*** dayou_ has joined #openstack-lbaas | 00:34 | |
*** dayou has quit IRC | 00:37 | |
*** fnaval has joined #openstack-lbaas | 00:43 | |
*** threestrands has quit IRC | 00:58 | |
*** threestrands has joined #openstack-lbaas | 01:02 | |
*** threestrands has quit IRC | 01:03 | |
*** threestrands has joined #openstack-lbaas | 01:04 | |
johnsom | Ok, back from a power outage... | 01:04 |
---|---|---|
*** threestrands has quit IRC | 01:11 | |
dayou_ | :-D, you fail over successfully | 01:14 |
johnsom | Yeah, I shut everything down because I could tell it would be a while from the way it went out. It was 1:30 | 01:16 |
dayou_ | A good way to protect stuffs, I just rely on the short circuit protection in the past. | 01:20 |
johnsom | Yeah, I have UPS, but if it seems like a long one I shutdown just so I don't abruptly get shutdown. | 01:21 |
*** numans has quit IRC | 01:24 | |
dayou_ | I remember your isp's network cable was cut down once last year, now it's the power cable or something, seems all kinds of miserable things happended around your area | 01:24 |
*** threestrands has joined #openstack-lbaas | 01:24 | |
*** threestrands has quit IRC | 01:24 | |
*** threestrands has joined #openstack-lbaas | 01:24 | |
dayou_ | Our power supply here is quite stable, I havn't been suffer from the provider failure in the past 10 years | 01:24 |
*** numans has joined #openstack-lbaas | 01:25 | |
*** threestrands has quit IRC | 01:25 | |
*** threestrands has joined #openstack-lbaas | 01:26 | |
dayou_ | I was trying to use wds to setup a load balanced dual wan in the last few days, now seems stable, going to watch it for while | 01:27 |
dayou_ | bricked several routers, going to try to use a usb to ttl cable to bring them back | 01:28 |
johnsom | Yeah, the ISP thing was bad, this wasn't too bad. They typically are pretty rare here. | 01:28 |
dayou_ | Our problem is the internet isp is very slow, it seems they never optimize their networks | 01:28 |
johnsom | Ah, I have ~250Mbps service. It's really pretty good. I'm too cheap to pay for 1gbps... | 01:29 |
johnsom | It's nice stacking, I can usually rebuild devstack in ~10 minutes | 01:29 |
*** numans has quit IRC | 01:29 | |
dayou_ | I have to envy you on that, the best chance I got a rebuild is like several hours | 01:30 |
johnsom | Have you tried the offline options with devstack? There are some settings that force it to use local cached content. It might help | 01:31 |
dayou_ | by setting that in local.conf? | 01:31 |
johnsom | https://www.irccloud.com/pastebin/f2nYtUBK/ | 01:32 |
dayou_ | Not yet, I haven't touch devstack since last time we talked about, now just use tox for development, since don't need to do really testing for the l3 active active yet | 01:32 |
johnsom | Ah, yes. | 01:32 |
johnsom | Well, something to keep in the back of your mind in case you need it. | 01:32 |
dayou_ | Thanks, I never know that kind of option exists without your mention | 01:33 |
dayou_ | I definitely will try that next time I rebuild my devstack | 01:33 |
*** numans has joined #openstack-lbaas | 01:35 | |
*** threestrands has quit IRC | 01:54 | |
*** threestrands has joined #openstack-lbaas | 01:56 | |
johnsom | rm_work Confirmed, it's not working.... | 02:09 |
johnsom | https://github.com/openstack/octavia/blob/master/octavia/controller/healthmanager/health_manager.py#L65 | 02:27 |
johnsom | That must not be working. | 02:27 |
*** fnaval has quit IRC | 02:27 | |
*** annp has joined #openstack-lbaas | 02:39 | |
*** numans has quit IRC | 03:05 | |
*** numans has joined #openstack-lbaas | 03:07 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/python-octaviaclient master: Updated from global requirements https://review.openstack.org/533051 | 03:22 |
*** amotoki has joined #openstack-lbaas | 03:30 | |
*** gans has joined #openstack-lbaas | 03:44 | |
*** yamamoto has joined #openstack-lbaas | 04:03 | |
*** links has joined #openstack-lbaas | 04:31 | |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: ACTIVE-ACTIVE: Initial distributor data model https://review.openstack.org/528850 | 04:34 |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: L3 ACTIVE-ACTIVE Data model impact https://review.openstack.org/524722 | 04:34 |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE ExaBGP rest api driver https://review.openstack.org/527009 | 04:34 |
*** sanfern has joined #openstack-lbaas | 04:34 | |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE ExaBGP rest api driver https://review.openstack.org/527009 | 04:40 |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE ExaBGP rest api driver https://review.openstack.org/527009 | 04:48 |
openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE with exabgp-speaker - Octavia agent https://review.openstack.org/491016 | 06:14 |
*** slaweq has joined #openstack-lbaas | 06:15 | |
*** slaweq has quit IRC | 06:19 | |
*** harlowja has quit IRC | 06:35 | |
*** annp has quit IRC | 06:57 | |
*** annp has joined #openstack-lbaas | 06:58 | |
*** harlowja has joined #openstack-lbaas | 07:10 | |
*** harlowja has quit IRC | 07:10 | |
*** rcernin has quit IRC | 07:28 | |
*** pcaruana has joined #openstack-lbaas | 07:29 | |
*** threestrands has quit IRC | 07:56 | |
*** tesseract has joined #openstack-lbaas | 08:08 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 08:23 | |
*** yamamoto_ has joined #openstack-lbaas | 08:45 | |
*** yamamoto has quit IRC | 08:48 | |
openstackgerrit | huangshan proposed openstack/python-octaviaclient master: Add failover an amphora client support https://review.openstack.org/532424 | 09:25 |
*** vaishali has joined #openstack-lbaas | 09:26 | |
openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE with exabgp-speaker - Octavia agent https://review.openstack.org/491016 | 09:30 |
*** numans has quit IRC | 10:05 | |
*** numans has joined #openstack-lbaas | 10:07 | |
*** yamamoto has joined #openstack-lbaas | 10:07 | |
*** dayou has joined #openstack-lbaas | 10:09 | |
*** yamamoto_ has quit IRC | 10:10 | |
*** dayou_ has quit IRC | 10:11 | |
*** salmankhan has joined #openstack-lbaas | 10:15 | |
*** dayou has quit IRC | 10:25 | |
*** dayou has joined #openstack-lbaas | 10:31 | |
*** yamamoto_ has joined #openstack-lbaas | 10:36 | |
*** yamamot__ has joined #openstack-lbaas | 10:38 | |
*** yamamoto has quit IRC | 10:39 | |
*** dulek has quit IRC | 10:40 | |
*** yamamoto_ has quit IRC | 10:42 | |
*** gans has quit IRC | 10:46 | |
*** sanfern has quit IRC | 10:46 | |
*** gans has joined #openstack-lbaas | 10:48 | |
*** dulek has joined #openstack-lbaas | 10:53 | |
*** AlexeyAbashkin has quit IRC | 10:57 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 11:12 | |
*** yamamot__ has quit IRC | 11:12 | |
*** dayou has quit IRC | 11:14 | |
*** yamamoto has joined #openstack-lbaas | 11:17 | |
*** yamamoto has quit IRC | 11:21 | |
gans | Need help here : http://paste.openstack.org/show/643772/ | 11:24 |
*** gans has quit IRC | 11:25 | |
*** dayou has joined #openstack-lbaas | 11:27 | |
*** yamamoto has joined #openstack-lbaas | 11:43 | |
*** yamamoto has quit IRC | 11:55 | |
*** sanfern has joined #openstack-lbaas | 12:00 | |
*** annp has quit IRC | 12:07 | |
*** yamamoto has joined #openstack-lbaas | 12:11 | |
*** yamamoto has quit IRC | 12:17 | |
*** gans has joined #openstack-lbaas | 12:31 | |
*** yamamoto has joined #openstack-lbaas | 12:42 | |
*** numans has quit IRC | 13:20 | |
*** numans has joined #openstack-lbaas | 13:20 | |
openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE with exabgp-speaker - Octavia agent https://review.openstack.org/491016 | 13:30 |
openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE create distributor flow https://review.openstack.org/527784 | 14:13 |
*** Alex_Staf has joined #openstack-lbaas | 14:27 | |
*** ndanl has joined #openstack-lbaas | 14:29 | |
ndanl | hi | 14:30 |
ndanl | is there any special configuration for DVR with lbaas ? | 14:30 |
ndanl | or for lbaas with dvr :) | 14:30 |
*** ndanl has quit IRC | 14:32 | |
*** numans has quit IRC | 14:32 | |
*** ndanl4 has joined #openstack-lbaas | 14:32 | |
*** mestery has quit IRC | 14:36 | |
*** gans has quit IRC | 14:56 | |
*** mestery has joined #openstack-lbaas | 14:57 | |
*** gans has joined #openstack-lbaas | 15:06 | |
*** Alex_Staf has quit IRC | 15:14 | |
*** sanfern has quit IRC | 15:16 | |
*** sanfern has joined #openstack-lbaas | 15:16 | |
*** AlexeyAbashkin has quit IRC | 15:34 | |
johnsom | ndanl Hi. Until the Pike release there were some bugs in DVR that caused floating IPs to fail when using neutron "allowed address pairs" ports which lbaas uses. Otherwise, not that I know of. | 15:41 |
johnsom | gans Is the code for this issue posted in a patch? | 15:42 |
gans | i have not posted it yet, as it was failing in py27 and py35 checks | 15:43 |
johnsom | Can you post it and just mark it Work In Progress (WIP) with the workflow -1 vote? | 15:44 |
johnsom | That way I can see the code and better help you | 15:44 |
gans | sure, will do that | 15:44 |
johnsom | It seems to be some kind of scoping problem, so should be a quick issue to identify | 15:45 |
gans | when i put the same function with a new name on top of the file, it worked | 15:52 |
*** Alex_Staf has joined #openstack-lbaas | 15:57 | |
johnsom | Looking at that file they are all in the module scope, so load order may matter. I.e. You may need to move the amphore method lower in the file, below the load balancer method definition | 15:58 |
gans | yeah, that will work. we may need to move couple of more dict variables too | 16:00 |
openstackgerrit | Ganpat Agarwal proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE ExaBGP rest api driver https://review.openstack.org/527009 | 16:05 |
openstackgerrit | Andreas Jaeger proposed openstack/octavia-tempest-plugin master: Rename tox_venvlist (1/2) https://review.openstack.org/533222 | 16:06 |
openstackgerrit | Andreas Jaeger proposed openstack/octavia-tempest-plugin master: Rename tox_venvlist (2/2) https://review.openstack.org/533223 | 16:06 |
*** Alex_Staf has quit IRC | 16:11 | |
sanfern | o/ | 16:23 |
*** armax has joined #openstack-lbaas | 16:37 | |
*** sapd_ has quit IRC | 16:37 | |
*** dosaboy has quit IRC | 16:37 | |
*** dosaboy has joined #openstack-lbaas | 16:37 | |
*** sapd_ has joined #openstack-lbaas | 16:37 | |
*** links has quit IRC | 16:38 | |
*** gans has quit IRC | 16:41 | |
*** links has joined #openstack-lbaas | 16:43 | |
-openstackstatus- NOTICE: Zuul has been restarted and lost queue information; changes in progress will need to be rechecked. | 16:45 | |
*** KeithMnemonic has joined #openstack-lbaas | 16:57 | |
*** pcaruana has quit IRC | 17:03 | |
xgerman_ | o/ | 17:05 |
xgerman_ | johnsom: guess we got some traction on https://storyboard.openstack.org/#!/story/2001426 | 17:05 |
johnsom | Yeah, I saw that. Hope we get an update soon | 17:06 |
xgerman_ | +1 | 17:07 |
*** tesseract has quit IRC | 17:12 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 17:12 | |
openstackgerrit | boden proposed openstack/neutron-lbaas master: use api attributes from neutron-lib https://review.openstack.org/533246 | 17:16 |
*** AlexeyAbashkin has quit IRC | 17:17 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 17:21 | |
*** links has quit IRC | 17:24 | |
*** AlexeyAbashkin has quit IRC | 17:25 | |
*** slaweq has joined #openstack-lbaas | 17:28 | |
*** slaweq has quit IRC | 17:29 | |
*** slaweq has joined #openstack-lbaas | 17:30 | |
*** slaweq has quit IRC | 17:30 | |
*** jniesz has joined #openstack-lbaas | 17:33 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 17:42 | |
*** AlexeyAbashkin has quit IRC | 17:44 | |
sanfern | Hi johnsom | 17:59 |
johnsom | Hi sanfern | 18:00 |
johnsom | FYI, I will probably post my base amphora driver patch today. Getting close to done with it. | 18:00 |
sanfern | nice | 18:00 |
sanfern | https://review.openstack.org/#/c/527784/ this patch directly return jul -1 | 18:01 |
sanfern | Zuul | 18:01 |
johnsom | Trying to bring together your needs and the other act/act driver needs | 18:01 |
sanfern | ok Thank you | 18:01 |
sanfern | I have many queries on connecting pieces | 18:01 |
sanfern | I told Jason to schedule meeting to discuss flows | 18:02 |
johnsom | Yeah, well, there are two parts there. Probably some rebases need, plus zuul has been very broken over the last two weeks, so not always reporting correctl | 18:02 |
sanfern | oh ok | 18:02 |
openstackgerrit | Michael Johnson proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE create distributor flow https://review.openstack.org/527784 | 18:03 |
johnsom | Let's try that | 18:03 |
sanfern | ok, To populate speaker and peer tables we need to provide user API's | 18:04 |
johnsom | Umm, I don't think so. There is just one field that I think you need to add to the API right? | 18:05 |
johnsom | Just a subnet or maybe a network ID right? | 18:05 |
sanfern | yes that is fronted subnet for create LB API | 18:06 |
johnsom | Right. There shouldn't be any other API changes right? | 18:06 |
sanfern | but I am not clear how do we add entries into speaker and peer tables | 18:06 |
sanfern | peer contains L3 switch details | 18:07 |
johnsom | Well, I think it is stuff that would come in from the config file and maybe later the flavor. | 18:07 |
johnsom | Users will not know those details right? | 18:07 |
sanfern | yeah atlas for Admins | 18:08 |
johnsom | I should say, users should not know those details | 18:08 |
sanfern | *atleast admins should know | 18:08 |
johnsom | So if I remember correctly, the only ToR switch detail you need is an IP and in your topology all of the switches will have the same config / IPs. | 18:09 |
sanfern | Yes, ToR with details has to be populated into BGP peer | 18:11 |
johnsom | I think most of that is all the same for all of the amps, so for now we can just add them to the octavia.conf, probably in an L3-active-active section | 18:11 |
sanfern | ok then at startup populate into DB | 18:12 |
johnsom | No, when a distributor is created | 18:12 |
sanfern | ok then my distributor creation flow is incomplete | 18:13 |
johnsom | These tables are all for tracking the configuration of the distributor, so they should only be updated when a distributor is created/deleted/updated. | 18:14 |
sanfern | ok I agree | 18:14 |
sanfern | I updated only distributor table need to update speaker and peer tables and DistributorL3BGPPeerRegistration now | 18:16 |
johnsom | Yeah, speaker is probably in the register amp flow I think | 18:17 |
sanfern | ok | 18:17 |
sanfern | Let me modify the patch | 18:17 |
sanfern | Next query - | 18:18 |
sanfern | When do we upload exabgp config and start service during amp registration | 18:19 |
sanfern | or as part of amp creation | 18:20 |
*** Alex_Staf has joined #openstack-lbaas | 18:21 | |
johnsom | Yeah, hmmm, it should be part of the load balancing amphora creation/deletion flow. Probably a subflow needs to be added like we do for network configuration. | 18:23 |
johnsom | Looking at: https://docs.openstack.org/octavia/latest/_images/LoadBalancerFlows-get_create_load_balancer_flow.svg | 18:24 |
johnsom | I think I would put a conditional in the flow creation that replaces "linear_flow.Flow: octavia-post-loadbalancer-amp_association-subflow-octavia-get-vrrp-subflow" if it is an L3 Act/Act LB creation | 18:25 |
sanfern | ok | 18:26 |
*** vaishali has quit IRC | 18:34 | |
*** salmankhan has quit IRC | 18:34 | |
*** slaweq has joined #openstack-lbaas | 18:35 | |
*** AJaeger has joined #openstack-lbaas | 18:36 | |
AJaeger | octavia team, could you review https://review.openstack.org/533222 for the tempest plugin Zuul setup, please? | 18:37 |
johnsom | yep, saw those. Thanks | 18:37 |
*** amotoki has quit IRC | 18:37 | |
johnsom | You have good zuul karma... ha | 18:38 |
AJaeger | ;) | 18:38 |
rm_work | dayou: AdvancedTomato has a really easy Dual-WAN setup if you are using OpenWRT/DD-WRT that is the firmware I generally prefer | 18:39 |
*** amotoki has joined #openstack-lbaas | 18:39 | |
rm_work | johnsom: yeah, not sure why the periodics thing is not doing what it's supposed to. | 18:41 |
rm_work | maybe a bug in periodics itself? | 18:41 |
rm_work | i can see if i can repro it with just a random periodics thing that prints hello | 18:42 |
rm_work | though right now i'm trying to track down an active/standby failover issue | 18:42 |
johnsom | Ok, yeah, you could always ping the master harlowj | 18:43 |
*** vaishali has joined #openstack-lbaas | 18:43 | |
*** slaweq_ has joined #openstack-lbaas | 18:43 | |
*** vaishali has quit IRC | 18:44 | |
*** vaishali has joined #openstack-lbaas | 18:45 | |
*** slaweq__ has joined #openstack-lbaas | 18:46 | |
*** slaweq has quit IRC | 18:47 | |
*** slaweq_ has quit IRC | 18:48 | |
*** Alex_Staf has quit IRC | 18:49 | |
sanfern | johnsom, when will be create distributor sub flow is called ? | 19:12 |
johnsom | LB creation | 19:13 |
sanfern | ok each LB will have distributor. if L3 switch we will have multiple of records of BGPpeers for same switch | 19:15 |
johnsom | Yes, in concept each LB will have a "distributor" if it is in active/active. | 19:16 |
sanfern | each LB will create new entry of peer record and id will be different but details will be same | 19:16 |
sanfern | ok thank you | 19:16 |
*** AlexeyAbashkin has joined #openstack-lbaas | 19:25 | |
*** harlowja has joined #openstack-lbaas | 19:27 | |
openstackgerrit | Santhosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE with exabgp-speaker - Octavia agent https://review.openstack.org/491016 | 19:30 |
*** sanfern has quit IRC | 19:39 | |
openstackgerrit | Merged openstack/octavia-tempest-plugin master: Rename tox_venvlist (1/2) https://review.openstack.org/533222 | 19:40 |
*** AlexeyAbashkin has quit IRC | 19:42 | |
*** harlowja has quit IRC | 20:04 | |
*** longstaff has joined #openstack-lbaas | 20:42 | |
*** openstackstatus has quit IRC | 20:46 | |
*** openstack has joined #openstack-lbaas | 20:50 | |
*** ChanServ sets mode: +o openstack | 20:50 | |
*** AJaeger has left #openstack-lbaas | 21:17 | |
*** harlowja has joined #openstack-lbaas | 21:26 | |
*** salmankhan has joined #openstack-lbaas | 21:29 | |
*** slaweq__ has quit IRC | 22:06 | |
*** slaweq has joined #openstack-lbaas | 22:08 | |
openstackgerrit | Michael Johnson proposed openstack/octavia master: ACTIVE-ACTIVE: Initial distributor data model https://review.openstack.org/528850 | 22:21 |
openstackgerrit | Michael Johnson proposed openstack/octavia master: ACTIVE-ACTIVE: Amphora driver updates https://review.openstack.org/529191 | 22:21 |
openstackgerrit | Michael Johnson proposed openstack/octavia master: ACTIVE-ACTIVE: Initial distributor data model https://review.openstack.org/528850 | 22:24 |
openstackgerrit | Michael Johnson proposed openstack/octavia master: ACTIVE-ACTIVE: Amphora driver updates https://review.openstack.org/529191 | 22:24 |
*** Alex_Staf has joined #openstack-lbaas | 22:25 | |
*** KeithMnemonic has quit IRC | 22:47 | |
openstackgerrit | Bar Elharar proposed openstack/python-octaviaclient master: Remove direct dependency on testtools https://review.openstack.org/533320 | 22:49 |
*** slaweq has quit IRC | 22:49 | |
*** AlexeyAbashkin has joined #openstack-lbaas | 22:54 | |
*** slaweq has joined #openstack-lbaas | 22:58 | |
*** AlexeyAbashkin has quit IRC | 22:59 | |
*** harlowja has quit IRC | 23:01 | |
*** yamamoto has quit IRC | 23:01 | |
*** slaweq has quit IRC | 23:02 | |
*** yamamoto has joined #openstack-lbaas | 23:04 | |
*** eandersson has quit IRC | 23:09 | |
*** yamamoto has quit IRC | 23:09 | |
*** eandersson has joined #openstack-lbaas | 23:09 | |
openstackgerrit | Bar Elharar proposed openstack/python-octaviaclient master: Remove direct dependency on testtools https://review.openstack.org/533320 | 23:17 |
rm_work | johnsom: got it | 23:36 |
rm_work | johnsom: the periodics decorator runs too early, the config isn't REALLY loaded yet | 23:36 |
rm_work | so it always has defaults | 23:36 |
rm_work | submitting a fix (tested in my actual cloud) once the tox stuff passes | 23:36 |
johnsom | Oh, interesting... | 23:37 |
rm_work | there's a standard workaround | 23:38 |
rm_work | it's lame | 23:38 |
rm_work | but | 23:38 |
rm_work | that is what people do | 23:38 |
rm_work | it has something to do with oslo.config laziness | 23:38 |
rm_work | meanwhile, sometimes on amphora failovers, the listeners don't get set up right >_< | 23:39 |
rm_work | time to track that down | 23:39 |
*** longstaff has quit IRC | 23:40 | |
rm_work | ugh pep8 checks take longer than py27 and functional combined | 23:40 |
*** longstaff has joined #openstack-lbaas | 23:41 | |
johnsom | What? Not on my rig | 23:42 |
johnsom | Functional is a beast IMO | 23:42 |
johnsom | The pylint has slowed pep8 for sure, but I also like what it catches. It called me on a totally valid issue today even. | 23:43 |
openstackgerrit | Adam Harwell proposed openstack/octavia master: Healthmanager health_check timer config fix https://review.openstack.org/533326 | 23:43 |
johnsom | That isn't that bad... | 23:45 |
*** longstaff has quit IRC | 23:45 | |
*** longstaff has joined #openstack-lbaas | 23:46 | |
*** longstaff has quit IRC | 23:46 | |
*** longstaff has joined #openstack-lbaas | 23:47 | |
*** longstaff has quit IRC | 23:51 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!