Tuesday, 2014-06-24

*** german_ has quit IRC00:06
*** sbfox1 has quit IRC00:09
*** vivek-ebay has quit IRC00:10
*** dlundquist has left #openstack-lbaas00:17
*** markmcclain1 has quit IRC00:22
dougwighmm, i must be the only one that thinks the tls/listener object split is premature optimization.  i wouldn't split it until too many new attributes or 1:n came into play.00:33
*** markmcclain has joined #openstack-lbaas00:36
*** markmcclain has quit IRC00:43
sbalukoffdougwig: Say that on the list, eh!00:53
sbalukoffI'm totally wiling to flip flop like a fish on this if you can come up with an defend solid reasoning on it, eh.  (Frankly, I don't actually care that much either way, other than that we arrive on a strategy without spending too much time on very minor details like this.)00:54
dougwighow many attributes is it right now?00:56
dougwigit's just a barbican id?  (not counting SNI.)  yeah, that's simple enough to warrant a reply.01:00
dougwigok, there's some mailing list dissension.01:03
dougwigwhat's with all the eh, eh?  you have some canadian in you?01:05
sbalukoffIt's totally my fake canadian accent, eh.01:11
sbalukoffI didn't tell that story at the hackathon?01:11
dougwigno, you sure didn't.01:11
sbalukoffBack in 2001 I worked for a company in Bend, Overgon called Cylant Technology where we were working on a new kind of linux kernel security (behavioral analysis-- it was pretty cool actually).01:12
sbalukoffAnyway, we imported a developer from Saskatook, Saskatchewan, eh...01:12
sbalukoffHis accent was so thick...  he'd try to tell a joke, and he'd be like, "So there were these three guys, eh, and they walked into a bar, eh..."01:13
sbalukoffAnd at this point we'd already be cracking up.01:13
sbalukoffHe'd protest, "I haven't even gotten to the punch line, eh!"01:13
dougwiglol01:14
sbalukoffAnd we'd say "You ARE the punch line, dude!"01:14
sbalukoffSo, of course we had to make fun of him incessantly.01:14
sbalukoffSix weeks later, he's speaking like an American-- no discernable accent at all.01:14
sbalukoffBut I picked up this fake Canadian accent thing, and haven't been able to shake it since.01:14
sbalukoffIt was awkward when I actually did visit Saskatoon--  the locals weren't sure if I was making fun of them or not. But they were too polite to ask or confront me on it.01:15
dougwigyou sound like my friend that adopts the accent of whoever he's talking to.  so listening to this californian conversing with someone from brooklyn and the UK at the same time was absolutely hilarious.01:15
sbalukoffHaha! That's awesome!01:16
sbalukoffAnd yeah, I do that somewhat.01:16
dougwigoh, goody, i love fucking with that.01:16
sbalukoffTotally, eh.01:16
*** fnaval has joined #openstack-lbaas01:16
dougwigok, off to grab some dinner.  might as well make the best of california.  no afghan restaurants in boise, e.g.01:17
dougwigthanks for the review feedback.01:17
sbalukoffNo worries.01:17
sbalukoffHave a good one!01:17
*** crc32 has quit IRC01:23
*** woodster__ has quit IRC01:25
VijayB_ctracey: Sorry, got deep into coding and stopped looking at the irc window.. I'm looking at the etherpad https://etherpad.openstack.org/p/juno-lbaas-mid-cycle-hackathon but don't see the git info there - wrong etherpad, eh ............. sbalukoff! Why eh!!!02:08
VijayB_ctracey: so I still need the git info... :)02:09
*** VijayB has joined #openstack-lbaas02:13
*** VijayB_ has quit IRC02:17
*** VijayB has quit IRC02:18
sbalukoff:)02:21
sbalukoffOk, folks, commuting home. Catch ya later.02:21
*** sbalukoff has quit IRC02:21
*** fnaval has quit IRC02:26
*** mestery has quit IRC02:31
*** mestery has joined #openstack-lbaas02:31
*** fnaval has joined #openstack-lbaas02:33
*** fnaval has quit IRC02:45
*** mestery_ has joined #openstack-lbaas02:52
*** mestery has quit IRC02:55
*** sbalukoff has joined #openstack-lbaas03:20
*** sbfox has joined #openstack-lbaas03:26
*** sbfox has quit IRC03:58
*** sbfox has joined #openstack-lbaas04:05
*** german_ has joined #openstack-lbaas04:19
*** min has quit IRC04:39
*** zz_blogan is now known as blogan04:51
*** german_ has quit IRC05:12
*** ptoohill_ has joined #openstack-lbaas05:14
*** sbfox has quit IRC05:22
*** ptoohill_ has quit IRC05:23
*** VijayB has joined #openstack-lbaas05:25
*** ptoohill_ has joined #openstack-lbaas05:25
*** VijayB has quit IRC05:36
*** vivek-ebay has joined #openstack-lbaas05:51
*** vivek-ebay has quit IRC05:57
*** evgenyf has joined #openstack-lbaas06:12
*** blogan is now known as zz_blogan06:41
*** openstackgerrit has quit IRC07:10
*** rm_work has quit IRC07:29
*** rm_work has joined #openstack-lbaas07:33
*** vjay2 has joined #openstack-lbaas08:35
*** evgenyf has quit IRC09:20
*** enikanorov_ has quit IRC09:37
*** enikanorov_ has joined #openstack-lbaas09:38
*** evgenyf has joined #openstack-lbaas09:44
*** openstackgerrit has joined #openstack-lbaas09:47
*** HenryG has quit IRC10:25
*** vjay2 has quit IRC11:49
*** vjay2 has joined #openstack-lbaas11:55
*** vjay2 has quit IRC12:02
*** vjay2 has joined #openstack-lbaas12:54
*** vjay2 has quit IRC13:35
*** vjay2 has joined #openstack-lbaas13:38
*** mestery_ is now known as mestery13:42
*** vivek-ebay has joined #openstack-lbaas13:47
*** HenryG has joined #openstack-lbaas14:04
*** vivek-ebay has quit IRC14:13
*** rolledback has joined #openstack-lbaas14:14
*** ptoohill_ has quit IRC14:24
*** fnaval has joined #openstack-lbaas14:36
*** fnaval has quit IRC14:36
*** fnaval has joined #openstack-lbaas14:36
*** ptoohill_ has joined #openstack-lbaas14:40
*** ptoohill_ has quit IRC14:40
*** ptoohill_ has joined #openstack-lbaas14:40
*** markmcclain has joined #openstack-lbaas14:45
*** markmcclain1 has joined #openstack-lbaas14:47
*** ptoohill_ has quit IRC14:49
*** markmcclain has quit IRC14:50
*** ptoohill_ has joined #openstack-lbaas14:52
*** ptoohill_ has quit IRC14:55
*** german_ has joined #openstack-lbaas15:04
*** vjay2 has quit IRC15:23
*** zz_blogan is now known as blogan15:34
*** woodster__ has joined #openstack-lbaas15:35
*** fnaval has quit IRC15:41
*** HenryG has quit IRC15:43
*** jorgem has joined #openstack-lbaas15:47
dougwigsbalukoff, blogan, others: no-op driver, updated.  abstract_driver was getting a bit unruly and unclear how to extend the managers, so it got a minor refactor.  https://review.openstack.org/#/c/101084/15:49
blogandougwig: did you get my comment about the super?15:49
blogandougwig: looks like you did bc the BaseStatusMixin doesn't have an init now15:50
dougwigwas that a comment in patch set 15?15:51
dougwigotherwise, no, i don't remember seeing one.15:52
dougwigbut if i accidentally addressed it, i guess that's a win?  ;)15:52
bloganlol15:53
bloganyes i believe you addressed it on accident15:53
dougwigi mean, i took out that init on purpose, but i didn't know you didn't want it.  :)15:53
bloganit was basically calling super with multiple inheritance15:53
bloganwasn't that i didnt want it, it was what you were expecting out of the super call in the BaseLoadBalancerManager15:54
bloganbc it would have only called one of the parent classes init15:54
dougwigthe unit test informed me that my expectations were not in line with reality.  :)15:55
*** markmcclain1 has quit IRC15:57
*** min has joined #openstack-lbaas15:57
*** vivek-ebay has joined #openstack-lbaas16:00
*** evgenyf has quit IRC16:00
*** vivek-ebay has quit IRC16:03
*** vivek-ebay has joined #openstack-lbaas16:04
*** fnaval has joined #openstack-lbaas16:06
*** rolledback has quit IRC16:10
ctraceysilly question...16:15
ctraceyhas anyone done any work to get the new API actually working?16:15
ctraceyie. devstack integration?16:16
bloganctracey: is that a two part question?16:22
ctraceyblogan: if you'd like it to be :)16:22
bloganctracey: because I've got parts of the new API "working" (just inserting to teh database for now), but nothing pushed to devstack right now that says this needs to be spun up16:23
bloganbut i do use devstack to spin up the environment and just change a few neutron config values to get the new API running16:23
ctraceyok - that is where i have been spinnin my wheel16:23
bloganah16:24
ctraceyI will get some devstack going16:24
*** rolledback has joined #openstack-lbaas16:24
bloganlet me know if you need the new config values as well bc that is not readily obvious16:24
ctraceywant to fire me your neutron.conf and any other configs?16:24
ctraceyjust make sure I got them right?16:25
bloganits just the neutron.conf right now and I'll just give you two lines to add in16:25
ctraceysounds like we need devstack at some point anyway - now's as good a time as any16:25
*** rolledback has quit IRC16:25
bloganwell I'm not sure if we are going to end up having both versions running side by side or just have v2 running with shims in place, in which case we won't need the extra lines16:26
bloganbut i wasn't going to cross that bridge until its time to make that decision16:27
bloganwhcih will be soon16:27
*** vivek-ebay has quit IRC16:31
ctraceyblogan: can you gist what you have?16:49
blogansure16:49
bloganhttps://gist.github.com/brandonlogan/5d28ee177a0dc917289d16:52
bloganof course the first line needs to replace the one that exists, the 2nd is just an additional line16:52
dougwigblogan: can i get a +1 or -1 when you get a chance?  i want to close this review and get it done.16:56
bloganoh whoops meant to do that16:56
blogandougwig: does a health monitor need a status?16:57
dougwigyep, but it has its own set of plugin functions for it.16:58
dougwigi assume you meant config/db status, not operational status?16:58
bloganyeah like pending create, update, delete17:02
blogani was hoping we could minimize the status to just loadbalancer and member, with member's status being an operational status17:03
bloganat least for the first iteration17:03
dougwigif you're reducing it, we can do so in the driver.  the current status mixins exactly mirror v1.17:03
bloganand if any entity is connected to a load balancer in any form it just puts that loadbalancer into a pending_update status17:03
dougwigwe were going to add operational status via get() calls later.17:04
bloganyeah i know, i guess this needs more discussion17:04
bloganbrb17:05
TrevorVDoes anyone know why when I'm trying to run "stack.sh" it decides to nuke my pip installation and then fails because it tries to use it after nuking it?17:07
TrevorVEven after a fresh pull from devstack?17:07
TrevorV(Meaning I re-cloned just to double check)17:08
*** VijayB has joined #openstack-lbaas17:11
*** openstackgerrit has quit IRC17:35
*** openstackgerrit has joined #openstack-lbaas17:36
*** vivek-ebay has joined #openstack-lbaas17:37
*** vjay2 has joined #openstack-lbaas17:40
ctraceyTrevorV: I saw that on Min's box.  I think it had to do with the version of pip.17:46
dougwigblogan: how so?  it's a tad frustrating that we didn't discuss that kind of thing via the bp, since the entry points were all listed.  we're designing by code review right now.17:46
*** sbfox has joined #openstack-lbaas17:47
ctraceyiirc to get her working, I installed pip, then 'pip install pip -U' then stack.sh17:47
ctraceybut I could be wrong.17:47
blogandougwig: i know i wish I would have thought about this during the meetup17:48
VijayBIs there a place where we can see the log of the mesgs on this channel? I get disconnected often and tend to lose some mesgs inbetween, so slip out of context..17:49
blogandougwig: but i figured there would be many little things that needed to be discussed17:49
bloganVijayB: i think there is now that the openstack bot is in here but i dont know where one would get that informatino17:51
VijayBblogan: Ok.. I'll see if I can find that out and let the group know.. btw did you get a chance to see my mesgs yesterday regarding the to_dict()?17:52
TrevorVctracey, thanks for the suggestion, but it still failed.17:52
TrevorVI sent a lengthy email with information17:52
TrevorVHopefully one of us can help me, and if not I'll re-send the email to the full openstack-dev list17:53
*** dlundquist has joined #openstack-lbaas17:55
dougwigsec, i'm in a meeting17:58
*** vivek-ebay has quit IRC17:59
*** vivek-ebay has joined #openstack-lbaas18:01
*** fnaval has quit IRC18:06
ctraceyTrevorV: precise?18:14
TrevorVctracey, its trusty, 14.04 right?18:21
TrevorVI'm about to verify ha ha18:21
TrevorVYeah, Trusty Tahr18:22
ctraceytrusty is 14.0418:22
TrevorVThat's what I'm running18:22
TrevorVHowever I already got an answer to the email I sent with a lot of information packed in18:22
TrevorVIt seems all I had to do was run "easy_install pip" after the first "./stack.sh" command failed, and then it seems to have worked.18:22
TrevorVNot sure why it fixes it, and neither does the person who suggested it, so I'm just going to go with "magic"18:23
blogandougwig: i'm sending an email to the ML about this and my thoughts18:23
*** fnaval has joined #openstack-lbaas18:25
*** vjay2 has quit IRC18:28
dougwigi'll answer there, but i think you're bundling too much into the LB object.  that's something that a UI should do, but the API should be working with smaller pieces, and each should have it's own db status.18:36
dougwig(still in meeting.)18:36
*** HenryG has joined #openstack-lbaas18:48
*** sbfox has quit IRC18:50
*** sbfox has joined #openstack-lbaas18:50
dougwigresponded.18:55
dougwigi have a meta-objection to being in a state of review paralysis at the moment.  it's tough for us all to iterate.  i see both sides.18:57
*** sbfox has quit IRC19:03
*** vivek-ebay has quit IRC19:03
*** sbfox has joined #openstack-lbaas19:03
*** dlundquist has left #openstack-lbaas19:13
*** vivek-ebay has joined #openstack-lbaas19:13
*** vivek-ebay has quit IRC19:16
*** vivek-ebay has joined #openstack-lbaas19:16
blogandougwig: i'm gong to +1 your review because if it is decided that health monitors dont need a status then those methods dont have to do anything in a subclass19:21
german_health monitors need a status since they can be provisioned19:23
german_so they might be BUILDING, ACTIVE, ERROR, ETC19:23
ptoohillbut, when provisioning most objects its config updates, the only thing thats being processed is the 'whole loadbalancer' object which in the ref impl is the haproxy config19:25
german_yeah, it's an edge case that provsioning the LB succeeds and say the health monitor fails19:26
german_but I just would like to prepare for all possible permutations19:27
ptoohillyea, i see both cases, guess im just attempting to keep the short timelines in mind19:28
german_or I create an LB, make a new Healthmonitor to replace the old one, the deploy of the new haproxy fails, but the old one is till running...19:28
ptoohillin case of haproxy19:28
german_yep, just keep it simple there :-)19:28
ptoohillthe old one will never be 'still' running, once you update the config, the config for that 'loadbalncer' is refreshed and update along with its healthmonitor config19:28
german_it's admittedly an edge case but I have seen stranger things where the control plane lost contact and the lb kept jugging along :-)19:30
german_my philosophy is that the more complex we can signal error cases the less roll back we need to do19:31
*** VijayB has quit IRC19:44
*** VijayB_ has joined #openstack-lbaas19:49
*** vivek-ebay has quit IRC19:53
*** VijayB_ has quit IRC19:53
*** fnaval has quit IRC20:17
*** fnaval has joined #openstack-lbaas20:18
*** fnaval has quit IRC20:22
*** fnaval has joined #openstack-lbaas20:27
*** blogan is now known as zz_blogan20:35
*** sbfox has quit IRC20:37
*** vivek-ebay has joined #openstack-lbaas20:54
*** VijayB has joined #openstack-lbaas20:58
*** vivek-ebay has quit IRC20:59
*** HenryG has quit IRC21:05
*** vivek-ebay has joined #openstack-lbaas21:13
*** jorgem has quit IRC21:25
*** enikanorov_ has quit IRC21:38
*** enikanorov_ has joined #openstack-lbaas21:38
*** sbfox has joined #openstack-lbaas21:45
*** sbfox1 has joined #openstack-lbaas22:10
*** sbfox has quit IRC22:12
*** sbfox1 has quit IRC22:19
*** sbfox has joined #openstack-lbaas22:19
*** zz_blogan is now known as blogan22:23
*** HenryG has joined #openstack-lbaas22:27
*** fnaval has quit IRC22:56
*** min has quit IRC23:14
*** dlundquist has joined #openstack-lbaas23:16
*** sbfox has quit IRC23:23

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!