Monday, 2014-12-22

*** ajmiller_ has joined #openstack-lbaas00:27
*** ajmiller has quit IRC00:30
*** hichtakk has joined #openstack-lbaas00:32
*** rm_work is now known as rm_work|away00:34
*** hichtakk has quit IRC00:41
openstackgerritAl Miller proposed stackforge/octavia: Interface specification for housekeeping manager
*** TrevorV_ has joined #openstack-lbaas01:14
*** TrevorV_ has quit IRC01:24
*** mlavalle has quit IRC01:48
*** ajmiller__ has joined #openstack-lbaas02:18
*** ajmiller_ has quit IRC02:22
*** ajmiller has joined #openstack-lbaas02:37
*** ajmiller__ has quit IRC02:39
*** hichtakk has joined #openstack-lbaas02:42
*** rm_work|away is now known as rm_work02:50
*** hichtakk has quit IRC02:53
*** woodster_ has quit IRC03:50
*** hichtakk has joined #openstack-lbaas03:54
*** hichtakk has quit IRC03:59
*** hichtakk has joined #openstack-lbaas04:07
*** rohara1 has quit IRC04:16
*** rohara1 has joined #openstack-lbaas04:17
*** hichtakk has quit IRC04:30
*** hichtakk has joined #openstack-lbaas04:30
*** hichtakk has quit IRC04:35
*** TrevorV has quit IRC05:05
*** TrevorV has joined #openstack-lbaas05:05
*** redrobot has quit IRC05:09
*** redrobot has joined #openstack-lbaas05:12
*** redrobot is now known as Guest8746305:12
*** blogan_ has quit IRC05:26
*** TrevorV_ has joined #openstack-lbaas05:30
*** blogan_ has joined #openstack-lbaas05:30
*** TrevorV_ has quit IRC05:35
*** TrevorV_ has joined #openstack-lbaas05:37
*** hichtakk has joined #openstack-lbaas05:39
*** TrevorV_ has quit IRC05:40
*** TrevorV_ has joined #openstack-lbaas05:40
*** TrevorV_ has quit IRC05:41
*** blogan_ has quit IRC06:21
*** blogan_ has joined #openstack-lbaas06:34
*** blogan_ has quit IRC06:51
*** apuimedo has joined #openstack-lbaas07:13
*** kobis has joined #openstack-lbaas07:19
*** hichtakk has quit IRC07:42
*** sbfox has joined #openstack-lbaas08:11
*** kobis1 has joined #openstack-lbaas08:14
*** sbfox has quit IRC08:14
*** sbfox has joined #openstack-lbaas08:15
*** kobis has quit IRC08:16
*** kobis has joined #openstack-lbaas08:19
*** kobis1 has quit IRC08:20
*** sbfox has quit IRC08:23
*** dorry has quit IRC08:23
*** kobis has quit IRC08:30
openstackgerritLi Yingjun proposed openstack/neutron-lbaas: Persistence lbaas http cookie
*** rm_work is now known as rm_work|away08:53
*** rm_work|away is now known as rm_work09:00
*** kobis has joined #openstack-lbaas09:41
*** rm_work is now known as rm_work|away09:44
*** Putns has joined #openstack-lbaas10:03
openstackgerritIhar Hrachyshka proposed openstack/neutron-lbaas: Moved lbaas-haproxy.filters from main neutron repo
openstackgerritIhar Hrachyshka proposed openstack/neutron-lbaas: Cleaned up requirements.txt
*** SumitNaiksatam has quit IRC12:16
*** SumitNaiksatam has joined #openstack-lbaas12:20
*** pc_m has joined #openstack-lbaas12:53
*** ajmiller has quit IRC13:26
*** 16WAAORL2 has quit IRC13:40
*** enikanorov_ has joined #openstack-lbaas13:40
*** TrevorV_ has joined #openstack-lbaas14:15
*** zeus has joined #openstack-lbaas14:27
*** woodster_ has joined #openstack-lbaas14:43
*** apuimedo has quit IRC15:04
sbalukoffMornin' folks!15:17
TrevorV_(dude says at most people's 7 am... :P   )15:26
sbalukoffHeh! Well, I'm here right now because of you, dude. ;)15:26
*** ajmiller has joined #openstack-lbaas16:02
a2hillWhy is TLS termination so low on the list when it should be part of .5?16:07
sbalukoffBecause we needed to prioritize things, and most everything else higher on the list is "core" functionality without which Octavia can't function.16:10
sbalukoffat all.16:10
sbalukoffHaving said this, if you want to write the TLS termination handling code, I won't reject a patch, eh. :)16:10
*** enikanorov_ has quit IRC16:27
*** xgerman has joined #openstack-lbaas16:29
sbalukoffIs not resolving for anyone else here?16:30
a2hillloading for me sbalukoff16:38
*** kobis has quit IRC16:43
*** sbfox has joined #openstack-lbaas16:43
*** xgerman has quit IRC16:48
*** xgerman has joined #openstack-lbaas16:52
*** mlavalle has joined #openstack-lbaas16:54
xgermansorry, for the disconnect I had to re-jigger my VPN...16:56
xgerman(I don't want the irc traffic run through HP)16:56
*** openstack has joined #openstack-lbaas17:01
*** fnaval has joined #openstack-lbaas17:18
*** ajmiller_ has joined #openstack-lbaas17:36
*** ajmiller has quit IRC17:39
*** sbfox has quit IRC17:49
*** xgerman has quit IRC17:53
*** pcaruana has quit IRC17:53
*** xgerman has joined #openstack-lbaas17:58
a2hillWhy is the driver only taking listener/vip? why not the whole LB config?18:05
*** hichtakk has joined #openstack-lbaas18:08
dougwiga2hill: i think i just heard the worm can busting open.18:15
a2hillIm sure this has been argued…cough… discussed. Im just so outta the loop and trying to catch back up18:16
a2hillsrry if im brining up old news :/18:16
dougwigit's a topic that keeps cropping up, because it has a lot of merit.18:17
a2hillI agree, theres things i would like to utilize that lives on the lb object itself in the jinja configs18:17
*** SumitNaiksatam has quit IRC18:19
dougwigif you have a listener, you can just do listener.loadbalancer to get to the LB18:20
a2hillthat works then, assuming its lazy loaded when it gets passed to me?18:20
dougwigi think it might get eager loaded, since the models are stripped of their sqlalchemyness.18:21
openstackgerritAl Miller proposed stackforge/octavia: Interface specification for housekeeping manager
a2hiller yea, eager i meant ><18:23
a2hillIn that case, i have no further questions18:23
*** sbfox has joined #openstack-lbaas18:31
*** hichtakk has quit IRC18:32
openstackgerritGerman Eichberger proposed stackforge/octavia: updates the amphora driver interface spec
openstackgerritAl Miller proposed stackforge/octavia: Interface specification for housekeeping manager
*** markmcclain has quit IRC18:39
*** SumitNaiksatam has joined #openstack-lbaas18:52
a2hillSo the 'driver returns a sub-flow that we kick off…', this means the caller of the driver is actually starting the flow/task/action?19:15
*** sbfox has quit IRC19:16
sbalukoffa2hill: Yes, I think that's how it's going to work. German knows more about taskflow than me, though.19:16
a2hillWho calls the driver then? And just to clear it up for me, this means whomever that caller is is the one actually maintaining the flows that the driver builds?19:17
*** crc32 has joined #openstack-lbaas19:17
a2hillso, besides building tasks/configs the driver does nothing in other words19:17
sbalukoffThe driver also receives events generated on the amphorae as well.19:20
sbalukoff(heartbeats, and status change notifications, that sort of thing.)19:20
a2hillwell yea, just thought more logic was in the driver19:20
a2hillSorry for lame questions19:20
sbalukoffBy returning the task flow, the driver is actually supplying the logic that the worker which calls it is going to use.19:21
xgermanyes, so to give you ane xample19:21
xgermanthe control system might build a flow like this:19:21
a2hillBut running the tasks isnt part of the driver was my point, or what i was trying to understand19:21
xgerman1) Create nov vm19:22
xgerman2) Call driver subflow to prepare vm19:22
xgermanso in 2) the driver would run code to do dtuff there19:22
a2hillto build the flow?19:22
xgermanwell the driver builds the flow but it build it with tasks from the driver19:22
xgermanand the controller builds the bigger flow19:23
a2hilli see19:23
sbalukoffxgerman: Did you mean to day "the controller builds the flow, but it builds it with tasks from the driver"?19:23
a2hillthat makes sense worded like that for me19:23
xgermanyes, sbalukoff +119:24
sbalukoffAnd by "controller", I assume you more specifically mean one of the managers we talked about last week.19:24
xgermanwe just wanted to gibe the driver the flexibility to make their own subflow19:24
xgermanin case the tasks need to be arrnged some way19:24
xgermanor there is more than 1 task...19:25
a2hillnice, clears that up for me. Thanks ;)19:25
xgermanawesome -- it's pretty much modeled after what i think the ML2 driver does19:26
xgermanwhich might not be what they do in reality :-)19:26
a2hillso, just to be more certain/run it by you guys. The haproxy driver (ignoring things like stats for now) will build the jinja configs and tasks that should call the amphora api, then return those tasks and the 'controller' and specific manager will take care of it from there19:27
xgermanyep, exactly19:28
a2hillkk, cool19:28
xgermanand tasks means that you can do a rollback + some sophisticated retry rules19:28
xgermanthough in reality we likely will just delete the whole vm19:28
a2hillyea, nice. me likey. Just wasnt hip to what was actually happening >< playing catch up19:29
xgermanyep, after we looked at what were basically state diagrams we decided to use some tool :-)19:29
sbalukoffxgerman: I think it's good to have the flexibility.19:29
xgermanagreed - so "in reality" means 0.519:30
*** hichtakk has joined #openstack-lbaas19:35
*** markmcclain has joined #openstack-lbaas19:53
*** crc32 has quit IRC19:57
*** hichtakk has quit IRC20:07
*** hichtakk has joined #openstack-lbaas20:07
*** crc32 has joined #openstack-lbaas20:08
openstackgerritCarlos Garza proposed openstack/neutron-lbaas: Partially-Implements: blueprint lbaas-tls Change-Id: I0805b6dae0dcc311b176e3017f0218e30b7d81f5
*** rm_work|away is now known as rm_work20:13
*** hichtakk has quit IRC20:17
*** rm_work is now known as rm_work|away20:20
* crc32 slaps rm_work around a bit with a large trout20:20
*** hichtakk has joined #openstack-lbaas20:25
*** enikanorov_ has joined #openstack-lbaas20:49
*** enikanorov has quit IRC20:51
*** crc32 has quit IRC21:01
*** crc32 has joined #openstack-lbaas21:06
*** busterswt has joined #openstack-lbaas21:07
*** hichtakk has quit IRC21:11
*** hichtakk has joined #openstack-lbaas21:11
*** busterswt has quit IRC21:16
*** ajmiller_ has quit IRC21:18
*** ajmiller has joined #openstack-lbaas21:19
*** zeus has quit IRC21:21
openstackgerritAl Miller proposed stackforge/octavia: Interface specification for housekeeping manager
*** hichtakk has quit IRC21:53
*** hichtakk has joined #openstack-lbaas21:53
*** rm_work|away is now known as rm_work22:01
*** rm_work is now known as rm_work|away22:06
*** pc_m has quit IRC22:07
*** rm_work|away is now known as rm_work22:07
crc32looking for something to work on. :/22:11
sbalukoffcrc32: There are a few things that need to happen on the amphora when it's first spun up. This includes starting the API and the heartbeat daemon. Would you like to work on the code which does that (and gets built into the amphora image)?22:15
sbalukoffFor now, some things will be place-holder "#TODO"s in the code.22:15
sbalukoffAlso, when building the image, I would like to get the list of locally installed packages and their versions dumped to the image somewhere in json format so we don't have to dump this everytime someone asks for the list of packages when they hit the /info URL on the amphora API22:18
sbalukoffI also know that Doug is looking for someone to work on the LBaaSv2 reference implementation.22:19
crc32sbalukoff: so I could right something that takes the output of "dpkg -l" and converts it to Json?22:20
crc32err write?22:20
crc32Or are we talking about python packages?22:21
sbalukoffcrc32: We're talking about dpkg -l22:23
sbalukoffAnd dumps it somewhere on the amphora image.22:23
sbalukoff(As part of the amphora image build process work that johnsom did)22:23
crc32Where is this code at at? I'd like to chip away at it.22:25
openstackgerritTrevor Vardeman proposed stackforge/octavia: Creation of Octavia API Documentation
*** TrevorV_ has quit IRC22:30
sbalukoffcrc32: It's in the repo, under diskimage-create22:30
crc32I guess I could also do blind research into the sever client x509 extensions you want to inject into your X509s if your still interested22:31
crc32That shell scripting foo looks intimidating.  Even scarier since it looks to be poly glot for RHEL CENTOS as well as Debian.22:37
sbalukoffcrc32: That works.22:41
crc32what works?22:41
sbalukoffAnd actually, yes, that's a really good idea and needs to be done. (Looking into the x509 client / server certificate extensions and how we insert these into the certs we generate.)22:42
crc32ok thanks. I'll be researching that since its in my domain of knowledge till my team returns from Limbo.22:44
*** Guest87463 is now known as redrobot_away22:48
dougwigcrc32: on the lbaas side, the agent-ified driver, or splitting tempest tests.22:51
crc32dougwig: The TLS commit keeps breaking cause devstack or something is trying to install "prettytable>0.7". Why is this failing. How do I resolve this. Its the same Error every hour.22:53
dougwigcrc32: all jenkins jobs are failing. infra is working on it.22:54
crc32sbalukoff: Do you have a client cert I could parse up. I'm tryingt to figure out what the OID of the extension is that uses it.22:59
sbalukoffcrc32: If you can give me a few minutes, I can generate one for you. (I need to go AFK for about 20 minutes)23:01
crc32I suspect the OID is but I'd like to see a X509 Client cert that I know works. Ok.23:01
crc32ok cool.23:01
crc32let me know how your generating one to. That would be awsome.23:01
*** apuimedo has joined #openstack-lbaas23:03
*** hichtakk has quit IRC23:19
*** hichtakk has joined #openstack-lbaas23:19
sbalukoffSorry-- got waylaid for a while (also, it's been a long time since I generated one of these by hand, so give me a bit to remember how to do that, eh.)23:34
*** apuimedo has quit IRC23:50
sbalukoffSorry crc32: I keep getting pulled off this. I'll get you something via e-mail tonight.23:54
crc32ok. carlos.garza@rackspace.com23:54

Generated by 2.14.0 by Marius Gedminas - find it at!