| rm_work | so in this test | 00:00 |
|---|---|---|
| rm_work | is raises a quota error | 00:00 |
| rm_work | *it | 00:00 |
| rm_work | then does a rollback | 00:00 |
| rm_work | ... but the object still shows up afterwards if i do a get | 00:00 |
| rm_work | so it was committed?! | 00:00 |
| johnsom | The empty record should be there | 00:00 |
| johnsom | but the load balancer should not be there | 00:01 |
| rm_work | right, the LB is there | 00:01 |
| rm_work | which is ... what | 00:01 |
| johnsom | That sounds a lot like the bug I was seeing, maybe | 00:01 |
| rm_work | it was created in the lock_session | 00:01 |
| rm_work | and the lock_session rolls back | 00:01 |
| johnsom | Yeah, that should not happen | 00:01 |
| rm_work | ... and the object exists after | 00:01 |
| johnsom | That is bad | 00:02 |
| rm_work | wtf | 00:02 |
| rm_work | this makes zero sense | 00:02 |
| rm_work | yeah wtf | 00:03 |
| rm_work | it's committing | 00:03 |
| rm_work | hold on gotta figure out why | 00:07 |
| rm_work | wtf | 00:08 |
| rm_work | umm | 00:08 |
| rm_work | wtf | 00:10 |
| rm_work | this is so weird | 00:10 |
| openstackgerrit | Merged openstack/octavia master: Fix cascade delete flow (pool quota requirements missing) https://review.openstack.org/437200 | 00:11 |
| rm_work | how do i tell if a session is dirty | 00:12 |
| *** fnaval has quit IRC | 00:16 | |
| rm_work | ok so err | 00:18 |
| *** yamamoto_ has joined #openstack-lbaas | 00:25 | |
| openstackgerrit | Merged openstack/octavia master: Allow to create vip in lb-mgmt-net https://review.openstack.org/431179 | 00:26 |
| *** catintheroof has quit IRC | 00:29 | |
| *** catintheroof has joined #openstack-lbaas | 00:30 | |
| *** catintheroof has quit IRC | 00:30 | |
| rm_work | ok so this: | 00:30 |
| rm_work | lock_session = db_api.get_session(autocommit=False) | 00:31 |
| rm_work | self.repos.create_load_balancer_and_vip( | 00:31 |
| rm_work | lock_session, lb_dict, vip_dict) | 00:31 |
| rm_work | print(lock_session.query(db_models.LoadBalancer).all()) | 00:31 |
| rm_work | lock_session.rollback() | 00:31 |
| rm_work | print(lock_session.query(db_models.LoadBalancer).all()) | 00:31 |
| rm_work | works fine | 00:31 |
| rm_work | so something else is causing a commit | 00:31 |
| rm_work | somehow | 00:31 |
| *** ducttape_ has quit IRC | 00:34 | |
| *** catintheroof has joined #openstack-lbaas | 00:42 | |
| rm_work | ok so | 00:43 |
| rm_work | check_quota_met does something wonky to lock_session and breaks it | 00:43 |
| rm_work | but not seeing how | 00:43 |
| *** fnaval has joined #openstack-lbaas | 00:45 | |
| rm_work | like, after that function has run once, anything we do with lock_session autocommits | 00:46 |
| rm_work | johnsom: https://gist.github.com/rm-you/1e0ee979774c4d7117a7063eb39ed23b | 00:53 |
| johnsom | Ummm | 00:54 |
| rm_work | i seriously can't imagine why | 00:55 |
| rm_work | is that a sqlite issue with SELECT FOR UPDATE ? | 00:55 |
| johnsom | It's even a different object location | 00:55 |
| rm_work | oh lol i didn't even notice that | 00:56 |
| johnsom | One odd thing is I did't think you could use a session after it had been rolled back | 00:56 |
| rm_work | lol well | 00:57 |
| johnsom | I wonder if this is sqlite issue | 00:57 |
| rm_work | apparently you can? :P | 00:57 |
| rm_work | right, yeah | 00:57 |
| rm_work | same | 00:57 |
| rm_work | let me narrow this down | 00:58 |
| rm_work | aaah hold on | 00:58 |
| rm_work | so | 00:58 |
| rm_work | we do: | 00:58 |
| rm_work | quotas = lock_session.query(models.Quotas).filter_by( | 00:58 |
| rm_work | project_id=project_id).with_for_update().first() | 00:58 |
| rm_work | so we've got this for_update quotas object | 00:58 |
| rm_work | we return False | 00:59 |
| rm_work | ... | 00:59 |
| rm_work | we never release that | 00:59 |
| rm_work | do we have to release that? | 00:59 |
| johnsom | I wondered the same, but I think the lock is allocated to the session. It should release on commit/rollback. | 00:59 |
| rm_work | looking it up | 01:00 |
| johnsom | Otherwise on single-call It would bomb out with repeated for_updates | 01:00 |
| johnsom | Still doesn't answer the mystery LB... | 01:00 |
| rm_work | i mean | 01:00 |
| rm_work | hmm | 01:00 |
| rm_work | yeah i think you should be right | 01:02 |
| rm_work | it's not the for_update | 01:04 |
| rm_work | this makes no freaking sense | 01:05 |
| *** yuanying_ has quit IRC | 01:05 | |
| korean101 | hi johnsom | 01:08 |
| johnsom | korean101 Hello | 01:08 |
| korean101 | johnsom: i try to Ubunru 16.04 newton releases | 01:09 |
| korean101 | a few days ago. i got trouble DVR (ARP) problem | 01:09 |
| korean101 | and i discover my 16.04 deployment doesn't any ARP problem | 01:09 |
| korean101 | so i try to deploy octavia on my 16.04 | 01:09 |
| rm_work | johnsom: updated | 01:10 |
| korean101 | and got some ERROR :( | 01:10 |
| rm_work | *updated the gist, check it out | 01:10 |
| rm_work | it's not the for_update query | 01:10 |
| rm_work | so then ... wtf else does CQM do | 01:10 |
| korean101 | i already succeed to deploy octavia on my CentOS 7 (Newton) | 01:11 |
| korean101 | but my 16.04 deployment got some errors... | 01:11 |
| rm_work | that's literally the only time it touches lock_session T_T | 01:11 |
| korean101 | do you have any time for glance to my ERRORS? (http://paste.openstack.org/show/602140/) | 01:11 |
| korean101 | another ERRORs (http://paste.openstack.org/show/602141/) | 01:12 |
| korean101 | my CentOS 7 Octavia works fine | 01:12 |
| korean101 | but my 16.04 Octavia has some trouble.... | 01:13 |
| rm_work | I just switched to centos7 as well for amps | 01:14 |
| johnsom | korean101 It looks like a keystone configuration issue | 01:14 |
| korean101 | johnsom: neutron.conf 's keystone configuration? | 01:15 |
| johnsom | yes | 01:15 |
| korean101 | johnsom: but... i can create instances | 01:15 |
| korean101 | johnsom: not LB | 01:15 |
| korean101 | johnsom: i already succeed deploy two instances | 01:16 |
| johnsom | nova instances? | 01:16 |
| korean101 | johnsom: yes! | 01:16 |
| korean101 | johnsom: but (neutron lbaas-loadbalancer-create --name test-lb demo-subnet) this commands some ERROS | 01:16 |
| johnsom | Yeah, that would not need the keystone token for neutron | 01:17 |
| johnsom | Give me a minute to look at that code to maybe narrow it down. | 01:17 |
| korean101 | johnsom: many thanks!!! | 01:17 |
| korean101 | rm_work: my CentOS 7's Octavia works fine! | 01:17 |
| korean101 | rm_work: but there is some (DVR+ARP) problems | 01:18 |
| korean101 | rm_work: DVR... :( | 01:18 |
| johnsom | DVR :( indeed | 01:18 |
| korean101 | johnsom: i use pip install octavia (0.9.0) on my 16.04 | 01:19 |
| korean101 | johnsom: i use Octavia rpm packages on my CentOS 7 | 01:19 |
| *** yuanying has joined #openstack-lbaas | 01:19 | |
| johnsom | korean101 look in your /etc/neutron/neutron.conf for a section called [keystone_authtoken] | 01:20 |
| johnsom | Make sure all of that looks right | 01:20 |
| korean101 | johnsom: hmmmmm. (http://paste.openstack.org/show/602142/) it looks ok | 01:22 |
| johnsom | korean101 Mine looks like: | 01:23 |
| johnsom | http://paste.openstack.org/show/602143/ | 01:23 |
| johnsom | It is using keystone version 2, which is odd | 01:25 |
| johnsom | most things are version 3 now | 01:25 |
| johnsom | korean101 It really looks like one of those first three lines is not right or the version isn't right. It's trying to do an HTTP POST to something and not finding it for keystone. | 01:27 |
| *** armax has joined #openstack-lbaas | 01:27 | |
| korean101 | johnsom: i followed this manuls (https://docs.openstack.org/newton/install-guide-ubuntu/neutron-compute-install.html#configure-the-common-component) | 01:28 |
| korean101 | Newton releases + Octavia 0.9.0 | 01:29 |
| johnsom | Ah, newton, yeah, keystone 2 is probably right then | 01:29 |
| johnsom | Same with the config file format | 01:29 |
| johnsom | I would try testing each of those urls to see if they work | 01:29 |
| korean101 | johnsom: but if that config is wrong, then nova boot command fails right? | 01:30 |
| korean101 | johnsom: but my nova boot command works fine TT | 01:30 |
| johnsom | No, nova has it's own configuration | 01:30 |
| johnsom | This is a neutron <-> keystone problem | 01:31 |
| korean101 | johnsom: but neutron net subnet router create works fine | 01:31 |
| korean101 | johnsom: just LB create problems | 01:31 |
| johnsom | Maybe that is not calling to keystone. | 01:31 |
| korean101 | johnsom: ah! OK | 01:31 |
| johnsom | I don't know of another neutron command off my head that calls to keystone, I'm sure there are some, I'm just not sure which | 01:32 |
| *** yamamoto_ has quit IRC | 01:32 | |
| korean101 | johnsom: yes. got it | 01:33 |
| *** ducttape_ has joined #openstack-lbaas | 01:34 | |
| johnsom | I need to take my wife to dinner now. You could try asking for help in the neutron channel, someone there might know how to help you get neutron configured correctly for keystone. Since nova is working I expect keystone is running ok, it's just a configuration problem in neutron.conf | 01:35 |
| korean101 | johnsom: yes. many thanks! | 01:36 |
| korean101 | johnsom: really thanks | 01:36 |
| korean101 | :) | 01:36 |
| *** ducttape_ has quit IRC | 01:40 | |
| rm_work | johnsom: whelp i'll keep hacking on this but i DO NOT see how this is breaking | 01:46 |
| *** blogan has joined #openstack-lbaas | 01:51 | |
| *** blogan has quit IRC | 01:51 | |
| *** sanfern has quit IRC | 01:53 | |
| *** sanfern has joined #openstack-lbaas | 01:54 | |
| *** fnaval has quit IRC | 02:01 | |
| *** ducttape_ has joined #openstack-lbaas | 02:05 | |
| *** yamamoto_ has joined #openstack-lbaas | 02:14 | |
| *** gongysh has joined #openstack-lbaas | 02:16 | |
| rm_work | johnsom: lb_count = self.session.query(db_models.LoadBalancer).filter( | 02:20 |
| rm_work | db_models.LoadBalancer.project_id == project_id).count() | 02:20 |
| rm_work | that | 02:20 |
| rm_work | when that runs, it breaks the *lock_session* | 02:20 |
| rm_work | no freaking clue why | 02:20 |
| *** yuanying has quit IRC | 02:20 | |
| rm_work | it's specifically the .count() | 02:21 |
| *** yuanying has joined #openstack-lbaas | 02:21 | |
| rm_work | well... anything that triggers it to actually fetch the dataset. all() does it too | 02:22 |
| *** _ducttape_ has joined #openstack-lbaas | 02:28 | |
| *** ducttape_ has quit IRC | 02:30 | |
| rm_work | johnsom: updated https://gist.github.com/rm-you/1e0ee979774c4d7117a7063eb39ed23b | 02:33 |
| *** yuanying has quit IRC | 02:48 | |
| *** _ducttape_ has quit IRC | 03:02 | |
| *** ducttape_ has joined #openstack-lbaas | 03:30 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/octavia master: Updated from global requirements https://review.openstack.org/444079 | 03:35 |
| *** links has joined #openstack-lbaas | 03:38 | |
| *** anilvenkata has joined #openstack-lbaas | 03:43 | |
| *** anilvenkata_ has joined #openstack-lbaas | 03:45 | |
| *** anilvenkata has quit IRC | 03:47 | |
| *** ducttape_ has quit IRC | 03:49 | |
| *** cody-somerville has joined #openstack-lbaas | 03:51 | |
| *** csomerville has quit IRC | 03:54 | |
| *** catintheroof has quit IRC | 03:57 | |
| *** catintheroof has joined #openstack-lbaas | 03:58 | |
| *** catintheroof has quit IRC | 03:59 | |
| *** ducttape_ has joined #openstack-lbaas | 03:59 | |
| *** cody-somerville has quit IRC | 04:06 | |
| *** korean101 has quit IRC | 04:10 | |
| *** armax has quit IRC | 04:10 | |
| *** yuanying has joined #openstack-lbaas | 04:10 | |
| *** ducttape_ has quit IRC | 04:13 | |
| *** amotoki has joined #openstack-lbaas | 05:13 | |
| *** gcheresh_ has joined #openstack-lbaas | 05:22 | |
| *** rm_mobile has joined #openstack-lbaas | 05:22 | |
| rm_mobile | johnsom: it is almost assuredly this: http://docs.sqlalchemy.org/en/latest/dialects/sqlite.html#serializable-isolation-savepoints-transactional-ddl | 05:22 |
| rm_mobile | I'll try that workaround when I get back home | 05:23 |
| rm_mobile | And I'll check using MySQL and see if this issue just doesn't happen | 05:23 |
| rm_mobile | Guessing it will not | 05:23 |
| rm_mobile | Oh, was that WebEx today? Did I miss it? :/ | 05:25 |
| *** cody-somerville has joined #openstack-lbaas | 05:26 | |
| *** cody-somerville has quit IRC | 05:26 | |
| *** cody-somerville has joined #openstack-lbaas | 05:26 | |
| johnsom | It was, you did | 05:28 |
| johnsom | Reading | 05:29 |
| rm_mobile | How did it go | 05:29 |
| rm_mobile | What's your impression | 05:29 |
| johnsom | I think they have a good idea. The next step is to write up a spec. | 05:35 |
| johnsom | It sounds like they have people to implement as well. | 05:36 |
| johnsom | They did an internal PoC that worked | 05:36 |
| johnsom | So I am pretty happy about it. My only worry is will they be ready before we have the base act/act patches merged | 05:37 |
| johnsom | This pysqlite stuff reminds me why I am not a huge fan of all things python.... | 05:38 |
| *** gcheresh_ has quit IRC | 05:38 | |
| johnsom | I mean really? We think you really want autocommit even though you explicitly ask for a transaction... fail | 05:39 |
| *** korean101 has joined #openstack-lbaas | 05:57 | |
| korean101 | johnsom: i got resolution | 05:58 |
| korean101 | johnsom: (https://docs.openstack.org/newton/config-reference/networking/samples/neutron_lbaas.conf.html) | 05:58 |
| korean101 | johnsom: above links, [service_auth] section only in neutron_lbaas.conf | 05:58 |
| korean101 | johnsom: but Newton on 16.04, neutron.conf also must contains [service_auth] section | 05:59 |
| korean101 | johnsom: must! | 05:59 |
| korean101 | johnsom: neutron.conf doesn't have [service_auth] section. then ERROR flood~ | 05:59 |
| korean101 | johnsom: i got this messages (http://paste.openstack.org/show/602177/) | 06:00 |
| johnsom | Ah, sorry I didn't catch that. | 06:00 |
| korean101 | johnsom: problem solved! | 06:00 |
| korean101 | johnsom: but this is so weird. | 06:00 |
| johnsom | Glad you got it figured out | 06:00 |
| korean101 | johnsom: incording to openstack manuals, that section only exists in neutron_lbaas.conf | 06:01 |
| rm_mobile | Johnsom: honestly, if they finish this act/act first... Well... Code wins... | 06:01 |
| johnsom | Yeah, the docs have some issues | 06:01 |
| korean101 | johnsom: Thank you every time. :) | 06:01 |
| rm_mobile | I have no attachment to any particular implementation, and the existing one seems a little clunky and black-boxy IMO | 06:02 |
| johnsom | rm_mobile I think they are planning to do a driver for the IBM stuff | 06:02 |
| johnsom | Thus will build on that code | 06:03 |
| rm_mobile | Ah | 06:04 |
| rm_mobile | Well maybe they'll help push that stuff then | 06:04 |
| *** aojea has joined #openstack-lbaas | 06:12 | |
| *** ducttape_ has joined #openstack-lbaas | 06:14 | |
| *** ducttape_ has quit IRC | 06:19 | |
| *** rcernin has joined #openstack-lbaas | 06:35 | |
| *** kobis has joined #openstack-lbaas | 06:38 | |
| *** aojea has quit IRC | 06:49 | |
| *** amotoki has quit IRC | 06:50 | |
| *** rm_mobile has quit IRC | 06:55 | |
| *** aojea has joined #openstack-lbaas | 07:20 | |
| *** amotoki has joined #openstack-lbaas | 07:23 | |
| *** tesseract has joined #openstack-lbaas | 07:42 | |
| *** aojea has quit IRC | 07:47 | |
| *** aojea has joined #openstack-lbaas | 07:53 | |
| *** gcheresh_ has joined #openstack-lbaas | 08:04 | |
| *** aojea has quit IRC | 08:14 | |
| *** aojea has joined #openstack-lbaas | 08:15 | |
| *** aojea has quit IRC | 08:19 | |
| *** yuanying has quit IRC | 08:58 | |
| *** eezhova has joined #openstack-lbaas | 09:02 | |
| *** yamamoto_ has quit IRC | 09:04 | |
| rm_work | johnsom: OK yeah tested with MySQL and it is working fine... SQLite bug, ffff | 09:05 |
| rm_work | I guess we just freaking make a note in the test and ... !?!?! | 09:06 |
| rm_work | we just can't test it | 09:06 |
| rm_work | this is so dumb | 09:06 |
| *** gcheresh_ has quit IRC | 09:39 | |
| *** yuanying has joined #openstack-lbaas | 09:41 | |
| *** ducnc has quit IRC | 09:54 | |
| *** ducnc has joined #openstack-lbaas | 09:55 | |
| *** aojea has joined #openstack-lbaas | 10:08 | |
| *** ducnc has quit IRC | 10:21 | |
| *** yamamoto has joined #openstack-lbaas | 10:22 | |
| *** yamamoto has quit IRC | 10:26 | |
| *** gongysh has quit IRC | 10:26 | |
| *** openstackgerrit has quit IRC | 10:33 | |
| rm_work | ugh screw this whole test, i think | 11:15 |
| *** yamamoto has joined #openstack-lbaas | 11:26 | |
| *** gcheresh_ has joined #openstack-lbaas | 11:32 | |
| *** yamamoto has quit IRC | 11:35 | |
| *** links has quit IRC | 11:37 | |
| *** gongysh has joined #openstack-lbaas | 11:41 | |
| *** links has joined #openstack-lbaas | 11:54 | |
| *** aojea has quit IRC | 12:02 | |
| *** ducttape_ has joined #openstack-lbaas | 12:19 | |
| *** ducttape_ has quit IRC | 12:24 | |
| *** yamamoto has joined #openstack-lbaas | 12:24 | |
| *** yamamoto has quit IRC | 12:26 | |
| *** yamamoto has joined #openstack-lbaas | 12:27 | |
| *** yamamoto has quit IRC | 12:27 | |
| *** catintheroof has joined #openstack-lbaas | 12:34 | |
| *** sanfern has quit IRC | 12:45 | |
| *** sanfern has joined #openstack-lbaas | 12:45 | |
| *** ducttape_ has joined #openstack-lbaas | 13:00 | |
| *** Andrew_jedi has joined #openstack-lbaas | 13:01 | |
| Andrew_jedi | Hello folks, I have a question on Octavia. Is it correct to assume that each amphora is always directly linked to its tenant? Or can a single service VM be shared across multiple tenants? | 13:03 |
| *** gcheresh_ has quit IRC | 13:12 | |
| *** links has quit IRC | 13:17 | |
| *** yamamoto has joined #openstack-lbaas | 13:27 | |
| *** links has joined #openstack-lbaas | 13:31 | |
| *** gongysh has quit IRC | 13:32 | |
| *** yamamoto has quit IRC | 13:33 | |
| *** yamamoto has joined #openstack-lbaas | 13:34 | |
| *** links has quit IRC | 13:39 | |
| *** zioproto has joined #openstack-lbaas | 13:51 | |
| zioproto | hello | 13:51 |
| zioproto | so good to find so many people here | 13:51 |
| zioproto | I really need help with LBaaS | 13:51 |
| zioproto | I cant upgrade to Newton if dont finish this migration from LBaaSv1 to LBaaSv2 | 13:52 |
| zioproto | at the moment I am trying to have a Horizon running with web dashboard for LBaaSv2, anyone has this thing working ? | 13:52 |
| *** aojea has joined #openstack-lbaas | 13:52 | |
| zioproto | I have the same problems described here: https://ask.openstack.org/en/question/96790/lbaasv2-dashboard-issues/ | 14:10 |
| zioproto | is anyone alive in this channel ? :) | 14:10 |
| johnsom | Andrew_jedi currently it is one or more service vms per load balancer. They cannot be shared. | 14:13 |
| Andrew_jedi | johnsom: In the same tenant ? | 14:14 |
| Andrew_jedi | johnsom: gotcha, thank you !! | 14:14 |
| johnsom | The service vms run under the octavia project, they are not in the user tenant. | 14:15 |
| johnsom | zioproto Maybe xgerman can help you, he recently tested out the panels | 14:18 |
| *** ducttape_ has quit IRC | 14:18 | |
| xgerman | hi | 14:37 |
| xgerman | recently is like 3 months ago | 14:37 |
| *** cpuga has joined #openstack-lbaas | 14:42 | |
| *** aojea_ has joined #openstack-lbaas | 14:48 | |
| *** yamamoto has quit IRC | 14:50 | |
| *** yamamoto has joined #openstack-lbaas | 14:50 | |
| *** Andrew_jedi has quit IRC | 14:51 | |
| *** aojea has quit IRC | 14:52 | |
| *** yamamoto has quit IRC | 14:55 | |
| *** ducttape_ has joined #openstack-lbaas | 14:56 | |
| *** catintheroof has quit IRC | 15:00 | |
| *** catintheroof has joined #openstack-lbaas | 15:01 | |
| *** gcheresh_ has joined #openstack-lbaas | 15:07 | |
| *** beardedeagle has joined #openstack-lbaas | 15:12 | |
| *** chlong_ has joined #openstack-lbaas | 15:15 | |
| *** kobis has quit IRC | 15:17 | |
| *** kobis has joined #openstack-lbaas | 15:18 | |
| *** kobis has quit IRC | 15:18 | |
| *** kobis has joined #openstack-lbaas | 15:18 | |
| *** kobis has quit IRC | 15:19 | |
| *** openstackgerrit has joined #openstack-lbaas | 15:22 | |
| openstackgerrit | German Eichberger proposed openstack/octavia master: [WIP] Adds the abiity to fsilover a load balancer https://review.openstack.org/444347 | 15:22 |
| *** ducttape_ has quit IRC | 15:34 | |
| *** Andrew_jedi has joined #openstack-lbaas | 15:37 | |
| Andrew_jedi | xgerman: Does horizon has support for Octavia ? | 15:39 |
| xgerman | yes, it supports whatever LBaaS V2 driver you put behind it | 15:39 |
| Andrew_jedi | well, getting lbaasv2 to horizon is itself challenging | 15:40 |
| xgerman | I just did a pip install and followed the instructions… | 15:40 |
| Andrew_jedi | wow, and it worked , thank you for letting me know ... | 15:41 |
| xgerman | yeah, but that was a few months ago and things change fast in OpenStack | 15:45 |
| Andrew_jedi | xgerman: Ok, do you have any idea about lbaasv2 horizon support in Mitaka release? | 15:47 |
| xgerman | that should work… since we made the Panel for mitaka | 15:47 |
| xgerman | I tested Newton and early Ocata and both worked | 15:48 |
| *** gcheresh_ has quit IRC | 15:52 | |
| Andrew_jedi | great, thanks :) | 15:54 |
| *** fnaval has joined #openstack-lbaas | 15:57 | |
| *** rcernin has quit IRC | 15:59 | |
| *** anilvenkata_ has quit IRC | 16:07 | |
| *** ducttape_ has joined #openstack-lbaas | 16:09 | |
| *** Andrew_jedi has quit IRC | 16:18 | |
| *** chlong_ has quit IRC | 16:19 | |
| johnsom | rm_work So the work around in the link you sent doesn't work? | 16:19 |
| *** ducttape_ has quit IRC | 16:23 | |
| xgerman | johnsom so with working on the failover I would like to put that under loadbalancer/123/failover | 16:24 |
| johnsom | Yeah, I think that is a good plan | 16:24 |
| xgerman | awesome, now I just need to put my patch on top of the LB API patch ;-) | 16:25 |
| xgerman | I am also deleting first the passive amphora and then the active — now I am not sure how an operator would figure out if the failover was successful (since I hand it off to the queue) | 16:25 |
| xgerman | also am I supposed to mark the LB pending whie we failover | 16:26 |
| xgerman | and if the failover fails mark it ERROR - which would probably be annoying | 16:26 |
| xgerman | since we don’t mark it ERROR if the amp failover in health fails | 16:27 |
| johnsom | Yeah, passive then active, we should be able to track failover with the amp ID changing and amphora_health messages coming in for the new amp. | 16:30 |
| *** blogan has joined #openstack-lbaas | 16:30 | |
| johnsom | LB should go to pending_update | 16:30 |
| johnsom | It should probably go to error if failover fails. | 16:31 |
| johnsom | Also, there is a busy flag that needs set, the failover flow sets that flag, so depending on what you reuse you may get it for free | 16:32 |
| xgerman | I am reusing those flows | 16:32 |
| xgerman | just wasn’t sure about what to do with the LB | 16:33 |
| *** eezhova has quit IRC | 16:48 | |
| *** yamamoto has joined #openstack-lbaas | 16:51 | |
| *** logan_ has joined #openstack-lbaas | 16:56 | |
| *** logan_ is now known as Guest29589 | 16:57 | |
| *** logan- has quit IRC | 16:59 | |
| *** Guest29589 is now known as logan- | 17:03 | |
| *** aojea_ has quit IRC | 17:05 | |
| *** catintheroof has quit IRC | 17:07 | |
| *** catintheroof has joined #openstack-lbaas | 17:08 | |
| *** yamamoto has quit IRC | 17:11 | |
| *** ducttape_ has joined #openstack-lbaas | 17:20 | |
| *** armax has joined #openstack-lbaas | 17:37 | |
| openstackgerrit | Sindhu Devale proposed openstack/octavia master: DB migration to add ID column to HM https://review.openstack.org/430451 | 17:49 |
| *** ducttape_ has quit IRC | 17:50 | |
| *** kobis has joined #openstack-lbaas | 17:53 | |
| *** kobis has quit IRC | 17:53 | |
| *** zioproto has quit IRC | 18:11 | |
| *** sanfern has quit IRC | 18:25 | |
| *** malos has quit IRC | 18:29 | |
| *** ducttape_ has joined #openstack-lbaas | 18:32 | |
| *** anilvenkata_ has joined #openstack-lbaas | 18:48 | |
| *** kobis has joined #openstack-lbaas | 18:56 | |
| *** kobis has quit IRC | 18:56 | |
| *** kobis has joined #openstack-lbaas | 19:05 | |
| *** aojea has joined #openstack-lbaas | 19:11 | |
| *** eezhova has joined #openstack-lbaas | 19:24 | |
| *** chlong_ has joined #openstack-lbaas | 19:42 | |
| diltram | johnsom: https://review.openstack.org/#/c/444482/ | 19:44 |
| *** anilvenkata_ has quit IRC | 19:51 | |
| *** eezhova has quit IRC | 20:06 | |
| *** aojea has quit IRC | 20:08 | |
| *** aojea has joined #openstack-lbaas | 20:09 | |
| *** ducttape_ has quit IRC | 20:13 | |
| *** aojea has quit IRC | 20:13 | |
| *** eezhova has joined #openstack-lbaas | 20:19 | |
| *** ducttape_ has joined #openstack-lbaas | 20:26 | |
| johnsom | diltram on the above patch, a few lines above there is also a lowercase true. Should it change? | 20:55 |
| diltram | johnsom: this previous one is not used by us | 20:55 |
| diltram | and because of this I made this lowercase word | 20:55 |
| diltram | and it was a mistake | 20:56 |
| diltram | it's internal python env | 20:56 |
| johnsom | Ok | 20:56 |
| diltram | johnsom: which cloud-init datasource is responsible just for reading /var/lib/cloud directory? | 20:57 |
| johnsom | I have no idea unless that is where the config drive device gets mounted. I would have to look. Config drive comes in as an sr0 device | 20:58 |
| diltram | yeah, I know | 20:58 |
| diltram | but lxd guys are making weird things and they already mounted the drive to this directory | 20:59 |
| diltram | and I can't configure it properly to read it from there | 20:59 |
| johnsom | diltram This is the source of information I have used previously: http://cloudinit.readthedocs.io/en/latest/topics/datasources.html#datasource-documentation | 21:00 |
| diltram | johnsom: I was reading this webiste | 21:01 |
| diltram | but completely can't find anything valueable there | 21:01 |
| diltram | what could be used to help | 21:01 |
| diltram | http://cloudinit.readthedocs.io/en/latest/topics/dir_layout.html | 21:01 |
| diltram | based on this they're mounting this openstack cloud config drive to this dir and it's all | 21:02 |
| *** ducttape_ has quit IRC | 21:08 | |
| johnsom | Looking in an amp, cloud-init.log has some interactions with that directory. I can post a log if you don't have one handy. | 21:15 |
| johnsom | It's not super helpful though | 21:15 |
| *** blogan_ has joined #openstack-lbaas | 21:35 | |
| *** blogan has quit IRC | 21:36 | |
| *** blogan_ is now known as blogan | 21:38 | |
| *** blogan has quit IRC | 21:49 | |
| *** aojea has joined #openstack-lbaas | 21:53 | |
| *** blogan has joined #openstack-lbaas | 21:53 | |
| *** beardedeagle has quit IRC | 21:58 | |
| *** eezhova has quit IRC | 21:59 | |
| *** ducttape_ has joined #openstack-lbaas | 22:08 | |
| *** ducttape_ has quit IRC | 22:14 | |
| *** chlong_ has quit IRC | 22:21 | |
| openstackgerrit | Ihar Hrachyshka proposed openstack/octavia master: Update for new pep8 rules E402 and W503 https://review.openstack.org/443298 | 22:21 |
| *** catintheroof has quit IRC | 22:23 | |
| *** dileepr has joined #openstack-lbaas | 22:28 | |
| *** aojea has quit IRC | 22:57 | |
| *** aojea has joined #openstack-lbaas | 22:57 | |
| *** aojea has quit IRC | 23:02 | |
| *** cpuga has quit IRC | 23:07 | |
| *** tesseract has quit IRC | 23:29 | |
| *** fnaval has quit IRC | 23:41 | |
| *** kobis1 has joined #openstack-lbaas | 23:42 | |
| *** kobis has quit IRC | 23:44 | |
| *** fnaval has joined #openstack-lbaas | 23:55 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!