourke ping | 13:06 |
*** jtriley has quit IRC | 13:11 | |
SiRiuS_ | mandre: thanks, i'll wait patiently :) | 13:11 |
ccesario_ | SamYaple, some idea? | 13:12 |
sbezverk | SamYaple I have a possible suspect for swift issue.. can we chat about it or you are busy? | 13:15 |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment commands https://review.openstack.org/292741 | 13:19 |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement deployment list command https://review.openstack.org/293444 | 13:20 |
openstackgerrit | Andrey Pavlov proposed openstack/kolla-mesos: CLI: implement commands list command https://review.openstack.org/293081 | 13:20 |
openstackgerrit | Merged openstack/kolla-mesos: CLI: implement deployment commands https://review.openstack.org/292741 | 13:30 |
openstackgerrit | Merged openstack/kolla-mesos: CLI: implement deployment list command https://review.openstack.org/293444 | 13:30 |
openstackgerrit | Merged openstack/kolla-mesos: CLI: implement commands list command https://review.openstack.org/293081 | 13:31 |
*** mbound has joined #kolla | 13:31 | |
*** mbound has quit IRC | 13:31 | |
*** jtriley has joined #kolla | 13:33 | |
*** mbound has joined #kolla | 13:33 | |
ccesario_ | SamYaple, the problem is related to Mysql Cluster, when I put multiple hosts in inventory file ([control] section) the deploy fails... But with only 1 host... the deploy it works | 13:38 |
SamYaple | ccesario_: i deploy only multinode | 13:42 |
SamYaple | but ill recheck | 13:42 |
SamYaple | just got into work | 13:42 |
SamYaple | sbezverk: feel free to talk about it | 13:43 |
sbezverk | SamYaple when you switch to keystonev3, do you know if swift deployment was updated to make sure that v3 api is used when creating domains/projects etc?? | 13:45 |
SamYaple | sbezverk: no idea | 13:45 |
SamYaple | thats possibly it though | 13:45 |
ccesario_ | SamYaple, yes yes I know, therefore I asked you about changes..... | 13:46 |
sbezverk | SamYaple I found a couple of reference saying that with keystonev3 for swift to work, it requires some modifications | 13:46 |
*** iceyao has joined #kolla | 13:46 | |
SamYaple | sbezverk: feel free to submit them in a review | 13:46 |
SamYaple | we can go over them there! | 13:46 |
sbezverk | SamYaple will do as soon as make it work locally :-) | 13:47 |
SamYaple | awesome | 13:47 |
SamYaple | sounds great | 13:47 |
SamYaple | mlima_: i responded to you instance_password concern | 13:48 |
SamYaple | perhaps you should clarify and give a recommendation | 13:49 |
SamYaple | ill default to your judgement here | 13:49 |
SamYaple | ccesario_: i think i may see the bug | 13:50 |
SamYaple | im testing now | 13:50 |
ccesario_ | SamYaple, thanks ;) | 13:50 |
*** inc0 has joined #kolla | 13:52 | |
inc0 | good morning | 13:52 |
SamYaple | inc0: morning! | 13:52 |
SamYaple | inc0: im pushing the fix for the comment on rabbitmq right now just fyi | 13:53 |
inc0 | you're excited today Sam....anything special?;) | 13:53 |
SamYaple | so we dont conflict | 13:53 |
SamYaple | inc0: yea i did cool stuff! | 13:53 |
SamYaple | my kolla deploy at home is kickin! | 13:53 |
SamYaple | i changed passwords, nuked a node, all while running | 13:53 |
SamYaple | instance was never affectede | 13:53 |
inc0 | that's cool | 13:54 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Test https://review.openstack.org/292869 | 13:54 |
inc0 | SamYaple, I'm about to write this hostname check, we still want it right? | 13:54 |
SamYaple | inc0: yes. let me push up this rabbitmq stuff so you have a good base | 13:54 |
inc0 | cool, let's do that | 13:55 |
inc0 | is there better way to do it than 2 tasks? | 13:56 |
inc0 | in which 1 will be command "host <<rabbit hostname" register rabbit ip | 13:56 |
inc0 | ahh | 13:56 |
inc0 | then it registers and failed_when ansible_hostname != ip | 13:57 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Playbook for rabbitmq upgrade https://review.openstack.org/287969 | 13:57 |
inc0 | or register and use module "fail" so we can customize message | 13:58 |
openstackgerrit | Artur Zarzycki proposed openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 13:58 |
SamYaple | inc0: i would probably use the fail message because this issue is tricky | 13:59 |
SamYaple | but as long as the failure is clear i dont see any problem with either way | 13:59 |
*** mbound has quit IRC | 13:59 | |
SamYaple | W00T!!!! | 13:59 |
inc0 | I'll use fail with message | 13:59 |
SamYaple | https://review.openstack.org/#/c/292869/ | 13:59 |
SamYaple | look! | 13:59 |
inc0 | since it's reconfig from user side so I want be super clear about that | 13:59 |
inc0 | nice, you don't break infra any more | 14:00 |
SamYaple | yay! | 14:00 |
SamYaple | thats great so now we can rapid fire docs patches | 14:00 |
SamYaple | no delay, 5m for gate is all | 14:00 |
mlima_ | SamYaple, you can add that variable here https://review.openstack.org/#/c/293797/1/ansible/roles/manila/defaults/main.yml | 14:03 |
*** mgoddard_ has joined #kolla | 14:04 | |
SamYaple | ok mlima_ i trust you on that. | 14:04 |
SamYaple | after i make the change, can you comment that way in the review as well? | 14:04 |
mlima_ | yes, i can | 14:04 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Add generate_passwords.py to generate passwords https://review.openstack.org/293728 | 14:05 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Move manila variable https://review.openstack.org/293797 | 14:05 |
SamYaple | mlima_: https://review.openstack.org/293797 | 14:05 |
mlima_ | just to clarify, look in manila.conf and you will see that this variables are in [generic] session | 14:05 |
SamYaple | yes i see it, i just wanted to make sure we dont control that password | 14:06 |
SamYaple | if we dont then its a known password | 14:06 |
*** mgoddard has quit IRC | 14:06 | |
*** mbound has joined #kolla | 14:06 | |
mlima_ | yes | 14:07 |
*** spisarski has joined #kolla | 14:07 | |
SamYaple | ccesario_: did yo urebuild your mariadb container? | 14:08 |
mlima_ | for others drivers define this variables isn't necessary SamYaple | 14:09 |
ccesario_ | SamYaple, yes | 14:10 |
ccesario_ | operator.local:4000/kolla/ubuntu-source-mariadb 2.0.0 09962241a2cd 2 hours ago 442.9 MB | 14:10 |
*** fthiagogv has quit IRC | 14:10 | |
SamYaple | ccesario_: did you pull that new image down to all your multinode? | 14:10 |
openstackgerrit | Sam Yaple proposed openstack/kolla: Playbook for rabbitmq upgrade https://review.openstack.org/287969 | 14:11 |
ccesario_ | yes .... I removed all images, configs, from all nodes | 14:11 |
SamYaple | ok im setting up multinode new vms to test this as we speak | 14:12 |
ccesario_ | SamYaple, ok, I'll try get some more infos in mys tests... | 14:13 |
ccesario_ | SamYaple, I 'm using Kolla Version: 2.0.0.0rc2.dev22 | 14:14 |
*** diogogmt has joined #kolla | 14:17 | |
SamYaple | so ntpttr i dug back through my logs. youll never believe it but neutron is _not_ allocating two ips | 14:18 |
SamYaple | ntpttr: i simple created and deleted those instances leaving me with a two ip spacing | 14:18 |
*** akwasnie has joined #kolla | 14:18 | |
SamYaple | it was just coincednce | 14:18 |
openstackgerrit | Merged openstack/kolla-mesos: Switch services to use mesos-dns names https://review.openstack.org/290181 | 14:19 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Only copy the needed config file https://review.openstack.org/294618 | 14:32 |
*** vinkman has joined #kolla | 14:32 | |
*** akwasnie has quit IRC | 14:33 | |
*** achanda has joined #kolla | 14:34 | |
*** Kennan has quit IRC | 14:36 | |
*** Kennan has joined #kolla | 14:37 | |
*** vinkman has left #kolla | 14:38 | |
*** haplo37 has joined #kolla | 14:39 | |
*** achanda has quit IRC | 14:47 | |
openstackgerrit | Sam Yaple proposed openstack/kolla: Enabled nova->glance snapshots https://review.openstack.org/294638 | 15:03 |
*** salv-orlando has joined #kolla | 15:10 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Only copy the needed config file https://review.openstack.org/294618 | 15:16 |
openstackgerrit | MD NADEEM proposed openstack/kolla: Fix inconsistancy in globals.yml https://review.openstack.org/294392 | 15:17 |
ntpttr | SamYaple: Oh gotcha, well that's sort of a relief - it was definitely just the split brain issue, which is all fixed now | 15:18 |
inc0 | ntpttr, rabbit split-brained? | 15:18 |
ntpttr | inc0: yes, my hosts file wasn't complete, so multiple rabbit instances thought they were in charge | 15:19 |
inc0 | I'm writing code that will break ansible deploy if hostnames aren't resolvable | 15:19 |
ntpttr | cool thanks, that and rabbit using IP addresses rather than hostnames would fix the issue | 15:20 |
inc0 | ntpttr, rabbitmq doesn't work with IPs :( | 15:21 |
inc0 | at all | 15:21 |
*** blahRus has joined #kolla | 15:22 | |
ntpttr | oh I see, maybe I misunderstood, I thought SamYaple was working on a patch for that | 15:22 |
inc0 | ntpttr, yeah, and he discovered that its impossibl | 15:23 |
*** vhosakot has joined #kolla | 15:23 | |
ntpttr | hah all right, well the precheck thing will be good enough then | 15:23 |
*** salv-orlando has quit IRC | 15:24 | |
inc0 | famous Sam's quotations vol.1 "Erlang sux" | 15:24 |
*** Marga_ has quit IRC | 15:27 | |
inc0 | guys, my docker networking breaks, maybe some of you experienced it | 15:30 |
inc0 | it's docker networking, so builds are broken | 15:30 |
inc0 | network on host work like charm, inside docker id doesn't | 15:30 |
*** dwalsh has quit IRC | 15:33 | |
SamYaple | ntpttr: i spent a few hours in the depth of erland, TL;DR multnode erlang with ips wont work | 15:34 |
SamYaple | inc0: regen rules for iptables | 15:35 |
SamYaple | stop docker service | 15:35 |
SamYaple | flush rules | 15:35 |
SamYaple | start docker service | 15:35 |
ccesario_ | SamYaple, got success in your deploy!? | 15:36 |
inc0 | SamYaple, iptables -F didnt help :( | 15:38 |
SamYaple | inc0: you need more than that | 15:38 |
*** mdnadeem_home has joined #kolla | 15:38 | |
SamYaple | inc0: its probably your nat stuff | 15:38 |
SamYaple | inc0: http://www.cyberciti.biz/tips/linux-iptables-how-to-flush-all-rules.html something like in that script | 15:39 |
SamYaple | ccesario_: not yet, got sidetracked with other work | 15:39 |
SamYaple | inc0: iptables -t nat -F (otherwise you are flushing nat rules) | 15:39 |
ccesario_ | SamYaple, np | 15:40 |
ccesario_ | it seems really that the problem is related to mysql cluster (in my tests, of course :P ) | 15:41 |
SamYaple | ccesario_: ubuntu or centos? | 15:41 |
ccesario_ | ubuntu | 15:41 |
sbezverk | anybody could point me to the location where keystone service catalog as well as domain/project get created by deployment script? | 15:41 |
SamYaple | sbezverk: docker/keystone/keystone_bootstrap.sh | 15:42 |
SamYaple | ansible/roles/keystone/tasks/bootstrap | 15:42 |
sbezverk | SamYaple thank you.. | 15:45 |
*** dwalsh has joined #kolla | 15:47 | |
*** CheKoLyN has joined #kolla | 15:50 | |
*** mbound has quit IRC | 16:01 | |
*** jtriley has quit IRC | 16:04 | |
*** jtriley has joined #kolla | 16:05 | |
mlima_ | hello guys, failed: [storage] => {"changed": true, "failed": true} | 16:07 |
mlima_ | msg: APIError(HTTPError(u'500 Server Error: Internal Server Error for url: http+docker://localunixsocket/v1.22/images/create?tag=2.0.0&fromImage=controller%3A4000%2Fkollaglue%2Fubuntu-source-heka',),) | 16:07 |
mlima_ | I have this error ^ | 16:07 |
mlima_ | in my multinode deployment ^ | 16:07 |
*** mkoderer___ is now known as mkoderer__ | 16:08 | |
*** dwalsh has quit IRC | 16:10 | |
*** mbound has joined #kolla | 16:14 | |
*** iceyao has quit IRC | 16:15 | |
vhosakot | mlima_: which task ? | 16:15 |
mlima_ | wait | 16:16 |
mlima_ | I lost the log | 16:16 |
*** mgoddard__ has joined #kolla | 16:17 | |
*** mgoddard_ has quit IRC | 16:18 | |
*** ssurana has joined #kolla | 16:20 | |
mlima_ | TASK: [common | Starting heka container] vhosakot | 16:20 |
vhosakot | mlima_: passes for me in this morning's master | 16:21 |
*** sdake has joined #kolla | 16:24 | |
*** sdake_ has joined #kolla | 16:24 | |
mlima_ | vhosakot my inventory multinode http://pastebin.com/AFcLbNvV | 16:27 |
vhosakot | mlima_: let me look into it.. thanks | 16:28 |
*** sdake has quit IRC | 16:28 | |
sdake_ | morning | 16:29 |
*** sdake_ is now known as sdake | 16:29 | |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Start the haproxy by using haproxy-system-wrapper https://review.openstack.org/294701 | 16:29 |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Fixes incorrect swift service endpoints https://review.openstack.org/294702 | 16:30 |
*** haplo37 has quit IRC | 16:31 | |
sbezverk | SamYaple this is a part of swift problem https://review.openstack.org/294702 | 16:31 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Start the haproxy by using haproxy-system-wrapper https://review.openstack.org/294701 | 16:32 |
*** mdnadeem_home has quit IRC | 16:33 | |
*** dims_ has joined #kolla | 16:33 | |
*** rstarmer has joined #kolla | 16:34 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Fixes incorrect swift service endpoints https://review.openstack.org/294702 | 16:34 |
*** dims has quit IRC | 16:34 | |
*** mdnadeem_home has joined #kolla | 16:36 | |
vhosakot | mlima_: your multinode inventory looks correct... can you check the volumes mounted by the heka container ? | 16:40 |
vhosakot | mlima_: can you paste the output of | 16:40 |
vhosakot | docker volume ls | grep 'kolla_logs\|heka' | 16:40 |
*** salv-orlando has joined #kolla | 16:41 | |
*** mikelk has quit IRC | 16:44 | |
*** jmccarthy1 has quit IRC | 16:44 | |
openstackgerrit | Serguei Bezverkhi proposed openstack/kolla: Fixes incorrect swift service endpoints https://review.openstack.org/294702 | 16:45 |
*** kproskurin has joined #kolla | 16:49 | |
mlima_ | root@controller:~/kolla# docker volume ls | grep 'kolla_logs\|heka' | 16:50 |
mlima_ | local heka_socket | 16:51 |
mlima_ | local heka | 16:51 |
mlima_ | local kolla_logs | 16:51 |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Add check if all the hostnames are resolvable https://review.openstack.org/294713 | 16:51 |
mlima_ | vhosakot ^ | 16:51 |
*** salv-orlando has quit IRC | 16:51 | |
*** kproskurin has quit IRC | 16:51 | |
*** salv-orlando has joined #kolla | 16:53 | |
*** Marga_ has joined #kolla | 16:55 | |
mlima_ | vhosakot i ran docker ps in storage and compute node and i have not any output | 16:56 |
vhosakot | ah, you have the docker volumes... can you rerun "kolla-ansible deploy" | 16:56 |
mlima_ | yes | 16:57 |
vhosakot | mlima_: I'd double check the mountpoints | 16:58 |
vhosakot | docker volume inspect heka heka_socket kolla_logs | 16:58 |
mlima_ | same error vhosakot | 17:00 |
vhosakot | mlima_: ah, not good.. what error ? | 17:00 |
mlima_ | the same | 17:00 |
mlima_ | msg: APIError(HTTPError(u'500 Server Error: Internal Server Error for url: http+docker://localunixsocket/v1.22/images/create?tag=2.0.0&fromImage=controller%3A4000%2Fkollaglue%2Fubuntu-source-heka',),)] | 17:00 |
vhosakot | kolla-ansible deploy -t heka -vvvv | 17:01 |
elemoine_ | mlima_: do you have more errors in the Docker logs? | 17:01 |
elemoine_ | vhosakot: there's no "heka" tag | 17:01 |
mlima_ | how can i check it elemoine_? | 17:02 |
mlima_ | btw, i'm trying multinode deployment | 17:02 |
*** achanda has joined #kolla | 17:03 | |
vhosakot | elemoine_: ah ok... | 17:03 |
mlima_ | vhosakot http://pastebin.com/duKztfyw | 17:05 |
mlima_ | :) | 17:05 |
mlima_ | docker volume inspect heka heka_socket kolla_logs ^ | 17:05 |
*** mlima_ is now known as mlima | 17:05 | |
vhosakot | mlima_: so, your heka volume mountpoints "/var/lib/docker/volumes/*/_data" look correct .. 2 sec | 17:06 |
mlima | guys, i have three nodes storage,compute and controller | 17:08 |
mlima | i installed docker in all nodes, just it | 17:08 |
vhosakot | elemoine_: where can I find a container's tag ? is this the right place? https://github.com/openstack/kolla/blob/d55be10ddd429c3c3e5ded2f95ad8df5edb701bc/ansible/roles/horizon/defaults/main.yml#L8 | 17:09 |
mlima | all configuration kolla, ansible were did in controller node | 17:09 |
*** rstarmer has quit IRC | 17:09 | |
SamYaple | Jeffrey4l: that haproxy change, we should not be doing it so late in the cycle. | 17:10 |
SamYaple | Jeffrey4l: in newton the ubuntu stuff will be systemd too, so that should be a fine change for netwon | 17:10 |
mlima | this tag https://github.com/openstack/kolla/blob/08101fd4654ffbf7ca6331427a5d895b1db0f008/ansible/group_vars/all.yml#L146 vhosakot ? | 17:10 |
Jeffrey4l | SamYaple, it is not related to the systemd system. Just a binary change. | 17:10 |
Jeffrey4l | it works on the ubuntu 14.04. | 17:11 |
SamYaple | im worried about making such a change this late in teh cycle though | 17:11 |
Jeffrey4l | I checked. ubuntu 14.04 has the haproxy-systemd-wrapper binary in the current package SamYaple | 17:11 |
inc0 | SamYaple, I'm thinking...do we really need rabbitmq address same as in api interface? | 17:12 |
vhosakot | mlima: that is the release tag.... I was asking where are the tags passed to the -t option for kolla-ansible set ? | 17:12 |
inc0 | it will be hard to check for other than local node | 17:12 |
Jeffrey4l | SamYaple, btw, the current implementaion has another issue. `docker stop haproxy` can not graceful. the haproxy container will be kill after 10 seconds. | 17:12 |
*** mgoddard__ has quit IRC | 17:12 | |
Jeffrey4l | Is this enough for this change? | 17:12 |
*** mdnadeem_home has quit IRC | 17:13 | |
inc0 | and not really important unless someone messed up big time | 17:13 |
SamYaple | inc0: yes we do, its doable (this is waht i was talking about yesterday) | 17:13 |
SamYaple | no its actually a big issue inc0 | 17:13 |
openstackgerrit | Hui Kang proposed openstack/kolla: Ceilometer needs mongoengine if build from source https://review.openstack.org/294725 | 17:13 |
SamYaple | because we bind to api_interface | 17:13 |
inc0 | ah ok | 17:13 |
SamYaple | if hostnaem doesnt resolve that way the it wont work | 17:13 |
vhosakot | SamYaple: elemoine_: where are the tags passed to the -t option for kolla-ansible set ? | 17:13 |
SamYaple | Jeffrey4l: hmmm let me dig into it a bit more. it worries me this type of change so late | 17:13 |
Jeffrey4l | SamYaple, OK. | 17:14 |
SamYaple | Jeffrey4l: but we can graceful stop, we just need changes to kolla_docker, which was planned for newton (this was discussed, not this wrapper specifically though) | 17:14 |
elemoine_ | vhosakot: Heka is part of the "common" playbook, which does not have a tag | 17:14 |
elemoine_ | it's just executed as a dependency when another playbook is executed | 17:14 |
inc0 | so how would we do it? | 17:14 |
inc0 | what's the proper way? | 17:15 |
vhosakot | elemoine_: ah I see.... and for non-common playbook, where/how is the tag set ? | 17:15 |
elemoine_ | I usually do kolla-ansible -t haproxy deploy | 17:15 |
elemoine_ | tags are defined in site.yml IIRC | 17:15 |
inc0 | (also fuck, more string processing in ansible) | 17:15 |
vhosakot | elemoine_: agreed, "kolla-ansible -t haproxy deploy" triggers/runs heka playbooks | 17:15 |
sdake | inc0 no | 17:15 |
inc0 | sdake, ...? | 17:15 |
sdake | reading scrollback | 17:16 |
sdake | re api binding | 17:16 |
sdake | shoud lbind to the management network | 17:16 |
SamYaple | sdake: no it shouldnt | 17:16 |
SamYaple | we bind nothing to management network (thats outside of kolla if it exists) | 17:16 |
Jeffrey4l | SamYaple, graceful should not be possible for the current implemention. The root process(after run ensure_haproxy_latest_config) can not handle the SIGTERM signal( I suppose it just ingore all the signal, because a new haproxy will run ). | 17:16 |
inc0 | well api for rabbitmq is mgmt normally | 17:16 |
SamYaple | its bind to api_interface | 17:16 |
vhosakot | elemoine_: thanks for the info | 17:17 |
SamYaple | Jeffrey4l: with kolla_docker changes we can send any signal, including USR2 | 17:17 |
SamYaple | Jeffrey4l: but that was planned for newton since that just came about a few weeks ago | 17:17 |
Jeffrey4l | USR2 single is used for haproxy-systemd-wrapper, not haproxy. | 17:18 |
sdake | which was planned for newton | 17:18 |
sdake | if its not in the bug tracker as a blueprint or bug its not mitaka material | 17:18 |
Jeffrey4l | sdake, https://review.openstack.org/294701 see this. | 17:18 |
sdake | Jeffrey4l personally i'd rather punt on lows that don't actually have a reproducer ;) | 17:18 |
sdake | Jeffrey4l let me lookie at your patch tho | 17:19 |
Jeffrey4l | sdake, roger. | 17:19 |
mlima | hello guys, someone help me to deploy kolla multinode | 17:19 |
ccesario_ | SamYaple, all tests report me erro when two or more hosts are defines in compute section | 17:19 |
SamYaple | ccesario_: prechecks? | 17:19 |
ccesario_ | its ok :D | 17:19 |
sdake | Jeffrey4l given that its some kind of wrapper loading haproxy which is totally unknown to me, and igven the bug is low priority, I am not keen on the solution | 17:20 |
ccesario_ | eheheh | 17:20 |
mlima | i have this error | 17:20 |
mlima | msg: APIError(HTTPError(u'500 Server Error: Internal Server Error for url: http+docker://localunixsocket/v1.22/images/create?tag=2.0.0&fromImage=controller%3A4000%2Fkollaglue%2Fubuntu-source-heka',),) | 17:20 |
mlima | storage node and compute node | 17:20 |
sdake | Jeffrey4l can't we just wait until newton and backport if mandatory? | 17:20 |
SamYaple | Jeffrey4l: i am keen on the solution for newton though, that wrapper looks pretty good (viewing source now) | 17:20 |
sdake | newton is wide open but we aren't in newton yet | 17:20 |
Jeffrey4l | sdake, I will add some comment to the PS to explain more. | 17:20 |
SamYaple | we know haproxy works now, i just dont want to introduce bugs | 17:20 |
sdake | i agree with SamYaple here - rather not rock apple cart | 17:21 |
sdake | its only 2 weeks to wait | 17:21 |
sdake | not like saying no to patch entirely ;) | 17:22 |
SamYaple | we do want to consider any potentially hard-to-change things now though | 17:22 |
SamYaple | like the volume names | 17:22 |
sdake | anything in the bug tracker is fair game | 17:22 |
vhosakot | elemoine_: I found "tags:" in ansible/site.yml.. cool.. so, all common containers (heka, kolla-toolbox and common) do not have tags ?, right ? | 17:22 |
SamYaple | we need to make sure those are spot on (i already adjusted the libvirt volume name) | 17:22 |
sdake | this is in the bug tracker | 17:22 |
sdake | SamYaple agree re volume names | 17:22 |
Jeffrey4l | SamYaple, sdake current implementation has a bug. after run `ensure_haproxy_latest_config` binary, the haproxy can not be stopped gracefully. the docker daemon kill the haproxy container after 10 seconds. | 17:22 |
sdake | i think they are all good to go from a single server persepctive | 17:22 |
SamYaple | i think all volume names are good, but they should get a twice over | 17:22 |
SamYaple | Jeffrey4l: thats not the behavoiur i see | 17:23 |
sdake | Jeffrey4l ok and the impact of that is what? | 17:23 |
SamYaple | Jeffrey4l: why would that be the case? | 17:23 |
SamYaple | how can i reproduce what youre asking | 17:23 |
sdake | Jeffrey4l impact is upgrade and reconfigure are slowed down? | 17:24 |
Jeffrey4l | OK. I will file a bug to describe this. wait for a moment. | 17:24 |
Jeffrey4l | sdake, yup. | 17:24 |
sdake | ya i think we can live with that for the short term | 17:24 |
sdake | as long as its not BROKEN | 17:24 |
sdake | but slower, thats probably ok | 17:24 |
*** achanda has quit IRC | 17:25 | |
sdake | i'm not keen to bikeshed on people's patch sets tho at this time - we only have 2 weeks to go | 17:25 |
sdake | but why not work on a high or critical prioirty bug ;) | 17:25 |
sdake | for example heat stacks cannot be created | 17:25 |
SamYaple | or even config change means haproxy can restart under some conditions, all the other contaienrs _always_ restart, so that should be fine as a bug | 17:25 |
Jeffrey4l | kill -9 process may be not a good idea. even though I can not find out the potential issue. | 17:26 |
Jeffrey4l | I just want to throw the exist issue and the solution. | 17:27 |
SamYaple | "not a good idea" == optimization for newton | 17:27 |
sdake | Jeffrey4l yup i hear that, your sayign docker daemon kill -9 the haproxy process and that could cuase a problem during reconfig | 17:27 |
Jeffrey4l | I am ok to fix this in Newton. | 17:27 |
SamYaple | broken is a "fix in mitaka-rc2" issue | 17:27 |
SamYaple | but i dont think its broken | 17:27 |
sdake | its suboptimal but we aren't optimizing | 17:27 |
sdake | Jeffrey4l if you come back and say kill -9 of haproxy disturbs reconfig in a non-recoverable way | 17:28 |
sdake | thats a different discussion | 17:28 |
Jeffrey4l | sdake, I do not know whether it will case any issue to kill -9 a haproxy process. But kill -9 any process, especially when there is a better way, is not a good idea. ( I am not mean we must fix this in the Mitaka cycle) | 17:30 |
*** unicell has joined #kolla | 17:30 | |
sdake | Jeffrey4l cool understood | 17:31 |
sdake | i think killing -9 keepalived could cause proboelsm | 17:32 |
sdake | fwiw Jeffrey4l I ran reconfig in a loop a week or so ago for 500 times and it didn't fail | 17:32 |
sdake | Took abut 8 minutes on my gear | 17:33 |
sdake | so easy to replicate in your environment | 17:33 |
SamYaple | kill -9 keepalived will leave vips around (resulting in potentially two vips on your hosts | 17:34 |
SamYaple | this is a fairly common support issue | 17:34 |
SamYaple | in general | 17:34 |
sdake | we should probably fix that somehow in the future via a task | 17:34 |
*** CheKoLyN has quit IRC | 17:34 | |
Jeffrey4l | sdake, it will never fail. the docker will kill(kill -9) the container process after 10 seconds if can not stop it( kill -SIGTERM ). | 17:34 |
sdake | Jeffrey4l yup very familiar with libcontainer | 17:34 |
sdake | i fixe dpid=host when it was busted ;) | 17:35 |
Jeffrey4l | :) | 17:35 |
sdake | or helped fix i suppose | 17:35 |
ccesario_ | SamYaple, http://sprunge.us/fMQB prechecks | 17:41 |
*** mbound has quit IRC | 17:45 | |
inc0 | damn...I probably just wrote the longest when statement in kolla | 17:53 |
SamYaple | git review | 17:54 |
Jeffrey4l | SamYaple, https://bugs.launchpad.net/kolla/+bug/1559238 reproduce step is here. | 17:54 |
openstack | Launchpad bug 1559238 in kolla "haproxy container can not be stopped/restarted gracefully" [Undecided,New] | 17:54 |
SamYaple | yea thats what i thought you were saying Jeffrey4l | 17:54 |
SamYaple | agree it could be fixed | 17:54 |
SamYaple | but its nto critical | 17:55 |
Jeffrey4l | yep. I am OK that it can be fixed until n cycle. | 17:55 |
sdake | Jeffrey4l this needs attention if you have spare cycles https://bugs.launchpad.net/kolla/+bug/1551992 | 17:59 |
openstack | Launchpad bug 1551992 in kolla "Make Magnum use Keystone v3" [Critical,Confirmed] | 17:59 |
sdake | it is unassigned atm | 17:59 |
openstackgerrit | Michal Jastrzebski (inc0) proposed openstack/kolla: Add check if all the hostnames are resolvable https://review.openstack.org/294713 | 17:59 |
sdake | folks if your looking for work, pleae work on critical confirmed bugs | 17:59 |
Jeffrey4l | sdake, ok. I will try to fix it. | 17:59 |
inc0 | this when is long and ugly but I have no idea how to make it better | 18:00 |
sdake | Jeffrey4l focus on the keystone v3 problem first, i'm ssure other problems will be uncovered | 18:00 |
sdake | file seapate bugs for those | 18:00 |
Jeffrey4l | sdake, ok, np | 18:00 |
openstackgerrit | Jeffrey Zhang proposed openstack/kolla: Start the haproxy by using haproxy-system-wrapper https://review.openstack.org/294701 | 18:03 |
*** JoseMello has joined #kolla | 18:08 | |
*** sdake has quit IRC | 18:09 | |
ccesario_ | SamYaple, did yoou see? no errors in precheck :) | 18:09 |
SamYaple | ccesario_: w00t | 18:10 |
SamYaple | always good to see no error | 18:10 |
*** rstarmer has joined #kolla | 18:12 | |
*** sdake has joined #kolla | 18:13 | |
SamYaple | who wants external journals for ceph? who wants "dont use my whole disk" for ceph? well just you wait and see! | 18:15 |
SamYaple | i have a suprise for oyu | 18:15 |
sdake | i want to be able to use an already deployed ceph SamYaple | 18:17 |
sdake | had alot of private feedback on that requirement | 18:17 |
sdake | and i htink you were in agreement at one point on that tpic as well that people want to manage their own ceph | 18:17 |
SamYaple | feel free to implement it | 18:17 |
*** dwalsh has joined #kolla | 18:19 | |
ccesario_ | SamYaple, but deploy show errors related to mysql cluster :( | 18:20 |
sdake | core reviewers - please feel free to -2 changes that are enhancements or optimizations that don't fix actual defects until Newton is open for business April 2nd. | 18:25 |
Jeffrey4l | so close :D | 18:28 |
sdake | pbourke you happen to be around | 18:34 |
ccesario_ | SamYaple, http://sprunge.us/QfiH?log | 18:35 |
*** mbound has joined #kolla | 18:38 | |
*** mbound has quit IRC | 18:39 | |
*** mbound has joined #kolla | 18:39 | |
sdake | SamYaple thoughts on https://bugs.launchpad.net/kolla/+bug/1558853 | 18:39 |
openstack | Launchpad bug 1558853 in kolla "Bootstrapping Ceph OSDs fails" [Critical,Triaged] | 18:39 |
Jeffrey4l | sdake, fyi, https://bugs.launchpad.net/kolla/+bug/1551992 this is not a bug of kolla. need a new release of python-magnumclient | 18:40 |
openstack | Launchpad bug 1551992 in kolla "Make Magnum use Keystone v3" [Critical,Invalid] - Assigned to Jeffrey Zhang (jeffrey4l) | 18:40 |
sdake | Jeffrey4l using master python-magnumclient fixes the problem? | 18:40 |
Jeffrey4l | sdake, yup. | 18:40 |
sdake | Jeffrey4l does magnum work otherwise? | 18:41 |
sdake | as in does it work with master client | 18:42 |
*** rstarmer has quit IRC | 18:42 | |
*** inc0 has quit IRC | 18:42 | |
sdake | Jeffrey4l i'd like to see if you can create bays and such | 18:43 |
Jeffrey4l | sdake, at least, when using the master code, the bug is disappeared( by using command line ). | 18:43 |
Jeffrey4l | sdake, I am not famillar with magnum ( never use it :( ). but I will try to create some in it. | 18:44 |
*** jtriley has quit IRC | 18:49 | |
Jeffrey4l | sdake, I can create a baymodel. But when creating bay, it raise errors "Failed to create trustee 85d1e386-51c6-4694-b702-19672fe3f129 in domain None" | 18:56 |
Jeffrey4l | Not sure what's the root cause. Seems related to keystone v3. | 18:56 |
*** Serlex has joined #kolla | 18:56 | |
Jeffrey4l | I will check this tomorrow. | 18:56 |
Jeffrey4l | need go to bed. | 18:56 |
Jeffrey4l | bye guys. | 18:57 |
openstackgerrit | Steven Dake proposed openstack/kolla: Explain how to use kolla-ansible pull https://review.openstack.org/284991 | 18:57 |
sdake | thanks Jeffrey4l | 18:57 |
sdake | that is the keystone v3 support i was tlaking about in tahtt bug | 18:57 |
sdake | i am going t o unmark it invalid | 18:58 |
sdake | also the containers team is going to release a new client | 18:58 |
Jeffrey4l | sdake, but that bug's description shuld be change. | 18:58 |
sdake | i'm in it what do you want it to be | 18:58 |
Jeffrey4l | the original description is invalid. | 18:58 |
Jeffrey4l | when creating bay, error raised http://paste.openstack.org/show/491151/ | 18:59 |
sdake | Jeffrey4l https://bugs.launchpad.net/kolla/+bug/1551992 | 19:00 |
openstack | Launchpad bug 1551992 in kolla "Magnum failed to create trustee" [Critical,Confirmed] - Assigned to Jeffrey Zhang (jeffrey4l) | 19:00 |
Jeffrey4l | maybe I missed some config? anyway, i will check this tomorrow. | 19:00 |
sdake | config is definately missing | 19:00 |
sdake | they went from keystone v2 to keystone v3 and used trusts along the way | 19:00 |
Jeffrey4l | roger that. | 19:01 |
sdake | have nice weekend lei | 19:01 |
Jeffrey4l | sdake, the same to you. | 19:02 |
* Jeffrey4l heading to bed. | 19:02 | |
SamYaple | sdake: looks like a dirty ceph environment | 19:04 |
*** dwalsh has quit IRC | 19:04 | |
sdake | SamYaple thanks would you mind responding in the bug about how to clean that up | 19:06 |
SamYaple | woah woah sdake, the passwrod generation thing was always going in mitaka https://review.openstack.org/#/c/293728/ | 19:06 |
SamYaple | why are you blocking that? | 19:06 |
SamYaple | it was back from liberty | 19:06 |
sdake | same reason as fernet got pushed | 19:07 |
sdake | disurptive feature in rcs | 19:07 |
SamYaple | that is much different as that affect the plabooks. this is not a disruptive feature dude | 19:07 |
SamYaple | this is a tool to populate the passwords yml, some thing that we didnt have before so the user would have had to do by hand | 19:08 |
sdake | yes but the file starts out empty | 19:08 |
sdake | so it becomes mandatory | 19:08 |
SamYaple | as always has been the plan | 19:09 |
sdake | no docs no blueprint not in cmd directory | 19:09 |
sdake | not written down plan | 19:09 |
sdake | point me at where it was written down | 19:09 |
sdake | atleast fernet was writetn down | 19:09 |
sdake | we pushed that | 19:09 |
sdake | features are getting pushed at this point | 19:09 |
SamYaple | we pushed fernet because it doesnt work and isnt done | 19:09 |
sdake | we have 40 bugs to fix | 19:09 |
SamYaple | even if we _dont_ put this tool in, the passwords file should be empty | 19:10 |
SamYaple | thats a huge security thing | 19:10 |
sdake | add to agenda if you like - and make a blueprint for it because its a feature nota trivial fix | 19:10 |
SamYaple | everyone having passwords being "password" | 19:10 |
sdake | i think operators are bright enough to make their own passwords in a file called "passwords.yml" | 19:10 |
SamYaple | fine ill remove the "feature" and blank out the file then | 19:11 |
SamYaple | but they wont and thats the security problem | 19:11 |
sdake | this is easily solvable by the QSG | 19:11 |
SamYaple | default configuration is not secure here | 19:11 |
sdake | just add instructions | 19:11 |
SamYaple | thats going to change for mitaka with or without a helpfultool | 19:11 |
sdake | nothing is secure if the default passwords are not set | 19:12 |
SamYaple | it wont deploy if they arent set.... | 19:12 |
sdake | what i mean is other tools not ours | 19:12 |
sdake | listen, I am open to securying passwords.yml | 19:12 |
sdake | but a feature at lasst minute with trivialfix is not the way to roll | 19:13 |
sdake | you should know that | 19:13 |
SamYaple | im happy to submit a patch that simply blanks the passwords.yml | 19:13 |
SamYaple | that _is_ the trivialfix | 19:13 |
sdake | wfm as long as there is a docimpact in it | 19:13 |
SamYaple | which this bug has | 19:13 |
sdake | then we can debate how much of a pita it is to set passwords | 19:13 |
SamYaple | without this tool, a big one | 19:14 |
SamYaple | but its hardly a feature | 19:14 |
SamYaple | or barely rather | 19:14 |
sdake | well file a bug then | 19:14 |
sdake | and fix the docs PLEASE and I'll remove -2 | 19:14 |
sdake | i want the qsg to work | 19:14 |
sdake | eveyrone does docimpact and then never fixes the documentation | 19:14 |
sdake | also it needs to go in cmd | 19:15 |
sdake | and a wrapper needs to be built for it in bin | 19:15 |
sdake | we don't install the tools directory | 19:15 |
sdake | if you need help with that ping me but its pretty straight forward | 19:16 |
*** Marga_ has quit IRC | 19:16 | |
SamYaple | https://bugs.launchpad.net/kolla/+bug/1559266 | 19:16 |
openstack | Launchpad bug 1559266 in kolla "passwords.yml does not have secure defaults" [Critical,Triaged] - Assigned to Sam Yaple (s8m) | 19:16 |
sdake | its a one liner | 19:16 |
SamYaple | the wrapper isnt a problem, i wasnt planning on doing that because it was going to get installed that way | 19:16 |
SamYaple | that i thought would violate the feature thing | 19:16 |
SamYaple | so in tools/ it was just a dev tool | 19:16 |
SamYaple | but if you want the tool in cmd thats a different story | 19:17 |
SamYaple | remember, tox -egenconfig doesnt exist for pip installed kolla | 19:17 |
sdake | what yu define above is a bug in your link | 19:17 |
sdake | our defualts are insecure | 19:18 |
sdake | i buy that if we need a feaeture to make that a reality then its still a bug :) | 19:18 |
SamYaple | well blanking the passwords solves the bug, adding a tool can be construed as a feature | 19:18 |
SamYaple | maybe | 19:18 |
SamYaple | but since it doesnt run by default and its not a requirement (and is from pre liberty....)its a steach to call it a feature | 19:19 |
sdake | lets go with what feels right | 19:19 |
sdake | expecting operators to figure out the format of passwords.yml doesnt feel right | 19:19 |
*** sdake has quit IRC | 19:20 | |
*** sdake has joined #kolla | 19:20 | |
sdake | [12:19:36] <sdake>lets go with what feels right | 19:20 |
sdake | [12:19:51] <sdake>expecting operators to figure out the format of passwords.yml doesnt feel right | 19:20 |
sdake | this is not a feature its also a bug - default cannot be deployed without setting of the passwords | 19:21 |
sdake | you cn ix that with your tool or qsg, up to you | 19:21 |
sdake | SamYaple when your done rebasing that patch with the bug id let me know and i'll remove the -2 | 19:22 |
sdake | or i'll just catch it during my normal reviews | 19:23 |
SamYaple | ive -2'd it, you can remove it now | 19:25 |
*** jtriley has joined #kolla | 19:27 | |
*** mbound has quit IRC | 19:29 | |
*** Jeffrey4l has quit IRC | 19:31 | |
sdake | thanks SamYaple - time for lucnh bbiaf | 19:34 |
SamYaple | o/ | 19:36 |
openstackgerrit | Merged openstack/kolla: Fixes incorrect swift service endpoints https://review.openstack.org/294702 | 19:38 |
*** banix has joined #kolla | 19:41 | |
*** vhosakot has quit IRC | 19:44 | |
*** vhosakot has joined #kolla | 19:48 | |
*** dwalsh has joined #kolla | 19:50 | |
mlima | kolla already supports lvm? | 19:55 |
sdake | already as in there is a patch in the queue to add it | 19:56 |
sdake | it is blocked because it is a feature - but epect it to land early april | 19:56 |
mlima | ok, nice | 19:57 |
mlima | this will make configuration easier manila | 19:57 |
SamYaple | yea its on track to be one of the first newton merges which is fantastic | 19:58 |
sdake | mlima did you know you have the lowest +/- ratio of any of our serious reviewers over the last 1 year? :) | 19:58 |
sdake | what is cool about it is sbezverk didn't know how to use git when he started contributing ;-) | 19:58 |
sdake | i can't tell yu how many people i've taught git basics that went on to be serious contributors :) | 19:59 |
vhosakot | yes, I remember sbezverk's first "DO NOT MERGE" test patch set to test gerrit with kolla repo... :) cool sbezverk | 20:00 |
sbezverk | sdake mlima lvm patch has been tested in my test bed, mlima would you be able to try it? | 20:00 |
*** dencaval has quit IRC | 20:02 | |
sbezverk | vhosakot that do not merge was not even done by my, sdkae was pressing on my fingers ;-) | 20:02 |
*** mlima has quit IRC | 20:03 | |
sdake | using my powers of telekensis over webex apparnetly ;-) | 20:03 |
*** banix has quit IRC | 20:04 | |
*** dwalsh has quit IRC | 20:04 | |
vhosakot | sbezverk: cool, I did that too internally in my company... to check my git/gerrit settings :) | 20:07 |
*** Serlex has quit IRC | 20:16 | |
ccesario_ | SamYaple, did you run the deploy ? any feedback!? | 20:20 |
*** banix has joined #kolla | 20:21 | |
SamYaple | ccesario_: im sorry, busy busy day | 20:21 |
SamYaple | ccesario_: it never completed, different error | 20:21 |
ccesario_ | SamYaple, np .. | 20:21 |
ccesario_ | SamYaple, but error related to mysql ? | 20:22 |
*** Marga_ has joined #kolla | 20:22 | |
SamYaple | no | 20:23 |
SamYaple | my registry | 20:23 |
SamYaple | i have network issue atm | 20:23 |
SamYaple | outside of kolla | 20:23 |
ccesario_ | oh... ok | 20:23 |
*** banix has quit IRC | 20:24 | |
*** mbound has joined #kolla | 20:30 | |
*** mbound has quit IRC | 20:35 | |
*** haplo37 has joined #kolla | 20:35 | |
*** nihilifer has quit IRC | 20:52 | |
*** nihilifer has joined #kolla | 20:53 | |
*** mbound has joined #kolla | 20:56 | |
*** gfidente has quit IRC | 21:07 | |
*** zaneb has quit IRC | 21:08 | |
*** diogogmt has quit IRC | 21:21 | |
*** vhosakot_ has joined #kolla | 21:23 | |
*** diogogmt has joined #kolla | 21:24 | |
*** gbraad has quit IRC | 21:25 | |
*** thomasem has quit IRC | 21:25 | |
*** thomasem has joined #kolla | 21:25 | |
*** vhosakot has quit IRC | 21:26 | |
*** m3m0 has quit IRC | 21:26 | |
*** gbraad has joined #kolla | 21:27 | |
*** m3m0 has joined #kolla | 21:28 | |
*** Marga__ has joined #kolla | 21:45 | |
*** sdake has quit IRC | 21:47 | |
*** Marga__ has quit IRC | 21:48 | |
*** Marga_ has quit IRC | 21:48 | |
*** Marga_ has joined #kolla | 21:48 | |
*** Marga_ has quit IRC | 21:56 | |
*** sdake has joined #kolla | 22:00 | |
*** Marga_ has joined #kolla | 22:13 | |
*** haplo37 has quit IRC | 22:13 | |
*** sdake_ has joined #kolla | 22:16 | |
*** sdake has quit IRC | 22:20 | |
*** sdake_ has quit IRC | 22:24 | |
*** SiRiuS_ has quit IRC | 22:24 | |
*** jtriley has quit IRC | 22:25 | |
*** salv-orl_ has joined #kolla | 22:28 | |
*** ccesario has joined #kolla | 22:30 | |
*** salv-orlando has quit IRC | 22:30 | |
*** jtriley has joined #kolla | 22:40 | |
*** mbound has quit IRC | 22:44 | |
*** jasonsb has joined #kolla | 22:51 | |
*** blahRus has quit IRC | 23:02 | |
*** jtriley has quit IRC | 23:04 | |
*** jtriley has joined #kolla | 23:06 | |
*** Marga__ has joined #kolla | 23:11 | |
*** vhosakot_ has quit IRC | 23:14 | |
*** Marga_ has quit IRC | 23:15 | |
*** ccesario has quit IRC | 23:17 | |
*** ccesario has joined #kolla | 23:17 | |
*** Marga__ has quit IRC | 23:23 | |
*** Marga_ has joined #kolla | 23:24 | |
*** alisonh has joined #kolla | 23:26 | |
*** JoseMello has quit IRC | 23:29 | |
*** mbound has joined #kolla | 23:44 | |
*** mbound has quit IRC | 23:49 | |
*** jtriley has quit IRC | 23:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!