Perl 6 - the future is here, just unevenly distributed

IRC log for #openstack-rally, 2015-03-02

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

All times shown according to UTC.

Time Nick Message
00:18 echoingumesh joined #openstack-rally
00:34 dmorita joined #openstack-rally
01:19 echoingumesh joined #openstack-rally
01:30 yingjun joined #openstack-rally
01:55 himangi joined #openstack-rally
02:22 echoingumesh joined #openstack-rally
02:35 mwagner_lap joined #openstack-rally
02:36 yingjun joined #openstack-rally
02:48 ilbot3 joined #openstack-rally
02:48 Topic for #openstack-rally is now ☁ Rally RoadMap: http://goo.gl/JZkmwY ☁ Rally IRC chat logs http://irclog.perlgeek.de/openstack-rally ☁ Key persons to ask:  boris-42, msdubov, rediskin, andreykurilin, amaretskiy  ☁ Documentation: https://rally.readthedocs.org/en/latest/ ☁ To publish changes to Rally:  https://rally.readthedocs.org/en/latest/contribute.html
03:37 echoingumesh joined #openstack-rally
03:38 panbalag joined #openstack-rally
04:30 himangi joined #openstack-rally
04:44 nkhare joined #openstack-rally
05:09 rdas joined #openstack-rally
05:13 yingjun joined #openstack-rally
05:19 yfried|afk joined #openstack-rally
05:37 echoingumesh joined #openstack-rally
05:46 yfried|afk joined #openstack-rally
05:54 tfreger joined #openstack-rally
06:27 anshul joined #openstack-rally
06:52 yfried|afk joined #openstack-rally
06:55 garyk joined #openstack-rally
06:56 yfried joined #openstack-rally
06:58 himangi joined #openstack-rally
07:05 yingjun_ joined #openstack-rally
07:08 psd joined #openstack-rally
07:31 neeti joined #openstack-rally
07:45 Miouge joined #openstack-rally
08:10 arxcruz joined #openstack-rally
08:17 yingjun joined #openstack-rally
08:22 fhubik_afk joined #openstack-rally
08:28 kairat_kushaev left #openstack-rally
08:36 yfried boris-42: "PhysicalNode.restart" doesn't exist yet, does it?
08:38 echoingumesh joined #openstack-rally
08:44 oanufriev joined #openstack-rally
08:50 kairat_kushaev joined #openstack-rally
09:00 boris-42 yfried: it doesn't exist
09:01 boris-42 yfried: but I don't think that it's hard to implement
09:01 msdubov_ joined #openstack-rally
09:02 yfried boris-42: it's not hard
09:03 yfried boris-42: the main issue is not to use ip/hostname since they are deployment specific
09:04 boris-42 yfried: actually what I was thinking is to use IP
09:04 boris-42 yfried: we can put it as parameter
09:04 boris-42 yfried: so user will need to provide id's of hosts that should be restarted
09:05 yfried boris-42: it's a little more complicated that. we should discuss this in specific spec/bp
09:05 yfried boris-42: I'm actually working on a private thing to inject HA events
09:05 boris-42 yfried: so the bigger picture is next
09:05 yfried boris-42: https://github.com/yfried-redhat/eventool
09:05 boris-42 yfried: I am workin on separating Rally from OpenStack
09:06 boris-42 yfried: e.g. you will be able to use it to generate load to something else
09:06 boris-42 yfried: so there will be benchmarks that use some inofrmation not related to OpenStack at all (like restarting hosts)
09:07 yfried boris-42: you might want to do more complicate logic than that, but fine
09:08 yfried boris-42: but aren't you working on being accepted into openstack?
09:08 boris-42 yfried: yep I am working and ?
09:08 boris-42 yfried: it will be simple as pervious to run workloads against OpenStack
09:09 boris-42 yfried: but you won't be forced to run task against deployment
09:10 yfried boris-42: oh
09:10 boris-42 yfried: so I am not removing case
09:10 boris-42 yfried: just adding new scope
09:10 boris-42 yfried: by improving our task validation mechanism
09:10 yfried boris-42: well, we can talk about "physical" node scenario later
09:10 yfried boris-42: as it's only valid with the new task your doing
09:10 boris-42 yfried: actually I think that it is right time
09:11 boris-42 yfried: you said that you won't more complicated behaviour
09:11 boris-42 yfried: the idea is of new task config is to make it possible to do on level of rally benchmark engine
09:11 boris-42 yfried: that operates with blocks of scenarios context sla and runners
09:11 kairat_kushaev boris-42: Hello, Boris
09:12 boris-42 yfried: do we miss something?
09:12 boris-42 yfried: related to HA testing ?
09:12 yfried boris-42: not that I'm aware of atm
09:12 boris-42 yfried: because HA testing is one of the most important case, that we would like to be able to cover by new task format
09:12 yfried boris-42: posted my comments to spec
09:12 boris-42 kairat_kushaev: hey hey
09:12 boris-42 yfried: thanks I will take a later
09:13 boris-42 yfried: a bit-=)
09:13 kairat_kushaev boris-42: I have a question about tests. Does VMBoot test fail only with my patch?
09:13 boris-42 take a look*
09:13 kairat_kushaev boris-42: Can I help somehow to improve it?
09:13 boris-42 kairat_kushaev: yep it fails only on your patch
09:13 boris-42 kairat_kushaev: nope it just fails on your patch
09:13 boris-42 kairat_kushaev: and it's very strange
09:13 yfried boris-42: my point is, that node-specific "scenarios" are unrelated to the task format, as you can add them now.
09:14 yfried boris-42: however, they only make sense if task format allows you to control the follow of when and how they are being executed
09:15 boris-42 yfried: so what we can control is in what second it can be executed
09:15 yfried boris-42: exactly
09:15 boris-42 yfried: like make load on keystone for 10 minutes
09:15 boris-42 yfried: and run every 2 minutes restart of one of contorlers
09:15 yfried boris-42: ack
09:16 boris-42 yfried: I think it's more then enough to test HA stuff
09:16 boris-42 yfried: as well we can add more parallel scenarios and test other stuff in the same scenario as well
09:16 yfried boris-42: I'm not arguing with you at all :)
09:16 yfried boris-42: we share this vision
09:16 yfried boris-42: however
09:17 yfried boris-42: have you looked at my link?
09:18 yfried boris-42: https://github.com/yfried-redhat/eventool/blob/master/etc/hosts_conf.yaml.sample
09:18 yfried boris-42: if we work with a deployment related conf-file like that
09:18 yfried boris-42: the scenarios can refer to a role/alias instead of an IP
09:18 kairat_kushaev boris-42: Does the problem with VMTasks look like a bug in rally?
09:19 kairat_kushaev boris-42: Or smth wrong in scenarios itself?
09:21 boris-42 yfried:  so I believe that this is only scenario related stuff
09:22 boris-42 yfried: so I mean nothing is blocking us to make such scenario
09:22 boris-42 yfried: yep?
09:22 boris-42 kairat_kushaev: no that scenarios works well
09:22 boris-42 kairat_kushaev: it is tested well in gates and for everbody else works well
09:25 kairat_kushaev boris-42: Hmm interesting... but how one test can affect another test in rally?
09:25 boris-42 kairat_kushaev: because in gates we are running one by one all tests including your new
09:26 boris-42 kairat_kushaev: so your new test can change somehow OpenStack resources that it doesn't work anymore
09:26 yfried boris-42: yep
09:28 olkonami joined #openstack-rally
09:30 boris-42 kairat_kushaev: so let's use logstach
09:31 boris-42 kairat_kushaev: uuu http://logs.openstack.org/46/155346/14/check/gate-rally-dsvm-rally/fff91d4/logs/screen-n-net.txt.gz?level=ERROR
09:33 rdas joined #openstack-rally
09:35 boris-42 kairat_kushaev: so use this http://logstash.openstack.org/#eyJzZWFyY2giOiJcIjxubyBpZD4gdG8gYmVjb21lIFwiIiwiZmllbGRzIjpbXSwib2Zmc2V0IjowLCJ0aW1lZnJhbWUiOiI2MDQ4MDAiLCJncmFwaG1vZGUiOiJjb3VudCIsInRpbWUiOnsidXNlcl9pbnRlcnZhbCI6MH0sInN0YW1wIjoxNDI1Mjg4Njk0NTU4fQ==
09:35 boris-42 kairat_kushaev: it finds 90 hits
09:48 pboros joined #openstack-rally
09:59 nmagnezi joined #openstack-rally
10:04 Miouge joined #openstack-rally
10:11 psd joined #openstack-rally
10:14 Miouge joined #openstack-rally
10:17 fhubik joined #openstack-rally
10:20 rvcehimanshu joined #openstack-rally
10:33 kairat_kushaev boris-42: I need some clarification for this scenario. Inside boot_server_with_fip method I see the self._boot_server method
10:34 kairat_kushaev boris-42: but i cannot find the method definition itself
10:34 kairat_kushaev boris-42: Looks like it should be called from nova_utils
10:38 boris-42 kairat_kushaev: hm? what where?
10:38 kairat_kushaev boris-42: the scenario is trying to ping the server but it fails
10:39 kairat_kushaev boris-42: by timeout if the server was not created
10:39 echoingumesh joined #openstack-rally
10:39 kairat_kushaev boris-42: Is it possible that smd forgot to inherit from nova_utils?
10:40 kairat_kushaev boris-42: I am not sure because it fails only for my patches
10:41 aix joined #openstack-rally
10:41 rvcehimanshu joined #openstack-rally
10:43 boris-42 kairat_kushaev: according to this logs something with network stuff was changed http://logs.openstack.org/46/155346/14/check/gate-rally-dsvm-rally/fff91d4/logs/screen-n-net.txt.gz?level=TRACE
10:43 boris-42 amaretskiy: ^
10:43 boris-42 akuznetsova: could you help kairat_kushaev
10:43 amaretskiy hi
10:45 akuznetsova boris-42, hi, what's the problem?
10:45 kairat_kushaev boris-42: But it could not: I create heat resources only (they are not related to any network resources). That's why I am asking for clarification.
10:46 kairat_kushaev akuznetsova: I need some clarification for this scenario. Inside boot_server_with_fip method I see the self._boot_server method  but i cannot find the method definition itself.
10:46 yingjun joined #openstack-rally
10:46 kairat_kushaev akuznetsova: Looks like it should be called from nova_utils
10:46 kairat_kushaev akuznetsova: Is it possible that my tests fail because of this reason
10:46 kairat_kushaev akuznetsova: http://logs.openstack.org/46/155346/14/check/gate-rally-dsvm-rally/fff91d4/rally-plot/results.html.gz
10:47 kairat_kushaev akuznetsova: You can also see a lot of references provided by Boris
10:47 openstackgerrit Mikhail Dubov proposed stackforge/rally: Unify RPS and constant runners and tune constant runner  https://review.openstack.org/155225
10:48 yingjun joined #openstack-rally
10:50 boris-42 akuznetsova: sorry*
10:51 boris-42 akuznetsova: I was trying to ping amaretskiy *
10:51 boris-42 amaretskiy: ^
10:51 boris-42 amaretskiy: could you help kairat_kushaev
10:51 amaretskiy boris-42 I'm currently inspecting logs
10:51 boris-42 amaretskiy: he is facing some issues with network in non related patch
10:51 amaretskiy kairat_kushaev hi
10:51 amaretskiy kairat_kushaev I'm working on that
10:52 kairat_kushaev amaretskiy: hi, ok
10:54 amaretskiy We have errors here http://logs.openstack.org/46/155346/14/check/gate-rally-dsvm-rally/fff91d4/rally-plot/results.html.gz#/NovaServers.boot_and_delete_server-4/failures
10:55 amaretskiy This seem to be related to the problem
10:59 amaretskiy this also looks suspicious, from console.html: 2015-03-02 00:02:08.603 | 2015-03-02 00:02:08.567 | 2015-03-02 00:02:08.543 24962 TRACE rally.benchmark.engine InvalidScenarioArgument: Invalid scenario argument: 'Image with pattern '^from_context_uploaded$' not found'
11:03 rvcehimanshu joined #openstack-rally
11:11 boris-42 amaretskiy: hm?
11:12 boris-42 https://www.irccloud.com/pastebin/EPTHFQON
11:14 rook__ joined #openstack-rally
11:18 psd_ joined #openstack-rally
11:19 yingjun joined #openstack-rally
11:27 openstackgerrit Pavel Boldin proposed stackforge/rally: Add FIO disk benchmark for VMs  https://review.openstack.org/152388
11:27 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the generic script-based benchmark for the VMs  https://review.openstack.org/141671
11:27 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the context custom_image  https://review.openstack.org/138466
11:29 amaretskiy boris-42, yes, looks server instance failed to spawn due to networking issue
11:29 amaretskiy I can't find the root cause for now, need to investigate
11:33 openstackgerrit Roman Vasilets proposed stackforge/rally: Adding Multiple Request Scenario  https://review.openstack.org/117705
11:37 rvcehimanshu joined #openstack-rally
11:40 openstackgerrit Pavel Boldin proposed stackforge/rally: Add FIO disk benchmark for VMs  https://review.openstack.org/152388
11:40 openstackgerrit Pavel Boldin proposed stackforge/rally: Add the generic script-based benchmark for the VMs  https://review.openstack.org/141671
11:42 yfried psd_:
11:42 psd_ yfried, Hey!
11:42 yfried psd_: are you Pavel Boldin?
11:43 psd_ yfried, haha no! I think he is pboldin or paboldin :)
11:43 yfried psd_: sorry
11:43 psd_ np :)
11:44 yfried boris-42: I wonder about "hidden" context - what does it mean?
11:50 andreykurilin yfried: hi! such context cann't be setup from user side (by yaml or json files)
11:51 paboldin joined #openstack-rally
11:51 paboldin i'm here
11:51 yfried andreykurilin: then why does it have input arguments? can it be set up from (for example) other contexts?
11:51 andreykurilin yfried: Pavel Boldin is here)
11:51 yfried andreykurilin: tnx
11:51 garyk joined #openstack-rally
11:51 andreykurilin yfried: It can be configured via decorator
11:52 paboldin yfried: I'm the ears now :)
11:52 paboldin yfried: the eyes, ofc, this is a text chat.
11:52 yfried paboldin: :)
11:52 yfried paboldin: wondering about your patch https://review.openstack.org/#/c/138466
11:52 yfried paboldin: and about hidden ctxt in general
11:53 paboldin yfried: yeah. what are the questions?
11:53 yfried paboldin: andreykurilin: so hidden ctxt is used as decorator? either it's allways used for a spcecific scenario, or it never is?
11:53 paboldin yfried: the problem with this is that I had to split it in two parts
11:54 paboldin yfried: and all the actual use is in the next patch
11:54 paboldin yfried: https://review.openstack.org/141671
11:54 yfried paboldin: no questions about uses yet
11:55 yfried paboldin: some general rally q
11:55 paboldin yfried: ok
11:55 paboldin yfried: hidden ctx is being inherited by the other context code
11:55 andreykurilin yfried: the example of hidden context  was TempestContext. There were no reasons to set up it from user side, so it was tempest scenario specific.
11:55 paboldin yfried: such as 'installer.py' context in the 141671
11:56 paboldin yfried: and the 'script_benchmark' there asks for it using a decorator
11:56 rvcehimanshu joined #openstack-rally
11:56 yfried paboldin: andreykurilin: also - if I subclass a hidden ctx will it be hidden as well?
11:56 paboldin yfried: but contexts are never setup based on the decorator, except for cleanup, allow_ssh (secgroup) and keypair contexts
11:57 yfried paboldin: lost you now ^
11:57 paboldin yfried: nope, if you will decorate it with base.context and leave hidden=False as default
11:57 yfried paboldin: ack
11:57 andreykurilin +1 to answer above:)
11:58 paboldin yfried: if you don't decorate the contextclass with base.context it won't be a context anyway.
11:58 paboldin yfried: (but i can be wrong on the discovery mechanism)
11:59 openstackgerrit Ravikumar Venkatesan proposed stackforge/rally: Adds Neutron floatingip tests  https://review.openstack.org/160312
11:59 paboldin yfried: yes, i'm wrong. it will be a context but it will be hidden
12:00 paboldin yfried: due to inherited _ctx_hidden variable
12:01 andreykurilin yfried, paboldin: but this variable can be overridden:)
12:01 paboldin andreykurilin: yes, but by default it is inherited
12:01 yfried paboldin: andreykurilin: thank you both
12:01 paboldin andreykurilin: and the name as well. so we really need to check for duplicates in get_by_name
12:02 paboldin andreykurilin: Context.get_by_name
12:02 andreykurilin paboldin: "but by default it is inherited". this is python way;)
12:02 paboldin andreykurilin: well, this can break a lot of code.
12:02 yfried paboldin: I suggest preventing implicit decorator inheritance
12:03 andreykurilin paboldin: yeah
12:03 paboldin yfried: because decorator only installs a few variables to the class itself it cannot be done at the moment.
12:04 paboldin yfried: or wee need to use some kind of metaclass magic, which is bad as well.
12:04 yfried paboldin: andreykurilin: well, at the very least, do it in hacking or in manual reviewing
12:05 yfried paboldin: now that I understand the risk, I'm gonna -1 inheritance like that
12:06 paboldin yfried: we will have to wait on boris-42 to talk about this issue
12:08 yfried paboldin: if you have multiple users in the same tenant, do you really need an image for each?
12:09 paboldin yfried: if they can't share the image then yes. only admin can make an image public
12:10 yfried paboldin: AFAIK, images are owned by tenant, so all users in the tenant share the same images
12:10 paboldin yfried: that is why we use utils.iterate_per_tenants
12:11 yfried paboldin: but it iterates over all users in the tenant, doesn't it?
12:11 paboldin yfried: only one user per tenant, afaik
12:12 paboldin yfried: and even this code is written incorrectly :-(
12:12 yfried ?
12:12 paboldin yfried: rally/common/utils.py like 435
12:12 paboldin yfried: spot an algorithmic error here
12:15 yfried paboldin: ?
12:15 paboldin yfried: it should be a `set()'
12:16 yfried paboldin: ack. fixing now
12:16 paboldin yfried: well, probably i'm wrong, but list takes O(n) for lookup
12:22 openstackgerrit Yair Fried proposed stackforge/rally: Fix iterate_per_tenants and adds docstring  https://review.openstack.org/160319
12:22 yfried paboldin: ^ added docstring for this method, as it's not trivial from name
12:22 psd__ joined #openstack-rally
12:24 Miouge joined #openstack-rally
12:27 e0ne joined #openstack-rally
12:28 yingjun_ joined #openstack-rally
12:39 tfreger1 joined #openstack-rally
12:40 echoingumesh joined #openstack-rally
12:44 panbalag joined #openstack-rally
12:47 rvcehimanshu joined #openstack-rally
13:01 himangi joined #openstack-rally
13:17 rvcehimanshu joined #openstack-rally
13:26 openstackgerrit Ivan Kolodyazhny proposed stackforge/rally: Implement create_and_upload_volume_to_image scenario for Cinder  https://review.openstack.org/152670
13:28 tfreger joined #openstack-rally
13:34 garyk joined #openstack-rally
13:42 psd_ joined #openstack-rally
13:44 exploreshaifali joined #openstack-rally
13:47 rvcehimanshu joined #openstack-rally
14:06 cdent joined #openstack-rally
14:10 yfried joined #openstack-rally
14:23 psd_ joined #openstack-rally
14:24 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add rally tests for Ceilometer  https://review.openstack.org/153994
14:29 yingjun joined #openstack-rally
14:41 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add rally tests for Ceilometer  https://review.openstack.org/153994
14:45 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add Ceilometer scenario for nova notifications  https://review.openstack.org/154803
15:25 psd_ joined #openstack-rally
15:27 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add Ceilometer scenario for nova notifications  https://review.openstack.org/154803
15:41 echoingumesh joined #openstack-rally
16:00 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add Ceilometer scenario for nova pollsters  https://review.openstack.org/160387
16:22 openstackgerrit Roman Vasilets proposed stackforge/rally: Adding Multiple Request Scenario  https://review.openstack.org/117705
16:24 rvcehimanshu joined #openstack-rally
16:29 rook__ joined #openstack-rally
16:47 openstackgerrit Merged stackforge/rally: Fix iterate_per_tenants and adds docstring  https://review.openstack.org/160319
16:53 yfried joined #openstack-rally
17:00 jaypipes joined #openstack-rally
17:41 echoingumesh joined #openstack-rally
17:54 yfried joined #openstack-rally
17:56 openstackgerrit Marian Krcmarik proposed stackforge/rally: Add scenarios for live migration of an instance with volumes  https://review.openstack.org/157844
18:02 psd_ joined #openstack-rally
18:05 boris-42 joined #openstack-rally
18:18 psd__ joined #openstack-rally
18:21 garyk joined #openstack-rally
18:32 rook joined #openstack-rally
18:33 rook sup boris-42 - so Alex is working on my BZ?
18:33 rook So less work for me. outstanding!
18:37 boris-42 rook: yep yep
18:37 boris-42 rook: I told him that he was wrong)
18:38 rook I will manually patch until he gets it out.
18:38 rook no problem.
18:38 boris-42 rook: by removing that functionallity
18:38 rook if someone wasn't working on it, i was going to push whatever I do us.
18:38 boris-42 rook: that a lot of people wants
18:38 boris-42 rook: so if you have something to push you can
18:39 boris-42 rook: and then you can collaborate with amaretsky
18:39 rook if need to write it, since I just noticed the us code is broke.
18:39 rook I*
18:39 boris-42 rook: ok you can as well just wait
18:39 rook i had to switch gears to CloudBench for a bit :D
18:39 boris-42 rook: btw did you see proposal for new input format?
18:39 boris-42 rook: heh=)
18:39 rook negative.
18:40 boris-42 rook: take a look https://review.openstack.org/#/c/159065/
18:40 boris-42 rook: the goal is to create format that will allow to do mulitiscenario load
18:41 rook # Without this line it's unclear what the hell it does
18:41 * rook giggles
18:42 boris-42 rook: =)
18:51 echoingumesh joined #openstack-rally
19:31 yfried joined #openstack-rally
19:32 e0ne joined #openstack-rally
19:35 himangi left #openstack-rally
19:37 jjmb1 joined #openstack-rally
19:42 psd_ joined #openstack-rally
19:43 boris-42 e0ne: hi
19:43 e0ne boris-42: hi
19:44 boris-42 e0ne: so about CinderVolumes
19:45 e0ne yep
19:47 psd__ joined #openstack-rally
19:50 boris-42 e0ne: what about finding better name?
20:00 e0ne joined #openstack-rally
20:01 e0ne boris-42: may be we need reorgonize scenarios?
20:01 e0ne boris-42: e.g.: components only, integrated
20:01 boris-42 e0ne: yep
20:02 boris-42 e0ne: as well it will be nice to split CinderVolume, CinderSnaphsot, Cinder....
20:02 boris-42 e0ne: benchmarks
20:02 boris-42 e0ne: what do you think?
20:02 e0ne boris-42: need to look on it. may be it will be too granular
20:03 boris-42 e0ne: so actually it can done after
20:03 boris-42 e0ne: could you address Pavel's comments?
20:03 e0ne boris-42: agree
20:03 boris-42 e0ne: and we will get this merged
20:03 e0ne boris-42: sure, i'll do it tonight or tomorrow morning
20:03 boris-42 e0ne: after that I will introduce depracation/rename mechanism
20:06 e0ne boris-42: sounds good
20:10 jjmb joined #openstack-rally
20:21 openstackgerrit Merged stackforge/rally: Add the context custom_image  https://review.openstack.org/138466
20:34 boris-42 jaypipes: pin
20:34 boris-42 jaypipes: ping
20:48 psd_ joined #openstack-rally
20:50 openstackgerrit Boris Pavlovic proposed stackforge/rally: Future improvment of plugin base  https://review.openstack.org/160142
20:50 openstackgerrit Boris Pavlovic proposed stackforge/rally: Switch to plugin base: context, sla, runners  https://review.openstack.org/150647
21:07 penguinRaider joined #openstack-rally
21:08 openstackgerrit Ivan Kolodyazhny proposed stackforge/rally: Implement create_and_upload_volume_to_image scenario for Cinder  https://review.openstack.org/152670
21:11 openstackgerrit Boris Pavlovic proposed stackforge/rally: Future improvment of plugin base  https://review.openstack.org/160142
21:11 openstackgerrit Boris Pavlovic proposed stackforge/rally: Switch to plugin base: context, sla, runners  https://review.openstack.org/150647
21:16 jaypipes boris-42: pong
21:18 boris-42 jaypipes: so you were interested in HA testing
21:18 boris-42 jaypipes: could you please review new rally input task format?
21:19 jaypipes boris-42: sure
21:19 boris-42 jaypipes: https://review.openstack.org/#/c/159065/
21:19 boris-42 jaypipes: thanks
21:19 boris-42 jaypipes: this was the most blocking part of work
21:24 openstackgerrit Boris Pavlovic proposed stackforge/rally: Future improvment of plugin base  https://review.openstack.org/160142
21:24 openstackgerrit Boris Pavlovic proposed stackforge/rally: Switch to plugin base: context, sla, runners  https://review.openstack.org/150647
21:32 jaypipes boris-42: done.
21:44 psd_ joined #openstack-rally
21:56 aix joined #openstack-rally
21:58 boris-42 jaypipes: replied on your comments
21:58 boris-42 jaypipes: and thanks for your comments
21:59 jaypipes boris-42: no prob! thx for inviting the feedback!
22:00 boris-42 jaypipes: so what do you think overall?
22:16 penguinRaider joined #openstack-rally
22:18 andreykurilin_ joined #openstack-rally
22:27 e0ne joined #openstack-rally
22:34 andreykurilin_ joined #openstack-rally
22:37 NikitaKonovalov joined #openstack-rally
22:38 e0ne joined #openstack-rally
22:53 e0ne joined #openstack-rally
22:54 openstackgerrit Ivan Kolodyazhny proposed stackforge/rally: Implement create_and_upload_volume_to_image scenario for Cinder  https://review.openstack.org/152670
23:03 echoingumesh joined #openstack-rally
23:24 echoingumesh boris-42: https://gist.github.com/anonymous/d64016df763d7bc9ecad on rally verify start
23:25 boris-42 echoingumesh: hi there
23:26 boris-42 echoingumesh: so are you going to refactor this code?)
23:26 boris-42 echoingumesh: I thought you are going to refactor this method for such cases where there are no subnets networks and so on to not fail
23:27 echoingumesh boris-42: 2 things.
23:27 echoingumesh boris-42: 1. I am planning to refactor the code so that the subnet issue is resolved.
23:28 echoingumesh boris-42: 2. This is a different error. I occurs even if there is a valid subnet.
23:28 boris-42 echoingumesh: =(
23:28 boris-42 echoingumesh: heh soo we need to fix as much as possible bugs in that piece of code=)
23:29 boris-42 echoingumesh: so this one means that there is no routers ..
23:29 echoingumesh boris-42: Gotcha.
23:29 boris-42 echoingumesh: maybe we need to create them?
23:29 boris-42 echoingumesh: automatically if there is no..
23:32 echoingumesh boris-42: ok. we should create a router. I will get back to you. I have to finish a meeting right now.
23:33 boris-42 echoingumesh: btw
23:33 boris-42 echoingumesh: did you see network context
23:33 boris-42 echoingumesh: this is how we set networks https://github.com/stackforge/rally/blob/master/rally/benchmark/context/network.py#L67-L69
23:33 boris-42 echoingumesh: maybe it makes sense to reuse this method
23:33 boris-42 echoingumesh: it will save a lot of time
23:38 openstackgerrit joined #openstack-rally
23:43 psd joined #openstack-rally

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