Perl 6 - the future is here, just unevenly distributed

IRC log for #openstack-rally, 2016-01-25

| Channels | #openstack-rally index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:38 chenli joined #openstack-rally
00:39 boris-42 joined #openstack-rally
01:10 bochi-michael joined #openstack-rally
01:19 yingjun joined #openstack-rally
01:57 baker joined #openstack-rally
01:59 Liuqing joined #openstack-rally
02:03 baker joined #openstack-rally
03:16 _lnk_ joined #openstack-rally
03:20 Deng joined #openstack-rally
03:25 harshs joined #openstack-rally
03:59 _lnk_ joined #openstack-rally
04:02 openstackgerrit li,chen proposed openstack/rally: Remove `None` as a redundant second argument to `dict.get()`  https://review.openstack.org/266684
04:14 baker joined #openstack-rally
04:15 baker_ joined #openstack-rally
04:23 rdas joined #openstack-rally
04:40 _lnk_ joined #openstack-rally
04:43 Liuqing joined #openstack-rally
04:46 bkopilov joined #openstack-rally
04:51 openstackgerrit Merged openstack/rally: Updated from global requirements  https://review.openstack.org/271633
04:57 harshs joined #openstack-rally
05:12 Liuqing joined #openstack-rally
05:18 _lnk_ joined #openstack-rally
05:28 boris-42 kun_huang: btw are you going to attend hackhathon ?)
05:37 tfreger joined #openstack-rally
05:46 shausy joined #openstack-rally
05:48 kun_huang boris-42: not 100% sure
05:49 anshul joined #openstack-rally
05:49 kun_huang current plan is to put all contributors to Chendu, another city of China
05:54 rcernin joined #openstack-rally
06:09 baker joined #openstack-rally
06:12 openstackgerrit Merged openstack/rally: Remove `None` as a redundant second argument to `dict.get()`  https://review.openstack.org/266684
06:14 Liuqing joined #openstack-rally
06:35 openstackgerrit Merged openstack/rally: Pass environment variables of proxy to tox  https://review.openstack.org/266639
06:37 _lnk_ joined #openstack-rally
06:56 anshul joined #openstack-rally
07:02 davidlenwell joined #openstack-rally
07:03 _lnk_ joined #openstack-rally
07:04 baker joined #openstack-rally
07:12 psuriset_ joined #openstack-rally
07:26 davidlenwell joined #openstack-rally
07:27 Liuqing joined #openstack-rally
07:33 davidlenwell joined #openstack-rally
07:41 psuriset_ joined #openstack-rally
07:45 davidlenwell joined #openstack-rally
08:00 openstackgerrit Tom Patzig proposed openstack/rally: Allow static fip in VMTasks.boot_runcommand_delete  https://review.openstack.org/271704
08:08 arxcruz joined #openstack-rally
08:10 Deng Could anyone please give me help about rally in local env. connecting with https server?
08:10 Deng Thanks so much!
08:10 boris-42 Deng: hi there
08:10 lezbar joined #openstack-rally
08:10 Deng hi
08:10 boris-42 Deng: you mean ssl keystone?
08:10 boris-42 Deng: of your cloud?
08:10 Deng Yes
08:11 boris-42 Deng: not sure why it makes sense how you have installed rally and  & https?)
08:11 boris-42 Deng: you should just set properly this https://github.com/openstack/rally/blob/master/samples/deployments/existing.json#L11-L12
08:11 Liuqing joined #openstack-rally
08:11 boris-42 Deng: that is all
08:16 Deng I will ask  our guys to change https to http, but i am not sure I can persuade them:)
08:20 Deng I have followed the exsiting.json and set "https_insecure" to  "true", but when I run the test cases, it always shows handshake error.
08:21 psuriset_ joined #openstack-rally
08:22 Deng When set "https_insecure" to  "false", it shows rally unable  connect with server.
08:23 Deng Do you mean we always use rally with http server, never use rally with https?
08:26 lezbar joined #openstack-rally
08:26 yfried_ joined #openstack-rally
08:29 Deng There is another question : do I should set the cacert path to "https_cacert": "" ?
08:30 Deng Thanks for your replying, boris.
08:32 psd joined #openstack-rally
08:32 boris-42 Deng: you need to set both
08:32 boris-42 Deng: and yep you need to put there path
08:32 boris-42 Deng: and don't forget to change address from http->https
08:32 boris-42 Deng: otherwise it won't work
08:34 yfried joined #openstack-rally
08:34 Deng OK, I will try these, thanks again.
08:41 adiantum joined #openstack-rally
08:46 boris-42 Deng: sure if you have any questions you are welcome
08:47 lezbar__ joined #openstack-rally
08:55 baker joined #openstack-rally
09:00 yfried_ joined #openstack-rally
09:01 yfried_ joined #openstack-rally
09:26 lezbar__ joined #openstack-rally
09:40 yingjun joined #openstack-rally
09:53 tfreger joined #openstack-rally
09:57 Liuqing joined #openstack-rally
09:59 openstackgerrit Tom Patzig proposed openstack/rally: Allow static fip in VMTasks.boot_runcommand_delete  https://review.openstack.org/271704
10:15 redixin joined #openstack-rally
10:19 DrugBug joined #openstack-rally
10:20 DrugBug does anyone know about rally test i can use that includes both neutron and nova clients at the same test?
10:29 vipul joined #openstack-rally
10:33 andreykurilin DrugBug: hi
10:33 openstackgerrit Allen Gao proposed openstack/rally: Remove pip <8 version pinning  https://review.openstack.org/272002
10:35 vipul joined #openstack-rally
10:35 andreykurilin DrugBug: what kind of test do you want? Create network, boot vm? or something like create several vms and launch iperf between them?
10:37 vipuls joined #openstack-rally
10:43 tosky joined #openstack-rally
10:45 baker joined #openstack-rally
10:47 rdas joined #openstack-rally
10:51 vipul joined #openstack-rally
10:57 bapalm joined #openstack-rally
11:06 openstackgerrit Raphaël Badin proposed openstack/rally: Add note about the fact that Rally won't use testtools' assertions  https://review.openstack.org/263759
11:28 zerda joined #openstack-rally
11:41 ilbot3 joined #openstack-rally
11:41 Topic for #openstack-rally is now ☁ Rally RoadMap: http://goo.gl/JZkmwY ☁ Review Dashboard: http://goo.gl/YuCJfl ☁ IRC logs http://irclog.perlgeek.de/openstack-rally ☁ Key persons to ask:  boris-42, yfried, msdubov, rediskin, andreykurilin, amaretskiy  ☁ Documentation: https://rally.readthedocs.org/en/latest/ ☁ To publish changes to Rally:  https://rally.readthedocs.org/en/latest/contribute.html
11:57 lgreg joined #openstack-rally
12:10 openstackgerrit Tom Patzig proposed openstack/rally: Allow static fip in VMTasks.boot_runcommand_delete  https://review.openstack.org/271704
12:20 DrugBug joined #openstack-rally
12:24 arxcruz joined #openstack-rally
12:31 openstackgerrit Tom Patzig proposed openstack/rally: Allow static fip in VMTasks.boot_runcommand_delete  https://review.openstack.org/271704
12:33 openstackgerrit Raphaël Badin proposed openstack/rally: Add note about the fact that Rally won't use testtools' assertions  https://review.openstack.org/263759
12:34 baker joined #openstack-rally
12:37 lgreg joined #openstack-rally
12:40 lgreg joined #openstack-rally
12:44 akrzos joined #openstack-rally
12:52 gubouvier_ joined #openstack-rally
13:00 openstackgerrit Illia Khudoshyn proposed openstack/rally: Integrate Rally & Alembic  https://review.openstack.org/264701
13:07 baker joined #openstack-rally
13:28 MaxPC joined #openstack-rally
13:41 diogogmt joined #openstack-rally
13:56 openstackgerrit Sergey Skripnick proposed openstack/rally: Heat helper class  https://review.openstack.org/272101
14:01 chris_19 joined #openstack-rally
14:02 boris-42 kun_huang yingjun meeting time
14:16 luz_ joined #openstack-rally
14:37 yingjun joined #openstack-rally
14:38 yingjun joined #openstack-rally
14:41 absubram joined #openstack-rally
14:43 yingjun joined #openstack-rally
14:53 absubram hi all.. quick debugging question.. are there any debugging tools in rally available to figure out issues I”m having with a new plugin that I’m working on?
14:54 absubram I’m getting some TypeErrors but I’m not sure what/ where it is and I’d like to know if there’s a nicer way than guessing and going with trial and error
14:54 stpierre joined #openstack-rally
14:57 boris-42 absubram: paste the error
15:09 absubram boris-42: Hi Boris!
15:09 absubram 2016-01-25 11:21:41.117 15056 INFO rally.task.runner [-] Task f4b85a3d-8577-4ef3-bf40-093ffdd49f2a | ITER: 0 START
15:09 absubram 2016-01-25 11:21:41.120 15056 INFO rally.task.runner [-] Task f4b85a3d-8577-4ef3-bf40-093ffdd49f2a | ITER: 0 END: Error TypeError: unexpected type <type 'dict'> for addr arg
15:10 absubram that’s what I’m seeing.. and I know it’s definitley somehting in the way I’ve coded something..
15:10 absubram 2016-01-25 11:21:41.120 15056 INFO rally.task.runner [-] Task f4b85a3d-8577-4ef3-bf40-093ffdd49f2a | ITER: 0 END: Error TypeError: unexpected type <type 'dict'> for addr arg
15:12 absubram are there ways to add debug messages in the plugin that I can look at?
15:15 stpierre yes, you can add logging
15:15 stpierre depending on where the error was raised, the full traceback should be in the task output
15:16 absubram stpierre: that’s great!.. I don’t see any tracebacks on the screen.. and I am using the -v option
15:16 absubram is there some other location where the task output is captured?
15:16 stpierre yeah, the task report
15:16 stpierre sorry, that's what i meant
15:16 stpierre s/output/report/
15:17 absubram ah ok
15:17 absubram rally task report <task-id> ?
15:17 stpierre yeah
15:18 absubram ah.. ok.. yep I see it.. thanks much!
15:20 baker joined #openstack-rally
15:31 absubram ah I see the problem.. I’m trying to boot an instance using a network created via neutron.. however when I invoke self._create_network from nova/servers.py - it’s picking up the _Create_network from nova/utils.py instead of neutron/utils.py
15:31 absubram so it’s using nova networks when I have neutron..
15:32 absubram I do have this line included in my plugin -     @validation.required_services(consts.Service.NOVA, consts.Service.NEUTRON)
15:32 absubram but it’s using the nova version instead of the neutron version
15:34 chris_19 joined #openstack-rally
15:43 openstackgerrit Illia Khudoshyn proposed openstack/rally: Integrate Rally & Alembic  https://review.openstack.org/264701
15:44 boris-42 absubram: actually
15:44 boris-42 absubram: no matter weather you are using nova api or neutron api
15:44 boris-42 absubram: if you have neutron it will go to neutron (via nova or directly)
15:45 chris_19 When trying to install rally, i get an error saying that tox<=1.6.1 cannot be found. Is there an easy fix? Removing the version requirement just seems to produce a cascade of other dependency errors (PrettyTable being next).
15:45 boris-42 chris_19: oh finally you are online and I am online
15:46 chris_19 ha
15:46 boris-42 chris_19: so I didn't see before that error
15:46 boris-42 chris_19: what is version of "pip" that you have?
15:47 chris_19 7.1.2, python 2.7
15:47 boris-42 chris_19: heh that should work well
15:47 boris-42 stpierre: ^ did you face this anytime?
15:47 chris_19 boris-42: I thought so
15:48 stpierre absubram: it's a bad idea to modify nova/servers.py directly. create your own plugin, and then you can either order the MRO however you want, or create your own _create_servers() that does explicit inheritance
15:49 stpierre boris-42: no, that's new to me
15:49 absubram boris-42: the tracebacks seem to indicate they are using the nova version for some reason.. but I need this for some internal testing so I’ve commented out those nova network versions.. and it’s now using neutron :)
15:49 absubram so I’m all set for now.. ty
15:49 chris_19 boris-42 I can go through and edit the dependencies one at a time, but it always seems like something fails. The tox directive is especially confusing since tox<=1.6.1 isn't available on pypi
15:49 boris-42 сchris_19 btw
15:49 boris-42 chris_19: are you using clean OS?
15:49 chris_19 boris-42 yes
15:50 boris-42 chris_19: and which one?
15:50 absubram stpierre: I understand.. I am trying to create my own version of _boot_Servers essentially :)
15:50 chris_19 boris-42 Centos 7. I usually try to install rally shortly after a fresh Vagrant deployment
15:50 absubram I won’t create noise to distract from chris_19’s questions.. I seem to be all set for now
15:51 chris_19 absubram didn't mean to interrupt :)
15:52 boris-42 absubram: it's okay just continue
15:52 boris-42 chris_19: hm Vagrant deployment of OpenStack?
15:53 boris-42 chris_19: or just vagrants boots VM ? and what OS do you use?
15:53 chris_19 boris-42 no. Vagrant deployment of CentOS 7 VM.
15:54 boris-42 chris_19: after that you are trying to run install_rally.sh and it fails?
15:54 chris_19 boris-42 correct
15:55 boris-42 chris_19: that sux =(
15:56 chris_19 boris-42 agreed :)
15:56 boris-42 chris_19: so it used to work =(
15:56 diogogmt joined #openstack-rally
15:56 boris-42 chris_19: can you pull the latest version of rally and try again?
15:57 chris_19 at some point, yes. like i said, tox<=1.6.1 isn't on pypi, so i'm so I'm confused as to why that's in the install-rally.sh script
15:57 chris_19 sure
15:57 chris_19 wish i'd thought of that :P
15:57 boris-42 chris_19: actually it is
15:57 chris_19 boris-42 hmm ok. I didn't see it listed, but i may just not have looked hard enough
16:00 boris-42 chris_19: I mean Rally is not DAT broken
16:00 boris-42 chris_19: you see https://pypi.python.org/pypi/tox/1.6.1
16:00 chris_19 boris-42 how do i install from the git clone? setup.py file?
16:00 boris-42 chris_19: there is install_rally.sh script
16:01 boris-42 chris_19: inisde repo
16:01 chris_19 boris-42 Great.
16:01 chris_19 boris-42 and yes I see that. I guess when I looked up tox on pypi, I didn't see that version so I assumed it wasn't there.
16:02 boris-42 chris_19: it's very strange error..
16:02 boris-42 chris_19: maybe you are under some proxy?
16:02 boris-42 chris_19: or you have local pypi server ?
16:03 chris_19 Could not find a version that satisfies the requirement tox<=1.6.1 (from versions: 2.0.0, 2.0.1, 2.0.2, 2.1.0, 2.1.1, 2.2.0, 2.2.1, 2.3.0, 2.3.1)
16:03 chris_19 boris-42 So, that's a thought. Let me try something
16:06 chris_19 boris-42 So I thought it may be because I was on my company's VPN, so I disconnected from that, but then I go the same error.
16:07 chris_19 boris-42 trying again after deleting the cache
16:08 boris-42 chris_19: btw it is good question why we are using such old tox.
16:09 chris_19 boris-42 it definitely downloads pip from pypi, but still fails at tox. weird.
16:09 boris-42 chris_19: so it doesn't fail in my case=(
16:09 boris-42 magic
16:09 boris-42 =)
16:09 chris_19 lol
16:10 boris-42 chris_19: actually you are quite rare person=)
16:10 boris-42 chris_19: you can be proud of yourself=)
16:10 chris_19 boris-42 hah yeah I have a track record of having issues that people on IRC can't help with :P
16:11 chris_19 boris-42 going to try one more time with some DNS changes
16:16 chris_19 boris-42: ok, i ran an install on my host machine, and it worked fine. There must be something about the setup of the Vagrant VM that is getting in the way
16:16 boris-42 chris_19: I believe it is wrong pypi
16:16 tosky joined #openstack-rally
16:16 boris-42 server
16:16 chris_19 probably
16:17 chris_19 although i would have thought getting off of the VPN would have fixed that
16:18 chris_19 boris-42 there's an internal server in /etc/pip.conf. I bet that's it
16:21 chris_19 thanks
16:23 harshs joined #openstack-rally
16:24 psd joined #openstack-rally
16:54 bkopilov joined #openstack-rally
16:55 chris_19 boris-42: is there a way to pass pypi as the index server to the install script?
17:34 psuriset_ joined #openstack-rally
17:56 KimJ joined #openstack-rally
18:01 openstackgerrit Castulo J. Martinez proposed openstack/rally: Re-run failed Tempest tests  https://review.openstack.org/244361
18:24 madhu_ak joined #openstack-rally
18:30 harshs joined #openstack-rally
18:41 yfried_ joined #openstack-rally
18:41 pvaneck joined #openstack-rally
18:51 josdotso joined #openstack-rally
18:51 pcaruana joined #openstack-rally
18:52 josdotso Hi boris-42.  Just checking in on progress toward "framework-generalization" ...  Or in specific case, the ability to run Rally tests without a hard dependency on openstack REST APIs being present.
19:02 openstackgerrit joined #openstack-rally
19:05 boris-42 josdotso: so we have some progress ikhudoshyn made integrated rally & alembic
19:05 boris-42 josdotso: so I am umblock to fix DB schema
19:06 pvaneck_ joined #openstack-rally
19:08 chris_19 joined #openstack-rally
19:10 anshul joined #openstack-rally
19:21 pvaneck joined #openstack-rally
19:36 pcaruana joined #openstack-rally
19:39 josdotso Thanks boris-42.  Maybe we need like a null deployment engine.. So I can skip Openstack deployment altogether and import my tests from tempest to build scenarios.. I'm guessing this won't work in reality, however
19:43 pcaruana joined #openstack-rally
19:44 rcernin joined #openstack-rally
19:45 chris_19 joined #openstack-rally
19:49 chris_19 joined #openstack-rally
20:01 baker_ joined #openstack-rally
20:08 linuxgeek joined #openstack-rally
20:17 openstackgerrit Piyush Raman Srivastava proposed openstack/rally: Adds support for listing subnets, routers and floatingIPs to NeutronWrapper class and listing floatingIPs to NovaNetworkWrapper. This extends capability for NetworkWrapper in Rally  https://review.openstack.org/272251
20:27 boris-42 josdotso: ya basically because of validation part
20:46 chris_19 joined #openstack-rally
21:03 chris_19 joined #openstack-rally
21:16 chris_19 joined #openstack-rally
21:18 davidlenwell joined #openstack-rally
21:26 stevelle joined #openstack-rally
21:32 stevelle Am I correct in assuming that new specs for Mitaka would not be accepted at this point?
21:33 stevelle thinking of a spec for new scenarios being introduced
21:34 chris_19 joined #openstack-rally
21:35 lutostag joined #openstack-rally
21:40 chris_19 joined #openstack-rally
21:42 phd2 joined #openstack-rally
21:56 chris_19 joined #openstack-rally
21:56 davidlenwell joined #openstack-rally
22:02 aix_ joined #openstack-rally
22:05 openstackgerrit Boris Pavlovic proposed openstack/rally: Move "tempest concurrency" feature request to implement  https://review.openstack.org/272294
22:06 chris_19 left #openstack-rally
22:08 boris-42 stevelle: nope you are not correct
22:09 boris-42 stevelle: first of all you don't need a spec for new scenarios
22:09 boris-42 stevelle: second thing is that we don't have feature freeze and so on
22:09 boris-42 stevelle: but it will be good to discuss what plugin you would like to add?
22:10 stevelle boris-42: great, so the scenarios involve using gnocchi as dispatcher / storage for ceilometer.
22:10 boris-42 stevelle: so rally is going to directly call gnocchi?
22:10 stevelle boris-42: for API query load, instead of querying ceilometer API yes
22:11 stevelle secondly ceilometer config needs to be tailored to the scenario
22:11 boris-42 stevelle: ok
22:11 diogogmt joined #openstack-rally
22:11 boris-42 stevelle: sounds good
22:13 stevelle boris-42: thanks much, I'll focus our energy on doing the work instead
22:14 boris-42 stevelle: yep I don't like bureaucracy when it's not required =)
22:15 stevelle boris-42: so because I'm really not familiar, this sounds like some work would likely be needed in rally but the scenarios would be added to ceilometer. is that right?
22:16 boris-42 stevelle: you can do complete work in ceilometer or in gnocchi or in rally
22:16 boris-42 stevelle: it depends on what you want to get as a result
22:16 boris-42 stevelle: so if you add scenarios to rally then it will be easy to run them everywhere
22:16 stevelle boris-42: and to that, I'm afraid I don't really have an answer yet
22:17 stevelle what do we want to do?
22:17 boris-42 stevelle: so does it make sense for operators and devops to have such tests? and would you like to run this in ceilometer and gnocchi CI?
22:19 stevelle boris-42: I think a non-voting check in CI would be helpful down the road. That isn't where we want to start, though.
22:19 stevelle our scenarios are concerned with the integration between the two projects which is what makes it a little more complex
22:21 boris-42 stevelle: we can try to implement it in Rally
22:21 boris-42 stevelle: btw Ceilometer had before Rally job
22:21 boris-42 stevelle: or it is in experimenetal queue maybe now
22:22 stevelle I believe I found a ceilometer spec for a non-voting check, maybe it was liberty or kilo.
22:29 boris-42 stevelle: we didn't add spec as far as I remember
22:29 boris-42 stevelle: it's actually openstack infra patch
22:30 stevelle boris-42: that sounds like it makes sense, being an infra patch. I would just think it might make sense to offer a spec if the telemetry community wants to discuss additional checks.
22:31 stevelle let them refuse the spec when we're ready
22:32 boris-42 stevelle: it's already there https://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L4044
22:32 boris-42 stevelle: you can write "check experimental" in review of any ceilometer patch and it will trigger ceilometer job
22:33 stevelle boris-42: thanks, you navigate this much quicker than I do.
22:36 davidlenwell joined #openstack-rally
22:59 openstackgerrit Allen Gao proposed openstack/rally: Remove pip <8 version pinning  https://review.openstack.org/272002
23:13 openstackgerrit Boris Pavlovic proposed openstack/rally: Move "tempest concurrency" feature request to implemented  https://review.openstack.org/272294
23:55 psuriset_ joined #openstack-rally

| Channels | #openstack-rally index | Today | | Search | Google Search | Plain-Text | summary