Thursday, 2018-02-15

*** fnaval has joined #openstack-lbaas00:10
*** fnaval has quit IRC00:52
*** fnaval has joined #openstack-lbaas00:52
*** jniesz has quit IRC01:20
*** sshank has quit IRC01:31
*** numans has quit IRC01:34
*** JudeC has quit IRC01:36
*** slaweq has joined #openstack-lbaas02:13
*** slaweq has quit IRC02:18
dayouCan somebody help to try this: https://review.openstack.org/#/c/520878/? Having some problems to get the load balancer up.02:46
*** harlowja has quit IRC03:08
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561203:19
openstackgerritAdam Harwell proposed openstack/octavia master: DNM: Remove PING type from HM options list  https://review.openstack.org/52843903:19
*** dougwig has quit IRC03:21
rm_workdayou: firstly, i'd recommend you switch to using the webserver setup we use in other testing -- the Golang binary we have checked in already03:39
rm_workbecause it is much less likely to break than the one you have03:39
rm_workwe went away from 'nc' for a reason03:39
rm_workhumorously you seem to also be using Golang for other stuff :P03:41
dayouyes, it's for lightweighted router reflector server03:42
rm_workyeah03:42
rm_workditch webserver.sh03:42
rm_workhttps://github.com/openstack/octavia/blob/master/octavia/tests/contrib/httpd.go03:43
rm_workwe checked in a binary somewhere...03:43
dayouSounds nice to use httpd, I remember the binary is in the tempest plugin side03:44
rm_workyeah i guess so03:44
rm_workyou could copy that binary into this tree as well if you need to03:44
rm_worki would be ok with that03:45
rm_workor you could compile it03:45
rm_worksince you have the source, and a golang compiler anyway03:45
dayouI guess I'll compile it and just copy the source, the be more space efficient for the git repository03:47
*** yamamoto has joined #openstack-lbaas03:47
*** yamamoto has quit IRC03:47
*** yamamoto has joined #openstack-lbaas03:51
*** fnaval has quit IRC04:08
openstackgerritHengqing Hu proposed openstack/octavia master: Provide devstack samples for l3 active active  https://review.openstack.org/52087804:11
openstackgerritHengqing Hu proposed openstack/octavia master: Provide devstack samples for l3 active active  https://review.openstack.org/52087804:22
*** fnaval has joined #openstack-lbaas04:24
*** links has joined #openstack-lbaas04:34
*** links has quit IRC04:42
*** links has joined #openstack-lbaas04:45
*** gcheresh has joined #openstack-lbaas05:04
*** threestrands_ has joined #openstack-lbaas05:09
*** threestrands_ has quit IRC05:09
*** threestrands_ has joined #openstack-lbaas05:09
*** threestrands has quit IRC05:09
*** threestrands_ has quit IRC05:10
*** threestrands_ has joined #openstack-lbaas05:11
*** threestrands_ has quit IRC05:11
*** threestrands_ has joined #openstack-lbaas05:11
*** harlowja has joined #openstack-lbaas05:15
*** threestrands has joined #openstack-lbaas05:15
*** threestrands has quit IRC05:15
*** threestrands has joined #openstack-lbaas05:15
*** threestrands_ has quit IRC05:18
*** threestrands has quit IRC05:28
*** gcheresh has quit IRC05:35
*** irenab has joined #openstack-lbaas05:52
*** harlowja has quit IRC06:14
*** slaweq has joined #openstack-lbaas06:15
*** kobis has joined #openstack-lbaas06:18
*** slaweq has quit IRC06:19
*** numans has joined #openstack-lbaas06:34
*** rcernin has quit IRC06:44
*** gcheresh has joined #openstack-lbaas07:09
*** threestrands has joined #openstack-lbaas07:09
*** threestrands has quit IRC07:09
*** threestrands has joined #openstack-lbaas07:09
*** JudeC has joined #openstack-lbaas07:21
*** JudeC has quit IRC07:26
*** rcernin has joined #openstack-lbaas07:28
*** pcaruana has joined #openstack-lbaas07:40
*** b_bezak has joined #openstack-lbaas08:04
*** AlexeyAbashkin has joined #openstack-lbaas08:08
*** slaweq_ has joined #openstack-lbaas08:15
*** tesseract has joined #openstack-lbaas08:17
*** slaweq_ has quit IRC08:20
*** links has quit IRC08:35
*** links has joined #openstack-lbaas08:52
*** threestrands_ has joined #openstack-lbaas09:21
*** threestrands has quit IRC09:21
rm_workLOL whoops09:56
rm_workjust realized I had fixed https://github.com/openstack/octavia/commit/4fb6fa2106f02b047b0710f70b8bd78c9d193894 in my flip driver a LONG time ago T_T09:56
rm_workshould probably have pulled that out into a patch :(09:56
*** b_bezak has quit IRC09:57
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561210:00
*** b_bezak has joined #openstack-lbaas10:02
cgoncalvesrm_work: shouldn't you be sleeping or something? :)10:02
rm_workI don't ... usually do that10:02
cgoncalvesHa :)10:03
rm_workgot charged up because i spent the last few hours helping my roommate optimize sqlalchemy queries10:03
rm_worki'm not sure if that's supposed to be "fun", but when it's not MY work it's a lot more interesting ;)10:05
cgoncalveswhen it's not our work we can always stop/give up at any time, so less pressure hence more fun10:06
rm_workprobably :P10:08
*** slaweq_ has joined #openstack-lbaas10:16
*** slaweq_ has quit IRC10:21
*** salmankhan has joined #openstack-lbaas10:22
*** salmankhan has quit IRC10:36
openstackgerritAllen proposed openstack/octavia-dashboard master: Fix bug when click into pool detail page  https://review.openstack.org/54388410:41
*** salmankhan has joined #openstack-lbaas10:50
*** b_bezak has quit IRC10:57
*** b_bezak has joined #openstack-lbaas10:58
*** salmankhan has quit IRC11:00
*** salmankhan has joined #openstack-lbaas11:00
*** yamamoto has quit IRC11:01
*** b_bezak has quit IRC11:02
*** b_bezak has joined #openstack-lbaas11:03
*** yamamoto has joined #openstack-lbaas11:06
*** b_bezak has quit IRC11:08
*** b_bezak has joined #openstack-lbaas11:21
*** isp has joined #openstack-lbaas11:22
*** b_bezak has quit IRC11:26
*** salmankhan has quit IRC11:34
*** salmankhan has joined #openstack-lbaas11:34
*** b_bezak has joined #openstack-lbaas11:40
*** yamamoto has quit IRC12:09
*** slaweq_ has joined #openstack-lbaas12:17
*** slaweq_ has quit IRC12:22
*** yamamoto has joined #openstack-lbaas12:22
*** sanfern has joined #openstack-lbaas12:28
*** salmankhan has quit IRC12:28
nmagnezicgoncalves, yeah rm_work never actually sleeps, he just pauses from time to time :D12:29
*** salmankhan has joined #openstack-lbaas12:30
*** yamamoto has quit IRC12:39
*** fnaval has quit IRC12:52
*** yamamoto has joined #openstack-lbaas12:54
*** yamamoto has quit IRC13:05
cgoncalvesnmagnezi: ha! he fails over to his own standby self :)13:22
*** fnaval has joined #openstack-lbaas13:27
*** sanfern has quit IRC13:36
*** isp has quit IRC13:40
*** pck has joined #openstack-lbaas13:41
*** pck has quit IRC13:41
*** pck has joined #openstack-lbaas13:41
*** salmankhan has quit IRC13:58
*** yamamoto has joined #openstack-lbaas14:00
*** salmankhan has joined #openstack-lbaas14:08
*** fnaval has quit IRC14:16
*** atoth has joined #openstack-lbaas14:32
*** fnaval has joined #openstack-lbaas14:34
*** threestrands_ has quit IRC14:39
*** yamamoto has quit IRC14:41
*** yamamoto has joined #openstack-lbaas14:48
*** yamamoto has quit IRC14:53
*** kobis has quit IRC14:53
*** kobis has joined #openstack-lbaas14:53
*** kobis has quit IRC14:54
*** kobis has joined #openstack-lbaas14:54
*** kobis has quit IRC14:54
*** rcernin has quit IRC15:01
xgerman_another day, anotheerLB: https://metallb.universe.tf/concepts/15:24
*** gcheresh has quit IRC15:24
*** atoth has quit IRC15:25
*** atoth has joined #openstack-lbaas15:27
*** yamamoto has joined #openstack-lbaas15:33
*** yamamoto has quit IRC15:38
mugsiexgerman_: that looks like a pretty good k8s baremetal LB15:38
mugsiewould have been handy in HDP days :P15:38
xgerman_yep, better than that VMWare crap15:39
dayouxgerman: A read to go LB which uses gobgp :P15:41
dayouready15:41
xgerman_yep, but in the k8s world using an LB, ingress, and the nodeport makes things loadbalanced three times15:45
dayouHow did you find this project?15:46
xgerman_the world of twitter15:46
dayouoops, that is like a wonder world to alice as it is to me :P15:49
*** pcaruana has quit IRC16:09
*** kobis has joined #openstack-lbaas16:15
*** yamamoto has joined #openstack-lbaas16:18
*** slaweq_ has joined #openstack-lbaas16:19
*** links has quit IRC16:22
*** yamamoto has quit IRC16:23
*** slaweq_ has quit IRC16:24
*** AlexeyAbashkin has quit IRC16:37
*** salmankhan has quit IRC16:55
*** salmankhan has joined #openstack-lbaas16:57
*** kobis has quit IRC17:02
*** yamamoto has joined #openstack-lbaas17:03
*** salmankhan has quit IRC17:06
*** yamamoto has quit IRC17:08
*** atoth has quit IRC17:15
*** b_bezak has quit IRC17:18
*** harlowja has joined #openstack-lbaas17:19
*** pcaruana has joined #openstack-lbaas17:20
*** atoth has joined #openstack-lbaas17:27
*** yamamoto has joined #openstack-lbaas17:32
*** yamamoto has quit IRC17:33
*** yamamoto has joined #openstack-lbaas17:41
*** sanfern has joined #openstack-lbaas17:45
*** yamamoto has quit IRC17:46
*** tesseract has quit IRC17:52
*** fnaval_ has joined #openstack-lbaas17:52
*** fnaval has quit IRC17:53
openstackgerritMichael Johnson proposed openstack/octavia master: Fix load balancers using IPv6 for the VIP subnet  https://review.openstack.org/54507017:56
*** AlexeyAbashkin has joined #openstack-lbaas18:01
sanfernHi johnsom,18:03
johnsomHi sanfern18:04
sanfernThank you for comments on distributor patch.18:05
johnsomSure, the driver stuff looks pretty good!18:06
*** AlexeyAbashkin has quit IRC18:06
sanfernbgp_driver - I was in impression we may need to change in future from exabgp to bird or any other implementation18:06
sanfernIf not needed I will remove that18:07
johnsomIt looked like that driver was replacing the rest_haproxy amphora driver.  I think we can just extend that driver and word those extensions in a way that is agnostic to the the actual implementation inside the amphora.18:07
sanfernThis is to load in the tasks18:09
sanfernok I will make that change18:10
*** SumitNaiksatam has joined #openstack-lbaas18:11
*** JudeC has joined #openstack-lbaas18:11
sanfernwe need to load the configurations into speaker and peer. There can be one speaker pairing with multiple peers. What's the best way to represent in the config?18:11
*** yamamoto has joined #openstack-lbaas18:11
sanfernSpeaker1 and Peer1 are sample type18:11
johnsomWould this list be unique per amp?18:12
johnsomYou could use a list type: https://github.com/openstack/octavia/blob/master/etc/octavia.conf#L60 or a dict type: https://docs.openstack.org/oslo.config/latest/reference/types.html#oslo_config.types.Dict18:14
sanfernnope it's a same for all amp18:14
johnsomhttps://docs.openstack.org/oslo.config/latest/reference/opts.html#oslo_config.cfg.DictOpt18:15
*** yamamoto has quit IRC18:16
*** slaweq has joined #openstack-lbaas18:20
*** slaweq has quit IRC18:25
*** yamamoto has joined #openstack-lbaas18:35
*** yamamoto has quit IRC18:35
*** atoth has quit IRC18:56
*** fnaval_ has quit IRC18:59
openstackgerritMichael Johnson proposed openstack/octavia master: Log output of subprocess check_output() errors  https://review.openstack.org/54509018:59
*** fnaval has joined #openstack-lbaas18:59
*** harlowja has quit IRC19:00
johnsomCores: I would like to get this into Queens, can you please review? https://review.openstack.org/#/c/545070/19:16
rm_work+219:22
*** kobis has joined #openstack-lbaas19:22
*** gcheresh has joined #openstack-lbaas19:26
xgerman_+2/A19:26
johnsomGroovy, thanks. I will get the backport going. I think we have two of those to do and I will cut the RC219:27
xgerman_k19:27
johnsomThis one: https://review.openstack.org/#/c/544526/19:28
rm_workaugh one of the client updates now forces some sort of version discovery thing for neutron endpoints, which doesn't work on liberty T_T19:32
rm_workso i can't run most neutron commands anymore19:33
*** slaweq has joined #openstack-lbaas19:33
johnsomOh, neat, yeah, mordred was working on that. He commented that our version discovery stuff isn't right too.19:35
*** harlowja has joined #openstack-lbaas19:35
rm_workuuuuugh19:35
rm_worknot neat19:35
rm_workeverything is broken19:35
johnsomThat is, interesting. I would file a bug against it frankly19:35
rm_workT_T19:35
rm_workso mordred would be a good contact?19:35
johnsomYes19:35
*** yamamoto has joined #openstack-lbaas19:35
johnsom#openstack-sdks19:35
rm_workk well i hit him up in neutron, will see19:36
*** harlowja_ has joined #openstack-lbaas19:37
xgerman_mmh, not sure if he hangs in neutron19:37
rm_workhe's technically *in there*19:38
* rm_work waves at harlowja 19:38
* rm_work waves at harlowja_ 19:38
* xgerman_ waves, too 19:38
harlowja_i am all19:38
harlowja_and none19:38
harlowja_lol19:38
harlowja_whats up19:38
*** harlowja has quit IRC19:39
*** yamamoto has quit IRC19:45
rm_workjohnsom: downgraded to openstacksdk==0.9.19 and it works again19:46
johnsomYep, I remember there were some late landing issues around that.  Note, you need a 0.10.x or newer for the dashboard19:47
johnsomThere was a bug fixed in our part of the sdk that didn't land until 10.x19:47
rm_workT_T_T_T19:54
*** sanfern has quit IRC19:54
rm_workwell this is for my local CLI and we don't use the dashboard anyway19:54
*** pcaruana has quit IRC20:03
rm_workjohnsom: where was that centos haproxy upgrade ticket they closed ...20:07
johnsomUmmm, more details please?20:08
rm_workremember we opened a ticket to get haproxy updated in the centos release20:08
johnsomOh, I did that in Ubuntu, I don't remember a CentOS one20:08
johnsomhttps://bugs.launchpad.net/ubuntu/+source/haproxy/+bug/174346520:08
openstackLaunchpad bug 1743465 in haproxy (Ubuntu) "Bionic should have haproxy 1.8-stable" [Undecided,Confirmed]20:08
rm_workah20:09
rm_worki remembered a centos one, i think nmagnezi had commented20:09
johnsomAh, yeah, I don't think I have that one20:09
rm_workand ryan ohara closed it20:10
rm_workand i was like20:10
rm_workdamnit ryan20:10
*** AlexeyAbashkin has joined #openstack-lbaas20:15
harlowja_how hard is it really to compile haproxy20:18
*** AlexeyAbashkin has quit IRC20:19
johnsomNot, but hey, they also have ancient versions of git too20:19
harlowja_ya, i was just thinking, like if they are ignoring u, then just like compile your own20:20
rm_workyeah i mean we could do a haproxy-source element20:20
rm_workand a flag for that20:20
rm_workwould get us 1.820:20
rm_workbut hey so will ubuntu20:20
rm_work<_<20:21
harlowja_or u know stop using distros completly20:21
harlowja_lol20:21
rm_workoctavia-linux? :P20:22
rm_workget some LFS going20:22
harlowja_no i mean more like containers20:22
rm_workactually I really want to experiment with nixOS20:22
rm_workerg20:22
rm_workyeah let me know when nova supports containers properly20:23
rm_work<_<20:23
harlowja_stop depending on nova20:23
johnsomYeah, same.20:23
rm_workwe already build a container image20:23
rm_workbut nothing will run it20:23
* rm_work shrugs20:23
johnsomhttps://www.irccloud.com/pastebin/8oRjO1pj/20:23
harlowja_like literally, stop depending on nova, lol20:24
harlowja_not really kidding, lol20:24
rm_workwhat compute service should we use?20:24
rm_workk8s?20:24
harlowja_u decide20:24
rm_worki literally don't know of a good one20:24
johnsomI am open to it. You going to support the dev work?20:24
harlowja_so, what does octavia need, does it need nova, i doubt it20:24
harlowja_it needs a thing that can provision something right20:25
harlowja_free yourselvessss!!!!!20:25
harlowja_lol20:25
johnsomharlowja No, not at all. We have a compute driver for a reason.20:25
johnsomI have toyed with the idea of a compute driver that just spins up local lxc.20:25
johnsomIt's just someone needs to take the time to make it happen20:26
harlowja_ya, idk, just feels like the whole 'we are bound to waiting to nova to do xyz or ....', just let it go20:26
rm_workyeah we need SOME sort of service to manage the containers20:26
rm_workwe have no attachment to nova really20:26
rm_workthere just isn't anything else either20:26
* harlowja_ is not saying its easy20:26
johnsomWe don't "need" it, but it's nice to have.20:26
rm_worki thought magnum would be good until i realized it doesn't actually do anything useful20:27
rm_workI guess we could start a "container compute management" service20:27
rm_workwhat should we name it?20:27
* johnsom looks at harlowja_ pointy hair type that can allocate someone to "make it so"20:27
harlowja_lol20:27
harlowja_i really don't know what octavia really needs out of nova or ... so i can't really say what would be needed to make it so20:28
harlowja_it just feels odd, that octavia is so bound to openstack20:28
rm_workwe need an API where we can request that a compute resource be created from an image resource20:28
rm_worki was looking at adding drivers for DigitalOcean actually, before they started their LB project20:29
rm_workyou could do it20:29
johnsomThere is a nice little interface that defines what we use: https://github.com/openstack/octavia/blob/master/octavia/compute/compute_base.py20:29
johnsomOk, grabbing lunch20:31
harlowja_ya, anyway, just ranting, like imho, even though i get it; and all, but i'd just start undoing the nova attachments, lol20:31
rm_workwe don't have any nova attachments20:31
harlowja_but thats just my 2 cents, lol20:31
rm_workit's all driverized20:31
rm_workwe just need another service to write a driver for20:32
rm_workthere's zero *hard* linkage to nova20:32
openstackgerritMerged openstack/octavia master: Fix load balancers using IPv6 for the VIP subnet  https://review.openstack.org/54507020:35
*** kobis has quit IRC20:45
*** atoth has joined #openstack-lbaas20:46
johnsomOur Queens IPv6 patch: https://review.openstack.org/#/c/545129/21:25
*** pcaruana has joined #openstack-lbaas21:31
*** gcheresh has quit IRC21:38
*** openstack has joined #openstack-lbaas21:47
*** ChanServ sets mode: +o openstack21:47
*** pcaruana has quit IRC21:48
*** rcernin has joined #openstack-lbaas21:50
*** dougwig has joined #openstack-lbaas21:56
*** slaweq has joined #openstack-lbaas22:42
*** threestrands has joined #openstack-lbaas22:43
*** threestrands has quit IRC22:43
*** threestrands has joined #openstack-lbaas22:43
*** slaweq has quit IRC22:44
*** fnaval has quit IRC23:46
*** fnaval has joined #openstack-lbaas23:57
*** fnaval has quit IRC23:57
*** fnaval has joined #openstack-lbaas23:58

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!