Monday, 2018-03-05

« Sunday, 2018-03-04 Index Tuesday, 2018-03-06 »
*** fnaval has quit IRC00:08
*** fnaval_ has joined #openstack-lbaas00:08
openstackgerritCarlos Goncalves proposed openstack/octavia master: [WIP] Add grenade support  https://review.openstack.org/54965400:13
openstackgerritCarlos Goncalves proposed openstack/octavia master: [WIP] Add grenade support  https://review.openstack.org/54965400:18
*** slaweq has joined #openstack-lbaas00:51
*** slaweq has quit IRC00:56
*** fnaval_ has quit IRC01:26
*** fnaval has joined #openstack-lbaas01:27
openstackgerritZhaoBo proposed openstack/octavia master: UDP jinja template  https://review.openstack.org/52542001:44
*** annp has joined #openstack-lbaas01:47
*** imacdonn has joined #openstack-lbaas02:02
*** threestrands_ has joined #openstack-lbaas03:12
*** threestrands_ has quit IRC03:12
*** threestrands_ has joined #openstack-lbaas03:12
*** threestrands has quit IRC03:16
*** links has joined #openstack-lbaas03:57
*** gcheresh has joined #openstack-lbaas04:31
*** gcheresh has quit IRC04:42
*** annp has quit IRC04:46
*** annp has joined #openstack-lbaas04:46
*** slaweq has joined #openstack-lbaas04:53
*** slaweq has quit IRC04:57
*** chandankumar has quit IRC06:04
*** links has quit IRC06:04
*** links has joined #openstack-lbaas06:05
*** slaweq has joined #openstack-lbaas06:06
*** slaweq has quit IRC06:11
*** chkumar246 has joined #openstack-lbaas06:22
*** threestrands_ has quit IRC06:57
*** AlexeyAbashkin has joined #openstack-lbaas07:07
*** kobis has joined #openstack-lbaas07:12
*** slaweq_ has joined #openstack-lbaas07:14
*** slaweq_ has quit IRC07:18
openstackgerritMin Sun proposed openstack/neutron-lbaas master: Update listener failed with: AttributeError: 'dict' object has no attribute 'tls_container_id'.  https://review.openstack.org/54969707:30
openstackgerritMin Sun proposed openstack/neutron-lbaas master: Update listener failed  https://review.openstack.org/54969707:33
*** b_bezak has joined #openstack-lbaas07:34
*** rcernin has quit IRC07:42
*** pcaruana has joined #openstack-lbaas08:05
*** tesseract has joined #openstack-lbaas08:11
*** AlexeyAbashkin has quit IRC08:22
*** AlexeyAbashkin has joined #openstack-lbaas08:25
*** slaweq_ has joined #openstack-lbaas09:15
*** zigo has quit IRC09:18
*** slaweq_ has quit IRC09:19
*** dosaboy has quit IRC09:20
*** dosaboy has joined #openstack-lbaas09:21
*** zigo has joined #openstack-lbaas09:23
*** salmankhan has joined #openstack-lbaas09:59
*** salmankhan has quit IRC10:03
*** salmankhan has joined #openstack-lbaas10:11
*** ivve has quit IRC10:58
*** ivve has joined #openstack-lbaas11:10
*** slaweq_ has joined #openstack-lbaas11:16
*** salmankhan has quit IRC11:19
*** slaweq_ has quit IRC11:20
*** salmankhan has joined #openstack-lbaas11:26
*** annp has quit IRC11:37
*** ivve has quit IRC11:39
*** salmankhan has quit IRC11:46
*** ivve has joined #openstack-lbaas11:55
*** salmankhan has joined #openstack-lbaas12:19
openstackgerritHengqing Hu proposed openstack/octavia master: Provide devstack samples for l3 active active  https://review.openstack.org/52087812:42
*** fnaval has quit IRC13:01
*** fnaval_ has joined #openstack-lbaas13:01
*** fnaval_ has quit IRC13:09
*** fnaval has joined #openstack-lbaas13:10
*** fnaval has quit IRC13:15
*** AlexeyAbashkin has quit IRC13:16
*** links has quit IRC13:23
*** fnaval has joined #openstack-lbaas13:26
*** atoth has joined #openstack-lbaas13:31
*** ianychoi has joined #openstack-lbaas13:41
*** AlexeyAbashkin has joined #openstack-lbaas13:48
*** slaweq_ has joined #openstack-lbaas14:01
*** slaweq_ has quit IRC14:05
xgerman_https://www.irccloud.com/pastebin/U99OuUcb/14:19
*** links has joined #openstack-lbaas14:19
xgerman_^^ feedback from Slack K8 OpenStack sig14:19
xgerman_(hodgepodge is the guy who gets paid by the OpenStack foundation to work on K8)14:21
nmagnezi_xgerman_, o/14:30
nmagnezi_xgerman_, Slack K8 OpenStack? what's that?14:30
xgerman_yes, exactly14:34
xgerman_it’s on the k8s slack workspace #sig-openstack14:34
xgerman_when I asked them an hour later they are all MIA14:35
*** links has quit IRC14:40
openstackgerritRalf Haferkamp proposed openstack/neutron-lbaas master: Fix missing device_id after recreating all listeners  https://review.openstack.org/54977914:56
*** raginbaji has quit IRC15:00
*** raginbajin has joined #openstack-lbaas15:03
openstackgerritNir Magnezi proposed openstack/neutron-lbaas master: [WIP]: Nuke lazy loaders and flush db sessions  https://review.openstack.org/54961315:27
*** openstackgerrit has quit IRC15:34
*** kobis has quit IRC15:40
*** kobis has joined #openstack-lbaas15:41
*** kobis has quit IRC15:41
*** bbzhao has quit IRC15:48
*** bbzhao has joined #openstack-lbaas15:48
*** bbzhao has quit IRC15:53
*** bbzhao has joined #openstack-lbaas15:54
*** AlexeyAbashkin has quit IRC15:59
*** openstackgerrit has joined #openstack-lbaas16:08
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Periodic job to build + publish diskimage  https://review.openstack.org/54925916:08
*** salmankhan has quit IRC16:11
*** salmankhan has joined #openstack-lbaas16:11
*** slaweq has joined #openstack-lbaas16:16
*** salmankhan has quit IRC16:17
*** salmankhan has joined #openstack-lbaas16:19
*** slaweq has quit IRC16:23
openstackgerritGerman Eichberger proposed openstack/neutron-lbaas master: Gate API test for the lbaasv2-proxy plugin  https://review.openstack.org/53935016:45
*** AlexeyAbashkin has joined #openstack-lbaas17:02
*** sapd has joined #openstack-lbaas17:06
*** fnaval_ has joined #openstack-lbaas17:07
*** fnaval has quit IRC17:07
*** beagles is now known as beagles|afk17:07
*** slaweq_ has joined #openstack-lbaas17:09
*** slaweq_ has quit IRC17:10
*** AlexeyAbashkin has quit IRC17:22
*** AlexeyAbashkin has joined #openstack-lbaas17:23
*** sshank has joined #openstack-lbaas17:25
*** AlexeyAbashkin has quit IRC17:30
*** sanfern has joined #openstack-lbaas17:36
*** AlexeyAbashkin has joined #openstack-lbaas17:52
*** sapd has quit IRC17:54
*** AlexeyAbashkin has quit IRC17:56
*** links has joined #openstack-lbaas18:10
*** b_bezak has quit IRC18:17
*** slaweq_ has joined #openstack-lbaas18:21
*** tesseract has quit IRC18:22
*** blake has joined #openstack-lbaas18:23
*** harlowja has joined #openstack-lbaas18:49
*** bbzhao has quit IRC18:50
*** bbzhao has joined #openstack-lbaas18:51
*** links has quit IRC18:51
*** AlexeyAbashkin has joined #openstack-lbaas18:51
*** beagles|afk is now known as beagles18:54
*** AlexeyAbashkin has quit IRC18:56
*** gcheresh has joined #openstack-lbaas19:10
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Switch amphora agent to use privsep  https://review.openstack.org/54929519:17
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Periodic job to build + publish diskimage  https://review.openstack.org/54925919:19
*** salmankhan has quit IRC19:21
*** sanfern has quit IRC19:27
*** gcheresh has quit IRC19:28
*** sshank has quit IRC19:42
*** blake has quit IRC19:44
*** AlexeyAbashkin has joined #openstack-lbaas19:51
*** AlexeyAbashkin has quit IRC19:55
*** gcheresh has joined #openstack-lbaas19:56
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Periodic job to build + publish diskimage  https://review.openstack.org/54925919:59
*** atoth has quit IRC20:11
*** gcheresh has quit IRC20:17
*** AlexeyAbashkin has joined #openstack-lbaas20:51
*** AlexeyAbashkin has quit IRC20:55
*** b_bezak has joined #openstack-lbaas20:56
*** aojea_ has joined #openstack-lbaas21:16
*** aojea_ has quit IRC21:28
*** aojea_ has joined #openstack-lbaas21:29
*** aojea_ has quit IRC21:29
*** aojea_ has joined #openstack-lbaas21:29
nmagnezi_xgerman_, rm_work, johnsom, I sure hope you guys made it home by now :)21:32
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Switch amphora agent to use privsep  https://review.openstack.org/54929521:32
xgerman_I am with my parents in germany and Michael is in Ireland with his wife on vacation21:33
bcafarelIreland without snow is a nice place to visit :)21:34
*** aojea_ has quit IRC21:37
xgerman_I hope so for him ;-)21:38
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Periodic job to build + publish diskimage  https://review.openstack.org/54925921:42
*** sshank has joined #openstack-lbaas21:43
*** threestrands_ has joined #openstack-lbaas21:45
*** threestrands_ has quit IRC21:45
*** threestrands_ has joined #openstack-lbaas21:45
openstackgerritNir Magnezi proposed openstack/neutron-lbaas master: [WIP]: Nuke lazy loaders and flush db sessions  https://review.openstack.org/54961321:50
*** AlexeyAbashkin has joined #openstack-lbaas21:51
*** AlexeyAbashkin has quit IRC21:55
*** sshank has quit IRC21:56
*** pcaruana has quit IRC21:57
openstackgerritGerman Eichberger proposed openstack/octavia master: [WIP] Periodic job to build + publish diskimage  https://review.openstack.org/54925922:19
*** rcernin has joined #openstack-lbaas22:34
*** b_bezak has quit IRC22:35
*** yamamoto has joined #openstack-lbaas22:44
*** yamamoto has quit IRC22:48
imacdonn2018-03-05 22:30:06.701 22264 ERROR octavia.controller.worker.controller_worker   File "/usr/lib/python2.7/site-packages/octavia/common/jinja/haproxy/templates/base.j2", line 32, in template22:56
imacdonn2018-03-05 22:30:06.701 22264 ERROR octavia.controller.worker.controller_worker     {% set found_ns.found = true %}22:56
imacdonn2018-03-05 22:30:06.701 22264 ERROR octavia.controller.worker.controller_worker TemplateSyntaxError: expected token 'end of statement block', got '.'22:56
imacdonnany known issues here ?22:56
imacdonnseems to be in the code added by change I83a3bf541c8c3c0f06544ef3228a5028756f51b322:56
openstackgerritCarlos Goncalves proposed openstack/octavia master: [WIP] Add grenade support  https://review.openstack.org/54965422:57
johnsomYour controller is not using the required jinja2 package from the upper constraints22:57
imacdonnOK, then I guess it's an RDO packaging bug22:57
johnsomimacdonn: upgrade your jinja on your computer ntrollers22:58
johnsomControllers, sorry, on mobile22:58
imacdonnRDO Queens comes with python2-jinja2-2.8.122:58
johnsomAh, old22:59
*** slaweq_ has quit IRC22:59
nmagnezi_ imacdonn the folks at #rdo might be able to assist with this22:59
nmagnezi_johnsom, o/ :)22:59
imacdonnYeah, going there next ... anyone know the required version off-hand ?22:59
johnsomo/ still in Ireland23:00
nmagnezi_johnsom, yeah xgerman_ told me. enjoy your vacation!23:00
imacdonnJinja2!=2.9.0,!=2.9.1,!=2.9.2,!=2.9.3,!=2.9.4,>=2.8 # BSD License (3 clause)23:00
johnsom2.10 if I remember correctly.  It is defined in the queens upper constraints file.23:00
imacdonnso .. in my math .. 2.8.1 > 2.823:00
imacdonnthis is from https://github.com/openstack/octavia/blob/master/requirements.txt23:00
johnsomhttps://github.com/openstack/requirements/blob/stable/queens/upper-constraints.txt23:02
johnsomI can’t paste the line from my mobile23:02
xgerman_You should be in the pub anyway ;-)23:03
johnsomBut search in there, it should be ==2.10 if I remember right, if not, it is whatever version is specified23:03
imacdonnYes, it is 2.10 in there .... I suspect that the RDO RPMs are built against the requirements.txt for the module in question, though23:03
johnsomxgerman_ no worries, in the bar with the wife23:04
imacdonnso it seems that really should be updated anyway23:04
xgerman_Probably...23:05
johnsomYeah, upper defines the exact versions for the release23:05
imacdonndoes it really? the name seems to suggest the *max* version, as opposed to the minimum23:06
johnsomWell, as pip always pulls the newest available, it basically defines the versions. Thus the ==23:07
*** yamamoto has joined #openstack-lbaas23:07
imacdonnthat may be find an dandy if you're using pip23:07
imacdonn:)23:07
johnsomWell, requirements.txt and upper constraints is defined by pip, so...23:08
*** jappleii__ has joined #openstack-lbaas23:09
johnsomAny which way, that is the solution23:10
imacdonnSo I still think the requirements.txt needs to be updated23:11
imacdonntrying to find a good definition of upper-constraints.txt .... some hints suggest that it's intended to specify the versions that should be used by CI23:11
imacdonnhttps://openstack.nimeyo.com/79453/openstack-dev-release-all-what-is-upper-constraints-txt23:12
*** jappleii__ has quit IRC23:12
*** threestrands_ has quit IRC23:12
*** threestrands has joined #openstack-lbaas23:12
*** threestrands has quit IRC23:12
*** threestrands has joined #openstack-lbaas23:12
imacdonnBTW, it appears to be an openstack thing, not a pip thing23:13
johnsomUpper is a flag on pip we pass in23:13
johnsomhttps://github.com/openstack/octavia/blob/stable/queens/tox.ini23:15
imacdonnyeah, and tox is used by CI for unit-testing23:15
johnsom-c23:15
johnsomYeah, but it is similar in other places23:16
imacdonnin any case, so you disagree that requirements.txt needs to be updated ?23:19
imacdonnArgggh.. not this again :/23:24
imacdonnLoad Balancer d3fe6073-4cce-476c-b968-9ed162d91e15 is immutable and cannot be updated. (HTTP 409) (Request-ID: req-3302c11f-0268-4096-aff1-ceea41efb29d)23:24
rm_workimacdonn: so technically, the requirements.txt will pull in the *newest*, which is bounded by upper-constraints23:36
rm_workusing our requirements.txt that you quoted, it would pull in 2.10 unless something else prevented it23:36
rm_work*that said*, we actually should change it, because the point is to say "what is the minimum version that will work with this software", and obviously it's wrong, if the minimum is really 2.10 not 2.823:37
rm_workbut it should already be pulling in the correct package23:37
*** aojea_ has joined #openstack-lbaas23:38
rm_workJinja2!=2.9.0,!=2.9.1,!=2.9.2,!=2.9.3,!=2.9.4,>=2.8 means:23:38
rm_workJinja2, >=2.8, but exclude/block a few versions that match that23:38
imacdonnrm_work: what is "it"?23:39
rm_workit's just blocking the 2.9 series which must have had some bugs23:39
rm_workit == pip23:39
imacdonnOK... I'm not sure that the RDO build process uses pip23:39
rm_workif you do `pip install Jinja2!=2.9.0,!=2.9.1,!=2.9.2,!=2.9.3,!=2.9.4,>=2.8` right now you should get 2.1023:39
rm_workno, it uses rpm packages23:39
rm_workbut that means they did the packaging wrong23:40
imacdonnyeah, I mean ... the process they use to build the RPMs23:40
rm_worktheir packaging process should use upper-constraint23:40
rm_workthis is stuff that nmagnezi_ or cgoncalves might know or be able to figure out23:40
imacdonnCan you show me something that makes that clear? So-far, all I've been able to find is statements that *CI* must use the version in upper-constraints23:41
rm_worki don't know who actually does the RDO packaging or what system it uses, but they would23:41
rm_worki mean, that's the openstack python version system23:41
rm_workif they are packaging for openstack, and not using requirements+upper-constraints, it is by definition wrong23:41
imacdonnI bet that they ARE using requirements.txt for package building23:42
rm_workif they were, they should still get newer23:42
imacdonnas far as I can tell, that's what they should be using.... for package build ... for *CI*, it's a different matter23:42
rm_workthey still need to calculate the correct version23:42
imacdonn"should get newer"... from where? A lot of the requirements are met by existing RPMs that are distributed as part of the OS23:43
rm_workand the only way to do that correctly is to look at a combination of requirements and upper-constraints23:43
*** aojea_ has quit IRC23:43
rm_workfrom pypi23:43
rm_workbecause the packaging of the RDO python libs should be based on pypi + requirements + upper-constraints as well23:43
imacdonnMy assertion is that the only correct way to do it is to ensure that the RPM has a dependency that agrees with requirements.txt for the module23:43
imacdonnagain, for *CI*, it's different23:44
rm_workso when they choose what pypi libs to package for the distro, they should be calculating them from requirements/U-C23:44
rm_workwell, if it is, it shouldn't be23:44
rm_workbecause being different would be wrong23:44
*** fnaval_ has quit IRC23:44
imacdonnI need to see a definition of U-C that shows that it pertains to anything other than CI23:44
rm_workit pertains to openstack, period23:45
imacdonnshow me the doc23:45
rm_workhttps://wiki.openstack.org/wiki/Requirements23:45
rm_work"Our project maintains global-requirements.txt, upper-constraints.txt and a set of jobs around these files. The goal of this project is to keep openstack up to date with newly released python packages without breaking Openstack CI. Packagers are advised to use the versions (upper-constraints) we have tested for their releases."23:45
imacdonnOK... "advised to" is not the same as "required to", though23:46
rm_workhttps://openstack.nimeyo.com/79453/openstack-dev-release-all-what-is-upper-constraints-txt23:46
imacdonnthat's where I got that it pertains specifically to CI23:47
imacdonnin any case, it'll be a lot easier to make the argument if requirements.txt is correct23:48
rm_workit pertains to anything23:48
rm_workIf you are installing openstack software -> You should use requirements and upper-constraints23:48
rm_workIf you dont -> Things may fail, and you are on your own23:49
rm_workIf you have objections to this, maybe take it up in #openstack-requirements but that's how it works23:49
rm_workthis isn't a new thing and it isn't very vague IMO :P23:49
rm_workif RDO is packaged without regarding U-C, then they are not following the correct process23:50
rm_workbut i feel like they know that, so something else is probably going on here23:50
imacdonnwell, clearly requirements.txt is wrong... we did agree on that much23:50
rm_workyes23:51
rm_workwe should update that23:51
rm_workbut this is exactly why we use upper-constraints23:51
rm_workbecause it will unify the packaging versions across all projects23:51
rm_workand guarantee everything works according to our CI23:52
*** AlexeyAbashkin has joined #openstack-lbaas23:52
rm_workbasically, the rest of openstack also requires jinja 2.10, and even if we had that in our requirements, they seem to have not packaged it?23:53
rm_workor else you would be getting that23:53
imacdonnmy interpretation is that "requirements.txt" specifies the requirements for running the code, and U-C specifies the version that CI should use ... because CI would pull the latest from pip unless contained to an *upper* version (the absolute latest may be "too new")23:53
rm_workso why would you ever want to use a version different than what the CI uses?23:54
rm_worktesting should === production23:54
imacdonnin that case, why does requirements.txt even exist ?23:54
rm_workit honestly doesn't matter a whole lot23:54
rm_worku-c is what matters23:54
rm_workg-r and u-c23:54
rm_workg-r being global-requirements23:55
imacdonnnote that G-R for queens says >= 2.823:55
rm_workhonestly it would maybe be less confusing if our requirements just had package names and not versions23:55
imacdonnsJinja2>=2.8,!=2.9.0,!=2.9.1,!=2.9.2,!=2.9.3,!=2.9.4 # BSD License (3 clause)23:56
rm_workyeah, >=2.8 at the date of queens release would be 2.1023:56
imacdonnhuh?23:56
rm_workyou see that there's a "greater than" there, right?23:56
rm_workqueens was released last week23:56
imacdonnyou can see that 2.8.1 >= 2.8, right ?23:56
*** AlexeyAbashkin has quit IRC23:56
rm_workyeah, but it isn't like ... ?>=23:56
rm_workthere's no "not greedy" operator23:57
rm_workpip will always get the newest matching package23:57
imacdonnargh .. this has nothing to do with pip23:57
rm_workas will any package manager23:57
rm_workand it has everything to do with pip, as pip is the python package manager23:57
rm_workso distros should be matching what pip does23:58
rm_workqueens released last week, and at that time the CI was running on whatever pip was pulling in23:58
rm_workif distros choose to not match that, they do so at their own peril23:58
rm_workwe helpfully define the *exact* version in upper-constraints23:58
rm_workso that is what distros should be looking at for their packaging23:59
« Sunday, 2018-03-04 Index Tuesday, 2018-03-06 »

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