Friday, 2017-12-15

openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561200:02
openstackgerritBar RH proposed openstack/octavia master: Fail-proof VIP deallocation task  https://review.openstack.org/52393100:03
rm_workand then to rebase it again…00:05
openstackgerritAdam Harwell proposed openstack/octavia master: Amphora API Failover call  https://review.openstack.org/52577800:10
openstackgerritAdam Harwell proposed openstack/octavia master: Add unit tests for neutron utils, add model/util for floating_ip  https://review.openstack.org/52535300:10
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561200:12
rm_workk00:12
*** bar_ has quit IRC00:13
*** SumitNaiksatam has quit IRC00:15
*** sshank has quit IRC00:16
*** bar_ has joined #openstack-lbaas00:17
*** bar_ has quit IRC00:28
*** fnaval has quit IRC00:29
*** bar_ has joined #openstack-lbaas00:32
*** bar_ has quit IRC00:34
*** sshank has joined #openstack-lbaas00:37
*** AlexeyAbashkin has joined #openstack-lbaas00:45
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561200:50
*** AlexeyAbashkin has quit IRC00:50
*** Swami has quit IRC00:58
dayourm_work I used ubuntu, I have no dhcp server also, but if cloud init networking is enabled, it will try to get an ip adress with dhcp which would cause the boot up time to last for about 3 minutes.01:06
johnsomNo01:06
dayourm_work, also I use ipv6 only01:06
rm_workcan you show me the logs for yours?01:06
johnsomThis is not the behavior01:06
johnsomPlease provide the logs I requested.  If you don't have it in the scroll back I can re-paste01:06
rm_worklet me see if it's the same as mine01:07
rm_workif you can, i need to see syslog and the cloud-init log01:07
johnsomdayou That is not how it works for me.  Please send the /etc/network/interfaces, /etc/network/interfaces.d/*, /etc/netns/amphora-haproxy/network/interfaces, /etc/netns/amphora-haproxy/network/interfaces.d/* , /var/log/cloud-init*, your config drive network_data.json file01:07
dayouMight not now, my env was destroyed several weeks ago due to the development of dashbard.01:08
dayouBut maybe you can put your files some where, then I try to get mine and we compare there01:09
dayouIf we were sit together it would make things much more eaiser01:10
rm_workwell, we really just need to see your logs01:10
rm_workand we should be able to understand what's happening and tell you how to fix it01:10
rm_workso you may need to recreate your env01:10
johnsomYeah, I have run this many times over the last few days, there is zero delay, no DHCP, and the configs are all loaded as I showed you yesterday.01:12
dayouAlso, in my env, I had the image with cloud init disabled01:14
johnsomWell, nothing will work with cloud init disabled.  The amp won't run at all01:15
dayouSo what you need is to reproduce the 3 minutes delay thing or?01:15
dayouWith an slaac ipv6 six and controller running in ipv6, it would work01:15
johnsomYeah, setup an amphora that reproduces the delay you saw then collect those logs01:16
johnsomsyslog woudl be helpful too01:16
dayouLet me see if I can find some tcpdump files01:16
johnsomIf you disable cloud init all of the files required for octavia will not load on the amphora, it will fail01:17
johnsomtcpdump will not help me01:17
johnsomI don't deny you are seeing that, I need to know what configuration is causing it since it should not happen and does not happen for me.01:18
*** yamamoto has quit IRC01:18
dayouIt requires me to destroy my current environment and might took a few days to rebuild the old one01:18
dayouSo that's why I am trying to avoid it.01:18
dayouCan you send me your those files first?01:18
rm_workyou can't reproduce in your current env?01:18
johnsomUnderstand.  It's up to you when/if01:18
dayouIt's a complicated one01:19
dayouThe external provider network is simulated in quagga01:19
*** sshank has quit IRC01:19
dayouand devstack is running on one of the servers01:19
dayouand devstack in running in linux bridge mode01:20
johnsomYeah, linux bridge doesn't matter here, it works with both OVS and LB01:21
dayouJust curious, could you post files? Let me take a look and then I will try to rebuild my env here01:23
rm_workdayou: i might recommend spending a little time writing up some ansible or something to create your env01:23
rm_workbecause a few DAYS to remake it is ... :/01:23
johnsomYeah, I rebuild devstacks in about 10 minutes01:24
dayourm_work, yes, that's something in my head,01:24
johnsomDid you not keep the files I sent from last night?01:24
dayoujohnsom, normal devstack will take me only several hours since I have a worse network condition01:25
dayoujohnsom, last night you sent only the cloud init metadata but not the complete files as you listed here01:25
dayoujohnsom, if you come to my country you'll understand the network difference, sometimes you have retry building several times due to network failures01:26
johnsomhttps://www.irccloud.com/pastebin/yjfwHgc8/01:27
*** yamamoto has joined #openstack-lbaas01:27
johnsomThat was from last night, it is the only interface file I have01:27
johnsomThis is the config drive data for cloud-init to use:01:27
johnsomhttps://www.irccloud.com/pastebin/VirRVS17/01:27
rm_workdayou: yeah i am somewhat familiar with that issue :P01:28
johnsomdayou There is an offline mode for devstack that tries to use cached versions.  You might look into that.01:28
rm_worki found it useful to make proxy caches for my devstack vms01:28
dayourm_work, that's cool01:29
dayoujohnsom, that's interesting, you got a static ip address assignment01:30
johnsomhttps://www.irccloud.com/pastebin/7ISNxNZl/01:30
johnsomThis is the script I gave to reproduce this01:30
dayoujohnsom, I think I know the difference, you are using openstack to manage do all the ip address allocation01:31
johnsomdayou yes, since neutron knows the prefix and the MAC of the port it will put the static in the config automatically.01:31
dayouSo it knows the ipv6 address in advance and can feed a static one01:31
dayouand in my environment, the network allocation is done by the external network simulated by quagga01:31
johnsomdayou It has to, it's a neutron network01:31
*** yamamoto has quit IRC01:32
dayoujohnsom, if you look at the ipv6 documentention, they distinguished these situations clearly01:32
johnsomdayou I have to go now.  It's dinner time.  I can only help you solve it if we get those logs.01:33
dayoujohnsom, enjoy your dinner, anyway I think the puzzle has already been resolved, have to do breakfirst also01:35
*** oomichi_afk is now known as oomichi01:36
*** sapd has joined #openstack-lbaas01:45
*** sapd has quit IRC01:55
*** yamamoto has joined #openstack-lbaas01:57
*** sapd has joined #openstack-lbaas01:57
*** sticker has quit IRC01:58
*** yamamoto has quit IRC02:02
*** yamamoto has joined #openstack-lbaas02:07
*** sanfern has joined #openstack-lbaas02:09
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561202:24
*** yamamoto has quit IRC02:29
*** yamamoto has joined #openstack-lbaas02:30
dayouN/S  slaac  Off  1,0,0  Guest instance obtains IPv6 address from non-OpenStack router using SLAAC.02:32
dayouslaac  slaac  1,0,0  Off  Guest instance obtains IPv6 address from OpenStack managed radvd using SLAAC.02:32
dayouhttps://docs.openstack.org/ocata/networking-guide/config-ipv6.html02:32
dayouSo my network is the first one and yours is the second one, actually in our labs, we also using the first one, so the same trick was used there02:33
*** annp has joined #openstack-lbaas02:40
*** harlowja has joined #openstack-lbaas02:41
*** harlowja has quit IRC02:42
*** harlowja has joined #openstack-lbaas02:43
*** atoth has quit IRC02:43
*** atoth has joined #openstack-lbaas02:45
*** sanfern has quit IRC02:47
*** harlowja has quit IRC03:15
*** jappleii__ has joined #openstack-lbaas03:17
*** jappleii__ has quit IRC03:19
*** jappleii__ has joined #openstack-lbaas03:19
*** threestrands_ has quit IRC03:20
*** SumitNaiksatam has joined #openstack-lbaas03:30
*** tongl has quit IRC03:31
*** armax has quit IRC03:47
*** jappleii__ has quit IRC03:56
johnsomYeah, both work fine here03:57
*** armax has joined #openstack-lbaas04:05
dayouOn proper amphora image with patch included, it should works fine04:06
dayouJohnsom, for https://review.openstack.org/#/c/521138/, is the final decision to keep both vrrp_ and frontend_ prefix for now?04:10
*** armax has quit IRC04:17
*** dmellado has quit IRC04:27
*** dmellado has joined #openstack-lbaas04:35
*** harlowja has joined #openstack-lbaas04:36
*** harlowja has quit IRC04:45
*** harlowja has joined #openstack-lbaas05:21
*** harlowja has quit IRC05:22
*** harlowja has joined #openstack-lbaas05:24
*** harlowja has quit IRC05:29
*** harlowja has joined #openstack-lbaas05:38
*** amotoki has joined #openstack-lbaas05:39
*** eN_Guruprasad_Rn has joined #openstack-lbaas05:51
*** sapd_ has joined #openstack-lbaas06:01
*** sapd has quit IRC06:01
*** links has joined #openstack-lbaas06:18
AJaegerdayou: everything clear on my changes now?06:23
dayouAjaeger: I am pretty sure it works now, :-)06:24
dayouAjaeger: it took a recheck to get those npm lint and test back06:24
AJaegerso am I ;) The whole goal of the exercise was to move those to project-config - and check that the new native jobs work fine so that I can move the right job06:24
dayouAjaeger: I see you removed tox_install.sh from other projects, any plan to remove it for us?06:25
AJaegerYeah, recheck was needed to show that they are there after the project-config change merged06:25
AJaegerdayou: I would love to - but don't know yet how ;(06:25
AJaegeryour tox_install installs some packages, and we need a plan on how to do this properly first06:25
AJaegerthe ones I removed were the easier ones, now need some time to taggle the more complex ones06:26
dayouCool, what's already done is hard enough for most people, I think, looking forward to see it, that's something really tough06:27
AJaegerso do I ;)06:28
*** harlowja has quit IRC06:32
*** rcernin has quit IRC06:47
openstackgerritOpenStack Proposal Bot proposed openstack/octavia-dashboard master: Imported Translations from Zanata  https://review.openstack.org/52816806:49
openstackgerritZhaoBo proposed openstack/octavia master: Support UDP load balancing  https://review.openstack.org/50360606:51
*** sri_ has quit IRC06:52
*** sshank has joined #openstack-lbaas07:04
openstackgerritMichael Johnson proposed openstack/octavia master: Fix PING health monitor for old amphora  https://review.openstack.org/52817107:08
*** sshank has quit IRC07:22
openstackgerritNir Magnezi proposed openstack/octavia master: Use get_rpc_transport instead of get_transport  https://review.openstack.org/47885907:35
*** AlexeyAbashkin has joined #openstack-lbaas07:59
*** dayou has quit IRC08:01
*** dayou has joined #openstack-lbaas08:32
openstackgerritZhaoBo proposed openstack/octavia master: Support UDP load balancing  https://review.openstack.org/50360608:58
*** gcheresh has joined #openstack-lbaas09:41
*** annp has quit IRC09:52
*** gcheresh has quit IRC09:54
dayourm_work, johnsom, Sigh, I deleted my old env and tried to bring back my old one, but failed to do. So I can't provide any logs for now.10:02
*** Alex_Staf has joined #openstack-lbaas10:34
*** dmellado has quit IRC10:39
*** dmellado has joined #openstack-lbaas10:44
*** salmankhan has joined #openstack-lbaas10:47
*** Alex_Staf has quit IRC11:00
*** yuanying has quit IRC11:00
*** salmankhan has quit IRC11:04
*** salmankhan has joined #openstack-lbaas11:05
*** sapd_ has quit IRC11:28
*** sapd_ has joined #openstack-lbaas11:28
*** yamamoto has quit IRC11:35
*** yamamoto has joined #openstack-lbaas12:09
*** sanfern has joined #openstack-lbaas12:11
*** dayou has quit IRC12:15
*** eN_Guruprasad_Rn has quit IRC12:17
*** eN_Guruprasad_Rn has joined #openstack-lbaas12:17
*** dayou has joined #openstack-lbaas12:41
*** yamamoto has quit IRC12:48
*** bbzhao has quit IRC12:55
*** bbzhao has joined #openstack-lbaas12:56
*** dayou has quit IRC12:56
*** dayou has joined #openstack-lbaas12:57
*** dayou has quit IRC12:57
*** dayou has joined #openstack-lbaas12:57
*** reedip has quit IRC13:07
*** salmankhan has quit IRC13:12
*** salmankhan has joined #openstack-lbaas13:13
*** atoth has quit IRC13:15
*** atoth has joined #openstack-lbaas13:16
*** reedip has joined #openstack-lbaas13:20
*** salmankhan has quit IRC13:27
*** salmankhan has joined #openstack-lbaas13:34
*** sanfern has quit IRC13:37
*** sanfern has joined #openstack-lbaas13:38
*** yamamoto has joined #openstack-lbaas13:39
openstackgerritMerged openstack/neutron-lbaas master: Fix the deprecated usage of "get_transport"  https://review.openstack.org/52560413:52
*** links has quit IRC14:31
*** gcheresh has joined #openstack-lbaas14:46
*** openstackgerrit has quit IRC14:48
*** eN_Guruprasad_Rn has quit IRC14:52
*** gcheresh has quit IRC14:56
*** rcernin has joined #openstack-lbaas14:57
*** harlowja has joined #openstack-lbaas15:11
*** harlowja has quit IRC15:13
*** armax has joined #openstack-lbaas15:31
*** rcernin has quit IRC15:41
*** Alex_Staf has joined #openstack-lbaas16:00
*** yamamoto has quit IRC16:03
sanfernhi johnsom ,16:21
johnsomsanfern Hi16:27
johnsomBTW, I plan to work on the data model patch today.16:27
johnsomI just have an API bug to address first this morning16:28
*** gcheresh has joined #openstack-lbaas16:32
*** salmankhan has quit IRC16:32
*** AlexeyAbashkin has quit IRC16:41
*** gcheresh has quit IRC16:44
*** salmankhan has joined #openstack-lbaas16:50
sanfernok thanks johnsom16:58
sanfernhttp://paste.openstack.org/show/BUKRtZvshLtqXWNhAMWb/16:58
sanfernDo you know this error ?16:58
*** yamamoto has joined #openstack-lbaas17:03
*** sapd_ has quit IRC17:06
johnsomYeah, it's a style check rule saying there are too many positional arguments on that method17:06
*** salmankhan has quit IRC17:08
*** yamamoto has quit IRC17:11
sanfern:(17:13
johnsomJust change some to named17:13
sanfernhttp://logs.openstack.org/84/527784/2/check/openstack-tox-pep8/5ee810f/job-output.txt.gz17:14
sanferni have fixed common spelling one17:15
*** Alex_Staf has quit IRC17:21
johnsomOk, I commented on the patch17:22
sanfernThank you17:24
johnsomSee if that works, I think that is the problem17:24
*** openstackgerrit has joined #openstack-lbaas17:34
openstackgerritMerged openstack/neutron-lbaas-dashboard master: Imported Translations from Zanata  https://review.openstack.org/52759117:34
sanfernDo we need to have common folder under distributor and move delete there17:34
sanfernbasically markfor deletion in17:35
sanfernHow to fix vrrp_ rename patch17:36
johnsomYes, we probably need a common folder for the distributor driver since some things, like marking deleted are the same in many of the drivers.17:38
sanfernok17:38
johnsomThe rename patch, we are going to extend it to maintain the old column name and the new one so we can do an upgrade without downtime17:38
johnsomThen later we will remove the old column17:39
johnsomThat is what was decided at the meeting17:39
sanferntoo-many-functional args didn't get fix with that change17:39
sanfernok17:39
johnsomHmm, bummer, let me look again17:40
sanfernok :)17:40
johnsomIs it still line 44?17:40
sanfernyes - E: 44,12: Too many positional arguments for constructor call (too-many-function-args)17:41
johnsomOk, I think it is that the name needs to be name=17:46
johnsomCommented again17:46
sanfernthanks17:47
sanfernjohnsom, solved.17:57
johnsomNice17:57
openstackgerritSanthosh Fernandes proposed openstack/octavia master: [WIP] ACTIVE-ACTIVE create distributor flow  https://review.openstack.org/52778417:59
*** sanfern has quit IRC18:01
*** salmankhan has joined #openstack-lbaas18:09
*** sshank has joined #openstack-lbaas18:13
*** AlexeyAbashkin has joined #openstack-lbaas18:16
*** AlexeyAbashkin has quit IRC18:21
*** sshank has quit IRC18:22
*** AlexeyAbashkin has joined #openstack-lbaas18:22
*** AlexeyAbashkin has quit IRC18:26
*** sshank has joined #openstack-lbaas18:45
*** salmankhan has quit IRC18:49
*** salmankhan has joined #openstack-lbaas18:52
*** salmankhan has quit IRC19:00
beaglesnoticing that cert_manager has defaulted to barbican_cert_manager for some time now, does that imply a hard dependency on barbican19:00
beaglesand what functionality is lost in the absence of barbican?19:01
johnsomNo19:01
johnsomWell, barbican is use as the securty TLS certificate store if you are enabling TLS offload19:01
beaglesfigured .. as we are able to create load balancers without it :)19:01
johnsomYeah, that is all it is used for.  There are two, soon, three options: local filesystem (not very secure, handy for testing), barbican, and soon castellan19:02
beaglesah19:02
johnsomOr you can just disable TLS offload options19:02
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561219:03
beaglesjohnsom, and configs like ca_certificate, etc. are for the local certificate generator correct?19:04
openstackgerritMichael Johnson proposed openstack/octavia master: Improve user error messages for duplicate objects  https://review.openstack.org/52837019:05
* beagles looks at code for hisself19:05
johnsombeagles Some of that is actually for the certificates we use to secure communications with the amphora agent.19:05
johnsomSo, internal use, with no dependency on barbican19:06
beaglesjohnsom, right.. some backstory, I'm working on finalizing integration with tripleo and we are getting questions about using FreeIPA/certmonger to provide the values for the ca_* configuration options and I'm trying to sort out if these will integrate19:08
beaglesjohnsom, previously we avoided it for other reasons, but now its a sticking point19:08
rm_workbeagles: yeah so there's two things19:09
beaglesjohnsom, and I'm just trying to sort out the details19:09
rm_workbeagles: cert GENERATOR (used internally) and cert MANAGER (used externally for TLS)19:09
rm_workyou can replace the cert GENERATOR with certmonger for certain19:09
rm_workor there's a driver for Anchor I believe19:09
rm_workas for cert MANAGER, there's a patch up to change that functionality a lot, that should merge this cucle19:09
rm_work*cycle19:09
rm_workwhich would enable Castellan, which makes using certmonger easy (if users have a way to store anything in there)19:10
rm_workbeagles: talk to alee19:10
rm_workI've worked with him a lot on this stuff19:10
beaglesrm_work, ack19:10
johnsomYeah, when messing with the cert generator stuff, note that the housekeeping process has auto rotation code for those, so you likely want to test that too19:12
johnsomJust to make sure a renew works with a different cert gen19:12
johnsomrm_work FYI, you are probably interested in those two patches I put up.19:13
rm_workk19:13
johnsomOne fixes your haproxy issue on centos, the other fixes duplicate health monitors from 500ing and gives the user a useful error19:13
rm_worksuch found19:16
beaglesrm_work, to be clear to work with certmonger, we'd need to write a new certificate generator correct? which is a relatively simple adapter that implements the interface defined by octavia.certificates.generator.CertGenerator ?19:17
rm_workyes19:19
beaglesrm_work cool19:19
beaglesthanks19:19
AJaegerany cores around to review https://review.openstack.org/527863 and https://review.openstack.org/527861 , please? That gets rid of now duplicated jobs19:22
johnsomrm_work Yeah, can you poke those?  I have already +219:23
rm_workk19:23
johnsomI thought Nir would get them last night, but I guess he didn't get to it19:23
AJaegerthanks, johnsom and rm_work19:24
openstackgerritMerged openstack/octavia-dashboard master: Use standard nodejs4 tests  https://review.openstack.org/52786119:34
openstackgerritMerged openstack/neutron-lbaas-dashboard master: Use standard nodejs4 tests  https://review.openstack.org/52786319:38
AJaegergreat, thanks^19:39
* AJaeger signs off here19:39
*** AJaeger has left #openstack-lbaas19:39
openstackgerritMerged openstack/octavia-dashboard master: Imported Translations from Zanata  https://review.openstack.org/52816819:49
*** sshank has quit IRC19:54
*** gcheresh has joined #openstack-lbaas20:09
*** gcheresh has quit IRC20:43
*** sshank has joined #openstack-lbaas20:46
*** openstackgerrit has quit IRC20:48
*** openstackgerrit has joined #openstack-lbaas20:50
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561220:50
*** sshank has quit IRC20:51
*** gcheresh has joined #openstack-lbaas20:52
*** sshank has joined #openstack-lbaas21:01
*** gcheresh has quit IRC21:01
*** openstackgerrit has quit IRC21:18
*** rcernin has joined #openstack-lbaas21:36
*** openstackgerrit has joined #openstack-lbaas21:47
openstackgerritOpenStack Proposal Bot proposed openstack/neutron-lbaas master: Updated from global requirements  https://review.openstack.org/52841921:47
*** openstackgerrit has quit IRC21:48
*** rm_mobile has joined #openstack-lbaas22:13
*** sshank has quit IRC22:25
*** AlexeyAbashkin has joined #openstack-lbaas22:42
*** AlexeyAbashkin has quit IRC22:47
*** sshank has joined #openstack-lbaas22:49
*** bar_ has joined #openstack-lbaas23:02
*** sshank has quit IRC23:23
*** sshank has joined #openstack-lbaas23:27
*** sshank has quit IRC23:28
*** openstackgerrit has joined #openstack-lbaas23:56
openstackgerritAdam Harwell proposed openstack/octavia master: WIP: Floating IP Network Driver (spans L3s)  https://review.openstack.org/43561223:56

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