Tuesday, 2017-05-23

*** salv-orlando has quit IRC00:01
*** kzaitsev_mb has joined #openstack-kuryr00:06
*** kzaitsev_mb has quit IRC00:11
*** hongbin has quit IRC00:22
*** tonanhngo has quit IRC00:36
*** feisky has joined #openstack-kuryr00:42
*** tonanhngo has joined #openstack-kuryr00:43
*** tonanhngo_ has joined #openstack-kuryr00:47
*** tonanhngo has quit IRC00:48
*** tonanhngo_ has quit IRC00:51
*** pmannidi has quit IRC01:28
*** gianpietro has joined #openstack-kuryr01:28
*** gianpietro has quit IRC01:33
*** pmannidi has joined #openstack-kuryr01:55
*** pmannidi has quit IRC02:00
*** hongbin has joined #openstack-kuryr02:06
*** kzaitsev_mb has joined #openstack-kuryr02:07
*** pmannidi has joined #openstack-kuryr02:10
*** vikasc has quit IRC02:10
*** kzaitsev_mb has quit IRC02:13
*** vikasc has joined #openstack-kuryr02:22
*** yedongcan has joined #openstack-kuryr02:35
*** pmannidi has quit IRC02:39
*** hongbin_ has joined #openstack-kuryr02:48
*** hongbin has quit IRC02:55
*** hongbin_ has quit IRC02:56
*** hongbin has joined #openstack-kuryr02:56
*** gianpietro has joined #openstack-kuryr03:04
*** gianpietro has quit IRC03:09
*** kzaitsev_mb has joined #openstack-kuryr03:10
*** aojea has joined #openstack-kuryr03:15
*** pmannidi has joined #openstack-kuryr03:15
*** kzaitsev_mb has quit IRC03:16
*** aojea has quit IRC03:19
*** pmannidi_ has joined #openstack-kuryr03:42
*** pmannidi_ is now known as pmannidi|Trg03:42
*** pmannidi|Trg is now known as pmannidi__03:43
*** pmannidi has quit IRC03:44
*** kzaitsev_mb has joined #openstack-kuryr04:13
*** hongbin has quit IRC04:20
*** kzaitsev_mb has quit IRC04:25
*** ltomasbo|away is now known as ltomasbo05:32
*** aojea has joined #openstack-kuryr05:38
openstackgerritvikas choudhary proposed openstack/kuryr-kubernetes master: k8s bearer token support  https://review.openstack.org/46381405:48
*** aojea has quit IRC05:57
*** kzaitsev_ws has joined #openstack-kuryr06:23
*** pcaruana has joined #openstack-kuryr06:40
vikascirenab, i was actually trying to add test for token use case whcih you asked for. But then it was taking time to figure out how to mock "open" system call. :)06:58
vikascirenab, will figue out and add06:58
irenabvikasc,  thanks07:00
vikascirenab, yw07:00
*** dimak_ has joined #openstack-kuryr07:06
*** salv-orlando has joined #openstack-kuryr07:10
*** salv-orlando has quit IRC07:25
*** danil has joined #openstack-kuryr07:26
*** aojea has joined #openstack-kuryr07:26
*** kzaitsev_ws has quit IRC07:39
*** ltomasbo is now known as ltomasbo|away07:43
*** kzaitsev_ws has joined #openstack-kuryr08:01
apuimedokzaitsev_ws: just finished with the spec. It is well defined. Just addressing a couple of comments and it will be ready to be taken08:08
kzaitsev_wsapuimedo: thanks. I'm trying to force myself to address the comments )08:09
kzaitsev_wsneed more coffee08:09
apuimedoirenab: we can probably take https://review.openstack.org/#/c/464394/6 in08:09
apuimedokzaitsev_ws: I added a couple of them just now08:09
apuimedobut irenab and vikasc had covered practically all08:10
kzaitsev_wsalso we suddenly had a fire drill here in the office %)08:10
apuimedovikasc: did you already find how to mock the open?08:10
apuimedokzaitsev_ws: I'm working from the train today08:10
apuimedoxD08:10
vikascapuimedo, not yet08:11
apuimedolet me take a look08:11
kzaitsev_wsapuimedo: vikasc: @mock.patch('six.moves.builtins.open')08:13
vikasckzaitsev_ws, i gave a quick try with builtin08:14
vikasckzaitsev_ws,  that seemed to require future08:14
*** garyloug has joined #openstack-kuryr08:14
*** egonzalez has joined #openstack-kuryr08:15
vikasckzaitsev_ws, i dint try much on this. i think it should not be very complicate problem. :)08:15
kzaitsev_wsah, no08:16
vikasckzaitsev_ws,  will try after sometime. in mid of some other activity08:16
kzaitsev_wsit's not that easy08:16
vikasckzaitsev_ws, then i am not disappointed that my quick try dint work :)08:17
kzaitsev_wshttps://github.com/openstack/murano-agent/blob/36313ef433e2d6e968827a1d89f05b16ecf74c3c/muranoagent/tests/unit/test_files_manager.py#L7208:17
kzaitsev_wshttps://github.com/openstack/murano-agent/blob/36313ef433e2d6e968827a1d89f05b16ecf74c3c/muranoagent/tests/unit/test_files_manager.py#L21208:17
kzaitsev_wsthat's how we did it in murano08:18
*** openstackgerrit has quit IRC08:18
vikasckzaitsev_ws, thanks a lot kiril , will refer there :)08:18
kzaitsev_wsnot sure its correct though, but it did work =P08:19
vikascgreat!!!08:19
*** ashishb has joined #openstack-kuryr08:28
kzaitsev_wsmaybe there are simpler ways, like http://boris-42.me/the-simplest-way-in-python-to-mock-open-during-unittest/ using create. I remember that we had to use six.moves for py3 compatibility though08:32
*** zengchen has joined #openstack-kuryr08:37
*** ltomasbo|away is now known as ltomasbo08:44
*** neiljerram has joined #openstack-kuryr09:17
*** vikasc has quit IRC09:47
*** vikasc has joined #openstack-kuryr10:01
*** salv-orlando has joined #openstack-kuryr10:49
*** salv-orlando has quit IRC10:53
*** yamamoto has quit IRC11:17
*** gouthamr_ has joined #openstack-kuryr11:35
*** yedongcan has left #openstack-kuryr11:37
*** salv-orlando has joined #openstack-kuryr11:42
*** atoth has joined #openstack-kuryr11:48
*** openstackgerrit has joined #openstack-kuryr11:49
openstackgerritOpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements  https://review.openstack.org/45447411:49
openstackgerritOpenStack Proposal Bot proposed openstack/kuryr master: Updated from global requirements  https://review.openstack.org/46210611:51
*** salv-orlando has quit IRC11:52
*** aojea has quit IRC11:56
*** yamamoto has joined #openstack-kuryr11:58
*** yamamoto_ has joined #openstack-kuryr12:00
*** gouthamr_ has quit IRC12:02
*** yamamoto has quit IRC12:04
irenabapuimedo, ping12:07
*** dimak_ has quit IRC12:09
openstackgerritBerezovsky Irena proposed openstack/kuryr-kubernetes master: Split nested drivers config options out of generic config.  https://review.openstack.org/45011312:19
apuimedoirenab: pong12:21
apuimedovikasc: did you manage to do the fs mocking?12:23
vikascapuimedo, i could not get time to try that yet12:24
vikascapuimedo, will do soon.12:24
apuimedook12:24
*** dimak_ has joined #openstack-kuryr12:26
*** garyloug has quit IRC12:28
irenabapuimedo, is there officially release version of kuryr-k8s?12:42
apuimedo0.1.012:44
irenabthe reason for question is if we need release note for moving config option12:44
apuimedohttps://tarballs.openstack.org/kuryr-kubernetes/12:44
apuimedofrom April 19th12:44
apuimedoof the nested stuff?12:44
apuimedomaybe we can just use the deprecated note and allow it12:44
irenabdo we need a deprecation cycle for the previous config ?12:45
apuimedoand drop it when we get to 1.0.012:45
irenabdeprecated note in release notes or in python config module?12:45
irenabfor some reason I though we are still flexible to move things around12:46
irenabno stable branch yet12:46
*** salv-orlando has joined #openstack-kuryr12:48
irenabapuimedo, let me know what do you think. I think keeping both locations for the nested settings and supporting it is not required at the current state. But if we are required once 0.1.0 is out, then we should do it12:52
apuimedotrue12:52
apuimedojust move it and put a note12:52
*** salv-orlando has quit IRC12:53
irenabwe should converge on config proper placement before stable branch is cut12:53
*** garyloug has joined #openstack-kuryr12:54
openstackgerritBerezovsky Irena proposed openstack/kuryr-kubernetes master: Split nested drivers config options out of generic config.  https://review.openstack.org/45011312:59
irenabapuimedo, check if this what you had in mind ^^13:02
apuimedook13:05
apuimedoirenab: kzaitsev_ws: I meant to put it in reno or something, so it would be in the release notes13:07
dmelladoapuimedo: +1 to reindeer13:08
dmelladoxD13:08
kzaitsev_wsI don't have a strong opinion on the config. However there is a README.rst, that's now inconsistent =)13:13
apuimedokzaitsev_ws: let's fix that in the same commit then13:13
kzaitsev_wsthat's was the problem that deprecated_group would solve or rather delay =)13:13
kzaitsev_wsyeah, I mentioned that in a review.13:13
*** salv-orlando has joined #openstack-kuryr13:15
dmelladobtw, apuimedo could you send the upstream bluejeans meeting invite for next week re: functional testing && tempest?13:18
*** salv-orlando has quit IRC13:20
apuimedodmellado: when is it supposed to happen?13:22
dmelladohmmm next week's Wed?13:22
apuimedodmellado: did you send me an invite?13:24
dmelladoapuimedo: actually I was thinking about having you act as my secretary and send it13:25
dmelladowait, I'll create and invite you xD13:25
*** gsagie has joined #openstack-kuryr13:26
dmelladoapuimedo: sent13:26
apuimedothanks dmellado13:26
dmelladoapuimedo: feel free to fwd it or share over the ML13:27
apuimedoI just wasn't sure if you used your thunderbird plugin that makes invites go to /dev/nulo (in spanish, as the invites)13:27
dmelladoapuimedo: don't make me do the ryme with nulo13:27
dmelladoxD13:27
apuimedodmellado: this one I'm doing with you in mind https://blueprints.launchpad.net/kuryr-kubernetes/+spec/devstack-support-api-accessing-pods13:44
apuimedoirenab: I'd appreciate approving the BP13:44
dmelladoapuimedo: lemme hava a look13:44
dmelladoheh, cool13:45
dmelladoactually I was going to add that to devstack myself13:45
dmelladoappretiate the bp, though xD13:45
* dmellado shouldn't be lazy and add bp...13:45
openstackgerritAntoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: Create infra for K8s API accessing pods  https://review.openstack.org/46724113:47
apuimedodmellado: too late ^^13:47
apuimedodmellado: but I'll let you volunteer for actually trying it out :-)13:47
apuimedovikasc: this patch should make your containerized kuryr-k8s testing easier13:48
dmelladoapuimedo: irenab13:49
dmelladohttps://blueprints.launchpad.net/kuryr-kubernetes/+spec/kuryr-tempest-plugin13:49
*** hongbin has joined #openstack-kuryr13:51
dmelladoltomasbo: ^^13:52
ltomasbothanks dmellado13:53
ltomasboI'll take a look later13:53
dmelladoit's quite straightforward, but just for tracking it :D13:53
openstackgerritKirill Zaitsev proposed openstack/kuryr-kubernetes master: Add kuryr-sriov spec proposal  https://review.openstack.org/46566113:54
*** gianpietro has joined #openstack-kuryr13:56
*** aojea has joined #openstack-kuryr13:57
kzaitsev_wsirenab: there are a couple of things we need to decide before I go further with updating my code.13:58
kzaitsev_ws1) I mentioned in a spec, that I wanted to use a separate annotation key to not interfere with regular vif and you mentioned in the code review, that it might be possible to not use one https://review.openstack.org/#/c/462455/3/kuryr_kubernetes/constants.py13:59
kzaitsev_wsthe thing is regular annotation holds a single object, while I want to pass a map of objects '{physnetX: [list-ov-direct-vifs]}'14:00
kzaitsev_wsso I thought a separate annotation key would be a good idea to keep things clean and separated.14:01
kzaitsev_wsother options would probably require modifying regular VIF format, which I wanted to avoid14:02
*** aojea has quit IRC14:03
apuimedoseparate is fine14:03
kzaitsev_ws2) adding SRIOV-VIF handler vs altering generic VIF handler.14:06
kzaitsev_wsmy biggest concern against super-vif handler is code-bloat14:06
apuimedoaddition14:07
apuimedoI don't want bloated handlers14:07
*** dimak_ has quit IRC14:08
*** salv-orlando has joined #openstack-kuryr14:10
kzaitsev_ws=) not trying to be stubborn, just trying to find the best possible compromise. )14:10
kzaitsev_wsvikasc: apuimedo: irenab: thanks for the reviews everyone btw =) I've updated the spec )14:11
vikascthanks kzaitsev_ws !14:13
*** salv-orlando has quit IRC14:14
*** danil has quit IRC14:15
kzaitsev_wsapuimedo: btw, I managed to hack Services support commit to handle kubernetes service automatically through kuryr-k8s-controller14:19
kzaitsev_wsis that what you're going to do with `devstack-support-api-accessing-pods` in the end?14:19
apuimedokzaitsev_ws: with which subnet for the member?14:20
kzaitsev_wsoh. I think you mentioned, that your installation does that semi-automatically14:20
kzaitsev_wsapuimedo: it used the same subnet as regular services14:21
apuimedokzaitsev_ws: baremetal?14:22
kzaitsev_wsyes, that would be a problem?14:25
*** dimak_ has joined #openstack-kuryr14:25
apuimedono, just making the picture whole14:25
apuimedoI added the patch to devstack because host networking services are controversial14:26
apuimedo:P14:26
apuimedoif we decide to translate for them as you did, we can then remove it from devstack14:26
apuimedo(if we get to add it at all)14:26
kzaitsev_wsI checked a couple of places where I could have left that code, and I can't find it anywhere =(14:28
kzaitsev_wswell, it was a hack anywat, but at least we know that it's possible14:28
kzaitsev_wscurrent Services handler only handles Services with Pods, which was the first step of my hack.14:30
apuimedokzaitsev_ws: not only it is possible. It was what kuryr-kubernetes did before the reimplementation that was done when moving it from Midokura PoC to openstack14:32
apuimedo:-)14:32
*** mattmceuen has joined #openstack-kuryr14:35
kzaitsev_wsregresssions regressions =)14:37
apuimedokzaitsev_ws: more like different decisions14:38
apuimedoit is not a bug14:38
apuimedoit is a feature!14:38
gsagieapuimedo: i hope you are not taking a connection to Israel as well :)14:54
apuimedogsagie: no, no14:54
apuimedostraight with Vueling14:54
apuimedoI only have layover in the customs waiting room14:55
apuimedoxD14:55
jgriffithThank you for driving this effort hongbin14:58
apuimedoindeed. Thank you hongbin14:59
gsagie:)15:01
kzaitsev_wswhere are you guys flying to? =)15:01
*** zengche has joined #openstack-kuryr15:02
hongbinjgriffith: apuimedo my pleasure15:02
apuimedokzaitsev_ws: Openstack Israel15:02
hongbinapuimedo: i think we could create a new repo (call it fuxi-go or something else), and por tthe code to it bit-by-bit15:03
kzaitsev_wshongbin: a separate branch maybe?15:05
apuimedohongbin: I thought about that too. I wonder if it would then be possible to move it over, or if there is some way to deprecate projects15:05
apuimedokzaitsev_ws: that could work15:05
hongbinkzaitsev_ws: yes, i guess that might work as well15:05
hongbinapuimedo: my concern is the amount of work to convert everything to golang, it might take a while15:07
kzaitsev_wsalthough it's always worth asking in #infra since they're always involved in all things, that touch repos in some way15:07
hongbinapuimedo: in addition, we still need to figure out how it works with additional runtime (i.e. hyper)15:08
apuimedohongbin: I think that fuxi (talking only about the docker volume api driver) has nothing to do with other runtimes15:10
hongbinapuimedo: ok, that is an option15:10
apuimedoas I said in the meeting, both things, k8s->openstack and docker->openstack look like separatable efforst15:11
hongbinapuimedo: ack15:11
hongbinapuimedo: yes, we could separate it15:11
hongbinapuimedo: ok, then we could start everything from scratch with golang for k8s15:12
hongbinapuimedo: and leave the old python code for docker?15:12
apuimedohongbin: it seems cinder would like to have the docker driver in golang too :-)15:14
hongbinapuimedo: i see15:14
apuimedoIt's quite a lot of work. But hopefully taking as a base jgriffith's docker driver and chaning it to fuxi api options would do it easier/cheaper15:15
hongbinok, we could try that15:15
hongbinapuimedo: we could do it step-by-step, take existing code and implement a flexvol driver in golang first15:17
hongbinapuimedo: then, implement the volume provisioner15:17
apuimedohongbin: as I said, separate efforts. Since the fuxi->golang would be keeping the API15:17
apuimedoboth things can be done at once15:17
apuimedos/at once/in parallel/15:18
hongbinapuimedo: ack15:18
hongbinapuimedo: ok, then let's do it :)15:19
apuimedo:-)15:20
hongbinapuimedo: let's try the fuxi->golang first?15:21
hongbinperhaps, we should list all hte TODO items15:22
apuimedohongbin: and make bps15:22
*** gianpietro has quit IRC15:23
*** gianpietro has joined #openstack-kuryr15:24
* hongbin is writing a bp15:24
*** yamamoto_ has quit IRC15:26
hongbinapuimedo: here you go: https://blueprints.launchpad.net/fuxi/+spec/convert-to-golang15:27
*** yamamoto has joined #openstack-kuryr15:27
apuimedothanks hongbin!15:27
hongbinapuimedo: besides the lauguage, do you think if there is any item that we should prioritize it for now?15:28
apuimedothe kubernetes handler15:30
apuimedofor volumes15:30
hongbini see15:32
hongbini think we agreed to do that in python?15:32
*** yamamoto has quit IRC15:32
hongbinlet me create another bp for that15:33
hongbin#link https://blueprints.launchpad.net/fuxi/+spec/k8s-volume-handler15:34
hongbinapuimedo: i need to leave for a while, ttyl15:35
*** hongbin has quit IRC15:36
*** zengche has quit IRC15:42
*** pcaruana has quit IRC15:47
*** kzaitsev_ws has quit IRC15:50
*** gsagie has quit IRC16:09
*** yamamoto has joined #openstack-kuryr16:30
*** salv-orlando has joined #openstack-kuryr16:31
*** gouthamr has joined #openstack-kuryr16:35
*** yamamoto has quit IRC16:39
*** ltomasbo is now known as ltomasbo|away16:44
*** egonzalez has quit IRC16:47
*** dimak_ has quit IRC16:50
*** gianpietro has quit IRC16:57
*** gianpietro has joined #openstack-kuryr16:57
*** aojea has joined #openstack-kuryr16:59
*** salv-orlando has quit IRC17:03
*** feisky has quit IRC17:14
*** salv-orlando has joined #openstack-kuryr17:19
*** salv-orlando has quit IRC17:23
*** aojea has quit IRC17:29
*** kzaitsev_mb has joined #openstack-kuryr17:30
*** kzaitsev_mb has quit IRC17:43
*** salv-orlando has joined #openstack-kuryr17:51
*** hongbin has joined #openstack-kuryr17:55
*** kzaitsev_mb has joined #openstack-kuryr17:59
openstackgerritOpenStack Proposal Bot proposed openstack/fuxi master: Updated from global requirements  https://review.openstack.org/45447418:00
*** tonanhngo has joined #openstack-kuryr18:42
*** salv-orlando has quit IRC18:46
*** egonzalez has joined #openstack-kuryr18:47
*** kzaitsev_mb has quit IRC18:48
*** gianpietro has quit IRC18:49
*** kzaitsev_mb has joined #openstack-kuryr18:49
*** tonanhngo has quit IRC18:49
*** salv-orlando has joined #openstack-kuryr18:56
*** kzaitsev_mb has quit IRC18:59
*** salv-orlando has quit IRC19:02
*** salv-orlando has joined #openstack-kuryr19:10
*** salv-orlando has quit IRC19:11
*** aojea has joined #openstack-kuryr19:29
*** aojea has quit IRC19:35
*** gianpietro has joined #openstack-kuryr19:41
jgriffithhongbin do you want me to take a shot at that or did you want to take it?19:42
jgriffithhongbin or even better, work on it together?19:42
jgriffithWRT the convert-to-golang bp19:42
hongbinjgriffith: sure, you can take it, i am happy to help if anything you need19:53
jgriffithhongbin ok, I’ll start on it; maybe sync up one morning later this week and make sure I’m not missing things :)20:12
jgriffithhongbin don’t you sleep?20:13
*** gianpietro has quit IRC20:25
*** gianpietro has joined #openstack-kuryr20:28
*** mattmceuen has quit IRC20:34
hongbinjgriffith: sorry, i am heading to a meeting room in before20:38
hongbinjgriffith: sure, we can do the sync up, and see how we can work together on that20:39
*** noname has joined #openstack-kuryr20:52
nonameHello, I got this channel link from openstack contribution wiki, and I would like to contribute for this project20:53
nonameI know python and shell scripting. And I have been working on linux since last six years. Can I contribute in any way.20:58
*** gouthamr has quit IRC21:05
*** noname has quit IRC21:13
*** salv-orlando has joined #openstack-kuryr21:50
*** gianpietro has quit IRC21:57
*** gianpietro has joined #openstack-kuryr22:03
*** gouthamr has joined #openstack-kuryr22:21
*** egonzalez has quit IRC22:43
*** salv-orlando has quit IRC22:48
*** garyloug has quit IRC23:05
*** aojea has joined #openstack-kuryr23:32
*** aojea has quit IRC23:37
*** hongbin has quit IRC23:45
*** salv-orlando has joined #openstack-kuryr23:59

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