Perl 6 - the future is here, just unevenly distributed

IRC log for #openstack-rally, 2015-02-25

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

All times shown according to UTC.

Time Nick Message
00:00 meteorfox boris-42: are you doing a lot of IO or something? I usually see kswapd at 100% when either I ran out of memory and the OS is swapping, or heavy IO which then has to write back pages to disks
00:01 boris-42 meteorfox: i just reboot it
00:01 abhirc joined #openstack-rally
00:02 boris-42 meteorfox: going to get new VM=)
00:02 meteorfox boris-42: ok, I run linux, because I get containers natively, and I want to code in the same OS that I write programs for
00:03 meteorfox :)
00:03 boris-42 meteorfox: and I don't like such kind of sex
00:03 boris-42 meteorfox: so I am using MacOs
00:03 boris-42 meteorfox: btw Rally works on it=)
00:04 meteorfox boris-42: cool!
00:04 meteorfox boris-42: so i actually run Linux on MacBookPro *shamelessly*
00:04 boris-42 meteorfox: uhoh
00:06 meteorfox boris-42: but for my performance work, I need Linux, for the tools. I constantly needed to try things in a Linux box, and I didn't want to spin up a VM
00:06 meteorfox is too slow, and uses to much memory
00:06 meteorfox too*
00:07 boris-42 meteorfox: hm but what about brew?
00:13 meteorfox boris-42: still they don't have all the things I need, because they are Linux specific
00:14 meteorfox boris-42: like perf tool, ftrace etc
00:14 boris-42 meteorfox: heh
00:37 penguinRaider joined #openstack-rally
00:46 penguinRaider joined #openstack-rally
00:49 penguinRaider joined #openstack-rally
00:50 penguinRaider boris-42, hi :-) around ?
00:52 boris-42 penguinRaider: oh hi
00:52 dmorita joined #openstack-rally
00:56 abhirc_ joined #openstack-rally
00:56 penguinRaider boris-42, long time :-), Openstack participating in GSOC again ! nice
00:56 boris-42 penguinRaider: =)
00:57 penguinRaider boris-42, so did GSOC people started flocking around here trying to solve those small bugs :-p ?
00:58 boris-42 penguinRaider: yep
00:58 boris-42 =)
00:58 boris-42 penguinRaider: 2 of them=)
00:58 penguinRaider boris-42, two so early lol nice :-) !!!!
00:59 penguinRaider boris-42, more to come your way then :-)
00:59 penguinRaider boris-42, wish I could get involved with the GSOC process this time too some way :-/
01:12 boris-42 meteorfox: if I make decision to make common context between multple scenarios
01:12 boris-42 meteorfox: somebody will find case where he need separated context=(
01:12 boris-42 meteorfox: ah I found the way
01:24 penguinRaider boris-42, you think I can get involved with the gsoc process someway this time ?
01:24 penguinRaider boris-42, I dont think I will be able to do it full time as student this time :-/
01:24 frickler joined #openstack-rally
01:24 boris-42 penguinRaider: heh I don't know
01:24 boris-42 penguinRaider: not an expert ..
01:24 boris-42 penguinRaider: you should ask dims
01:24 penguinRaider boris-42, maybe I can help people around here with small bugs and getting started
01:24 penguinRaider boris-42, ^ I mean the new gsoc people who come here
01:29 boris-42 penguinRaider: sure you can
01:29 boris-42 penguinRaider: that will be nice
01:52 dmorita_ joined #openstack-rally
01:52 frickler joined #openstack-rally
01:52 penguinRaider joined #openstack-rally
01:52 openstackgerrit joined #openstack-rally
01:52 garyk joined #openstack-rally
01:52 yfried joined #openstack-rally
01:52 harlowja joined #openstack-rally
01:52 afazekas joined #openstack-rally
01:52 klindgren joined #openstack-rally
01:52 dmellado joined #openstack-rally
01:52 ibiris_away joined #openstack-rally
01:52 aarefiev joined #openstack-rally
01:52 boris-42 joined #openstack-rally
01:52 vaidy_ril joined #openstack-rally
01:52 zhiyan joined #openstack-rally
01:52 akuznetsova joined #openstack-rally
01:52 jlk joined #openstack-rally
01:52 ilyashakhat joined #openstack-rally
01:52 meteorfox joined #openstack-rally
01:52 verdurin joined #openstack-rally
01:52 mkoderer joined #openstack-rally
01:52 tellesnobrega joined #openstack-rally
01:52 frobware joined #openstack-rally
01:52 rvasilets__ joined #openstack-rally
01:52 jseutter joined #openstack-rally
01:52 andreykurilin joined #openstack-rally
01:52 mitz joined #openstack-rally
01:52 rmk joined #openstack-rally
01:52 avico- joined #openstack-rally
02:06 boris-42 penguinRaider: maybe tomororow?)
02:06 boris-42 penguinRaider: and yep we have a lot of task=)
02:09 harlowja_ joined #openstack-rally
02:12 mwagner_lap joined #openstack-rally
02:16 abhirc joined #openstack-rally
02:22 penguinRaider joined #openstack-rally
02:42 himangi joined #openstack-rally
02:49 ilbot3 joined #openstack-rally
02:49 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
03:39 rook joined #openstack-rally
05:08 chen123 joined #openstack-rally
05:15 psd joined #openstack-rally
05:21 Mehrnaz joined #openstack-rally
05:21 Mehrnaz You can't imagin what had happen to our open stack!
05:21 Mehrnaz imagine*
05:21 neeti joined #openstack-rally
05:22 Mehrnaz when I commaned " rally task start /path/to/scenario/tempest/single_tes.json", all of our instances were terminated....Why????????
05:22 Mehrnaz boris-42: ^^^^
05:23 Mehrnaz andreykurilin: ^^^^^
05:24 Mehrnaz and we are so nervous...just imagine! All of our instances were terminated completely!!!
05:24 Mehrnaz in that test code, I didn't find any thing that relates to instance! It was just about images!
05:25 Mehrnaz boris-42:  are you there?
05:25 Mehrnaz :(
05:25 rdas joined #openstack-rally
05:27 rook joined #openstack-rally
05:37 tfreger joined #openstack-rally
05:38 penguinRaider joined #openstack-rally
05:43 abehl joined #openstack-rally
05:50 Mehrnaz_ joined #openstack-rally
05:51 Mehrnaz_ boris-42: Did you read my message here?
05:52 Mehrnaz_ You can't imagin what had happen to our open stack!
05:52 exploreshaifali joined #openstack-rally
05:52 Mehrnaz_ imagine*
05:52 Mehrnaz_ when I commaned " rally task start /path/to/scenario/tempest/single_tes.json", all of our instances were terminated....Why????????
05:52 Mehrnaz_ and we are so nervous...just imagine! All of our instances were terminated completely!!!
05:53 Mehrnaz_ in that test code, I didn't find any thing that relates to instance! It was just about images!
05:58 himangi joined #openstack-rally
06:33 exploreshaifali andreykurilin: around ?
06:36 boris-42 Mehrnaz_: hithere
06:36 boris-42 Mehrnaz_: sorry I was sleeping
06:36 Mehrnaz_ boris-42: hi
06:36 Mehrnaz_ boris-42: ok, np. ;)
06:36 boris-42 Mehrnaz_: so
06:37 boris-42 Mehrnaz_: don't use tempest inside rally
06:37 Mehrnaz_ boris-42: did you read my message? I think it's dilemma!
06:37 Mehrnaz_ why?
06:37 boris-42 Mehrnaz_: I mean not "rally verify command"
06:38 boris-42 Mehrnaz_: but don't run benchmarks of "tempest"
06:38 boris-42 Mehrnaz_: because this is tempest cleanup
06:38 boris-42 Mehrnaz_: not rally
06:38 boris-42 Mehrnaz_: it deletes everything
06:38 boris-42 Mehrnaz_: and as far as tempest is creating resources we are not able to do cleanup by rally mechanism
06:39 Mehrnaz_ I wish you told this on rally document before we use it for benchmarking .... :(
06:40 boris-42 Mehrnaz_: You can use Rally for benchmarking
06:40 boris-42 Mehrnaz_: everything except tempest based benchmarks
06:40 Mehrnaz_ So why you put tempest in scenarios?
06:41 boris-42 Mehrnaz_: to show tempest team that tempest is not consumable for benchmarking*
06:41 boris-42 Mehrnaz_: especially in production
06:41 boris-42 Mehrnaz_: Tempest wish to split Rally to few parts and integrate it in Tempest
06:42 boris-42 Mehrnaz_: so Like we won't duplicate their efforts
06:42 boris-42 Mehrnaz_: and I disagree with this point
06:42 boris-42 Mehrnaz_: because as you see it's not safe to use tempest for this purpose..
06:43 Mehrnaz_ OMG! we lost every thing...every instances.... It's terrible...we may lose our customers...
06:43 boris-42 Mehrnaz_: ooOO fuck
06:43 boris-42 Mehrnaz_: really sorry about that=(
06:44 garyk joined #openstack-rally
06:44 Mehrnaz_ thanks ... we won't use tempest at all..
06:45 anshul joined #openstack-rally
06:45 boris-42 Mehrnaz_: so I think I have to add special key
06:45 boris-42 Mehrnaz_: "rally task start --force"
06:45 boris-42 Mehrnaz_: for such kinds of benchmarks
06:45 Mehrnaz_ boris-42: I do agree...
06:47 boris-42 Mehrnaz_: okay let me add but to Rally for this
06:47 boris-42 Mehrnaz_: we will fix this ASAP so it won't repeat
06:49 yfried joined #openstack-rally
06:53 Mehrnaz_ boris-42: thanks for your response. I hope so ... :)
06:53 yfried_ joined #openstack-rally
06:56 nkhare joined #openstack-rally
06:56 openstackgerrit Boris Pavlovic proposed stackforge/rally: Add warnings to not use Tempest benchmarks against production cloud  https://review.openstack.org/159017
06:56 boris-42 Mehrnaz_: ^ would this help you to avoid such situation?
06:57 boris-42 Mehrnaz_: actually hopefully soon I will drop requriments of admin for most of benchmarks in Rally
06:57 boris-42 Mehrnaz_: so you will be able to benchmark most of stuff just using existing users with setted quota and so on
06:58 boris-42 Mehrnaz_: so that will be the most safe way for production
06:59 exploreshaifali boris-42: morning :)
06:59 exploreshaifali boris-42: to fetch data of n tasks I need to do something like https://github.com/stackforge/rally/blob/master/rally/cmd/commands/task.py#L537-L543
06:59 exploreshaifali boris-42: right ?
07:00 boris-42 exploreshaifali: it's enough https://github.com/stackforge/rally/blob/master/rally/cmd/commands/task.py#L543
07:00 boris-42 exploreshaifali: as you see it's mapped
07:00 Mehrnaz_ boris-42: I'll see it. I think by considering your edition, it's better that we define new user and new deployment that does'nt need admin privillage, in isolate zone and isolate project!
07:01 boris-42 Mehrnaz_: so in such case you won't be able to run these tempest scenarios at all
07:01 boris-42 Mehrnaz_: because they requires admin
07:02 Mehrnaz_ boris-42: ok!
07:02 exploreshaifali boris-42: yes, then to generate trends I need to create a module rally/benchmark/processing/trend.py
07:05 exploreshaifali boris-42: I am searching for D3.js code in rally
07:05 exploreshaifali boris-42: how is it used for plot
07:06 boris-42 exploreshaifali: this part is actually underconstruction
07:07 boris-42 exploreshaifali: https://review.openstack.org/#/c/146814/
07:07 boris-42 exploreshaifali: so you should search for rally/ui stuff
07:07 boris-42 exploreshaifali: it's mako template that is used for generation of stuff
07:07 boris-42 exploreshaifali: HTML pages with JS
07:07 boris-42 exploreshaifali: and data is passed from processing/
07:08 exploreshaifali boris-42: okay, Thanks :)
07:11 exploreshaifali boris-42: for writing code to generate trends I should read plot.py module https://github.com/stackforge/rally/blob/master/rally/benchmark/processing/plot.py#L10 right ?
07:11 boris-42 exploreshaifali: actually new code that is on review is better
07:11 boris-42 exploreshaifali: so you can cherry pick it and read
07:12 exploreshaifali boris-42: coooool :D
07:12 boris-42 exploreshaifali: so you need to understand how works StackedAreaa
07:12 boris-42 exploreshaifali: and use it few times=)
07:12 boris-42 exploreshaifali: with various data
07:15 exploreshaifali boris-42: okay, how  stackedArea works with D3.js...... right?
07:16 boris-42 exploreshaifali: we are actually using this http://nvd3.org/
07:16 boris-42 exploreshaifali: not directly d3js
07:16 exploreshaifali boris-42: okay :)
07:17 rook joined #openstack-rally
07:18 exploreshaifali boris-42: but once we discussed about the output, it should be a line graph
07:18 exploreshaifali boris-42: generated trends will be shown as line graphs or stackedArea graphs ?
07:19 nmagnezi joined #openstack-rally
07:19 boris-42 exploreshaifali: In case of atomic actions it will be stacked area I think
07:20 boris-42 exploreshaifali: in case of duration/idle_duration as well
07:20 boris-42 exploreshaifali: in case of errors it will be just line
07:20 exploreshaifali boris-42: stackedArea looks more good :P
07:21 exploreshaifali boris-42: so output will be like on y axis is our parameter let say error and on x axis tasks?
07:22 boris-42 exploreshaifali: exactly
07:23 exploreshaifali boris-42: and for each parameter we will create separate graph ?
07:23 nmagnezi joined #openstack-rally
07:23 exploreshaifali boris-42: like one for errors, one for atomic actions, one for dudration etc
07:24 exploreshaifali boris-42: or we will show all parametere based trends only in single graph
07:30 Miouge joined #openstack-rally
07:41 exploreshaifali boris-42: so output graph will look something like http://tinyurl.com/llns6jw
07:43 boris-42 exploreshaifali: nope there will be separated graphs for errors and for durations...
07:43 exploreshaifali boris-42: okay graphs will be separate for each parameter, rest all is fine in sample graph ?
07:44 exploreshaifali boris-42: apart from stagingarea graph and line graph issue :P
07:44 boris-42 exploreshaifali: sorry but I see only one graph
07:44 boris-42 exploreshaifali: in that file
07:44 exploreshaifali boris-42: yes, there is only 1 graph
07:44 boris-42 exploreshaifali: so yep that looks ok
07:45 exploreshaifali boris-42: should I careate sample graphs for all paramteres ?
07:46 openstackgerrit Merged stackforge/rally: Add base for Swift API Benchmarks - Patch-1  https://review.openstack.org/154225
07:47 boris-42 exploreshaifali: could you just use nvd3.js ?
07:47 boris-42 exploreshaifali: instead of using googlepaint?
07:47 exploreshaifali boris-42: yea sure :)
07:50 lordd joined #openstack-rally
07:51 amaretskiy joined #openstack-rally
07:57 e0ne joined #openstack-rally
07:59 lordd joined #openstack-rally
08:03 himangi joined #openstack-rally
08:22 yfried_ boris-42: good morning
08:22 yfried_ ran into a weird issue today
08:23 lordd joined #openstack-rally
08:24 yfried_ boris-42: http://paste.openstack.org/show/181718/
08:24 yfried_ boris-42: this happens when I'm trying to use rally normally
08:24 yfried_ boris-42: if I'm using root it works fine
08:43 himangi joined #openstack-rally
08:44 boris-42 yfried_: hm
08:44 boris-42 yfried_: so you need to change owner of db file
08:45 boris-42 yfried_: or re run rally-manage db recreate from user that you are using..
08:46 boris-42 yfried_: and we have as well https://bugs.launchpad.net/rally/+bug/1393887 bug related to this
08:46 boris-42 yfried_: I believe it should be fixed
08:49 openstackgerrit OpenStack Proposal Bot proposed stackforge/rally: Updated from global requirements  https://review.openstack.org/158764
08:52 e0ne joined #openstack-rally
08:53 yfried_ boris-42: what do you mean "[10:46] <boris-42> yfried_: I believe it should be fixed " ?
08:53 ekarlso joined #openstack-rally
08:53 yfried_ boris-42: was it fixed? I know LP has problems tracking rally
08:54 boris-42 yfried_: nope it was not fixed
08:54 boris-42 yfried_: I believe we should fix it to avoid such situations
08:54 yfried_ boris-42: also - this is new. yesterday everything worked fine
08:54 boris-42 yfried_: hehe
08:54 boris-42 yfried_: I know one patch
08:55 yfried_ boris-42: problem is, I deleted my VMs from yesterday so I can't prove it
08:55 yfried_ boris-42: did anything merge yesterday that handles installation/db/permissions?
08:57 boris-42 yfried_: there was one BIG BIG patch
08:57 boris-42 yfried_: https://github.com/stackforge/rally/commit/387dc356e3792f5cf24145bf381c0c30268a9e02
08:57 boris-42 yfried_: it removes olso incubator
08:58 boris-42 yfried_: from rally
08:58 fhubik joined #openstack-rally
08:59 yfried_ yeah - no idea how to search for this\
08:59 yfried_ boris-42: ^
09:00 boris-42 yfried_: maybe just revert this patch locally and try to reproduce on new vm?
09:01 kairat_kushaev joined #openstack-rally
09:03 yfried_ boris-42: I'll try. tnx
09:04 yfried_ boris-42: right now I don't have time/resource to triage this. I'll open a bug and add details if I can
09:05 openstackgerrit Kairat Kushaev proposed stackforge/rally: Add create-suspend-resume stack scenario  https://review.openstack.org/155346
09:06 rook joined #openstack-rally
09:08 yfried_ boris-42: https://bugs.launchpad.net/rally/+bug/1425439
09:09 yfried_ boris-42: I think this is HIGH prio but feel free to change that
09:09 yfried_ boris-42: are the ci jobs using root to execute rally?
09:17 andreykurilin_ joined #openstack-rally
09:23 tnurlygayanov joined #openstack-rally
09:23 psd joined #openstack-rally
09:35 e0ne joined #openstack-rally
09:42 Miouge joined #openstack-rally
09:43 himangi joined #openstack-rally
09:54 andreykurilin_ joined #openstack-rally
09:56 boris-42 yfried_: yep agree that it's high priority
09:56 boris-42 yfried_: I have some idea let me check
09:59 openstackgerrit Kairat Kushaev proposed stackforge/rally: Add create and check stack scenario  https://review.openstack.org/155685
09:59 andreykurilin_ joined #openstack-rally
10:05 andreykurilin_ joined #openstack-rally
10:06 anshul joined #openstack-rally
10:07 rook joined #openstack-rally
10:08 boris-42 yfried: so we had this https://github.com/stackforge/rally/commit/9b6acab919e8811dbcbd333bd647e2ce30177aed code
10:08 boris-42 yfried: maybe it was deleted/broken during the removing of oslco incubator
10:23 openstackgerrit Boris Pavlovic proposed stackforge/rally: Proposal for new task input file format  https://review.openstack.org/159065
10:24 boris-42 yfried|afk: ^
10:25 openstackgerrit Boris Pavlovic proposed stackforge/rally: Proposal for new task input file format  https://review.openstack.org/159065
10:27 Miouge joined #openstack-rally
10:34 yfried boris-42: how is this patch related to the bug?
10:36 yfried boris-42: oh. it's UNrelated. lol
10:43 openstackgerrit Roman Vasilets proposed stackforge/rally: Refactor ImageGenerator and validation  https://review.openstack.org/158797
10:45 yfried boris-42: do you have a fix for my bug?
10:45 boris-42 yfried: nope I am still not able to reporduce it
10:45 yfried really?
10:48 boris-42 yfried: yep
10:48 boris-42 yfried: could you provide information how did you install rally?
10:49 yfried boris-42: are you familiar with ansible?
10:51 boris-42 yfried:  not too much
10:51 boris-42 yfried: but maybe the issue is in your script?
10:51 yfried boris-42: I have an ansible playbook to recreate it. if you'd like to try it. if not, I'm doing it manually now
10:52 boris-42 yfried: let me try manually via install_rally.sh
10:53 boris-42 yfried: downloading centos
10:53 yfried boris-42: I'm working on fedora
10:54 boris-42 yfried: ah ok
11:02 openstackgerrit Merged stackforge/rally: Add warnings to not use Tempest benchmarks against production cloud  https://review.openstack.org/159017
11:02 aix joined #openstack-rally
11:05 yfried boris-42: do you have a f20 VM?
11:06 kairat_kushaev Hello Guys, do you have any idea why VMTasks.boot_runcommand_delete test can fail?
11:06 boris-42 kairat_kushaev: OpenStack
11:06 boris-42 yfried: nope but I can get it
11:06 boris-42 yfried: let me try ubuntu  first
11:07 yfried boris-42: using this image https://dl.fedoraproject.org/pub/fedora/linux/updates/20/Images/x86_64/Fedora-x86_64-20-20140407-sda.qcow2
11:09 yfried boris-42: this "script" http://paste.openstack.org/show/181797/
11:09 yfried I've just recreated the issue
11:14 boris-42 yfried: hm
11:14 boris-42 yfried: can you run one more command?
11:14 boris-42 yfried: rally-manage db recreate
11:15 yfried boris-42: same failure. want me to try sudo?
11:16 boris-42 yfried: nope without sudo
11:16 yfried can't be done :(
11:16 yfried boris-42:
11:17 boris-42 yfried: okay no worries we will find the issue
11:17 yfried boris-42: 2015-02-25 11:14:55.414 2667 CRITICAL rally [-] OperationalError: (OperationalError) attempt to write a readonly database u'\nDROP TABLE deployments' ()
11:18 boris-42 yfried: so easy fix is to change access to DB
11:18 boris-42 yfried: sudo chown ...
11:18 boris-42 yfried: but I am not sure why this happens at all
11:18 yfried boris-42: no idea how to change that, since I've no DB knowledge
11:19 yfried boris-42: could you please try to reproduce? I've given you image and script. let's make sure it's not just me
11:19 yfried boris-42: should take you 3min
11:27 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano benchmarks  https://review.openstack.org/137661
11:27 openstackgerrit Roman Vasilets proposed stackforge/rally: Add murano base for murano benchmarks  https://review.openstack.org/137650
11:28 Miouge joined #openstack-rally
11:32 boris-42 yfried: I will I will
11:32 boris-42 yfried: sorry can't work on your case 100% =)
11:33 yfried boris-42: ack
11:39 openstackgerrit Merged stackforge/rally: Optional arguments for boot_and_delete_server_with_secgroups  https://review.openstack.org/158798
11:39 openstackgerrit Merged stackforge/rally: Updated from global requirements  https://review.openstack.org/158764
11:43 oanufriev joined #openstack-rally
11:46 exploreshaifali joined #openstack-rally
11:55 rook_ joined #openstack-rally
12:04 openstackgerrit Roman Vasilets proposed stackforge/rally: Refactor ImageGenerator and validation  https://review.openstack.org/158797
12:04 cdent joined #openstack-rally
12:04 openstackgerrit Roman Vasilets proposed stackforge/rally: Refactor ImageGenerator and validation  https://review.openstack.org/158797
12:04 Miouge joined #openstack-rally
12:11 e0ne joined #openstack-rally
12:15 openstackgerrit Roman Vasilets proposed stackforge/rally: Refactor ImageGenerator and validation  https://review.openstack.org/158797
12:18 jaypipes joined #openstack-rally
12:31 msdubov_ joined #openstack-rally
12:33 openstackgerrit Merged stackforge/rally: Removed **kwargs for volume in create_and_attach_volume  https://review.openstack.org/158808
12:33 openstackgerrit Merged stackforge/rally: Don't launch stress cleanup in TempestContext  https://review.openstack.org/154408
12:37 wtakase joined #openstack-rally
12:39 pboros joined #openstack-rally
12:43 aix joined #openstack-rally
12:46 andreykurilin exploreshaifali: yes
12:47 andreykurilin exploreshaifali: hi! I'm here. can I help you?
12:47 exploreshaifali andreykurilin, hello, I got solution of my problem. Thanks :)
12:57 rook_ joined #openstack-rally
13:03 abhirc joined #openstack-rally
13:19 psd joined #openstack-rally
13:22 psd boris-42, Hey!
13:23 mwagner_lap joined #openstack-rally
13:32 openstackgerrit Yusuke Ide proposed stackforge/rally: Catch a connection refused exception not to show traces  https://review.openstack.org/157659
13:32 Miouge joined #openstack-rally
13:54 openstackgerrit Andrey Kurilin proposed stackforge/rally: WIP: try to launch rally in py34 env  https://review.openstack.org/158743
13:57 yfried_ joined #openstack-rally
14:04 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add Ceilometer scenario for nova notifications  https://review.openstack.org/154803
14:09 abhirc_ joined #openstack-rally
14:22 Miouge joined #openstack-rally
14:23 openstackgerrit Sergey Skripnick proposed stackforge/rally: Fix db configuration in rally_install.sh  https://review.openstack.org/159141
14:30 openstackgerrit nayna-patel proposed stackforge/rally: Adds Designate server-list scenario  https://review.openstack.org/157691
14:31 openstackgerrit Mikhail Dubov proposed stackforge/rally: Unify RPS and constant runners and tune constant runner  https://review.openstack.org/155225
14:33 nmagnezi amaretskiy, hi again
14:35 openstackgerrit Yusuke Ide proposed stackforge/rally: Catch a connection refused exception not to show traces  https://review.openstack.org/157659
14:38 nmagnezi amaretskiy, i re-wrote run_command (which is now actually, run_action)
14:39 nmagnezi amaretskiy, but i'm having hard time with the unittests (Mocks are giving me hard time)
14:39 nmagnezi amaretskiy, can you advise?
14:39 amaretskiy hi
14:39 amaretskiy yes
14:40 amaretskiy what is the question? (or, if required, submit a patch set so I can review/help with tests there)
14:41 Miouge joined #openstack-rally
14:47 rook_ joined #openstack-rally
14:47 nmagnezi amaretskiy, ok so I'll submit the patch in few minutes.. I expect unittests to fail but after we discuss them I'll fix it :-)
14:47 amaretskiy nmagnezi: okay
14:54 openstackgerrit Boris Pavlovic proposed stackforge/rally: Fix db configuration in rally_install.sh  https://review.openstack.org/159141
14:56 psd joined #openstack-rally
14:56 openstackgerrit Nir Magnezi proposed stackforge/rally: Modifies VMScenario._run_action function, to handle single bash cmd  https://review.openstack.org/154825
14:56 nmagnezi amaretskiy, ok please take a look
14:56 amaretskiy ok
14:56 nmagnezi amaretskiy, i'm new to Mock so bear with me :-)
14:57 amaretskiy okey
14:57 amaretskiy please wait a while, I'm finishing another review :)
14:57 nmagnezi amaretskiy, np at all
14:57 nmagnezi amaretskiy, when you check it, my questions relate to https://review.openstack.org/#/c/154825/14/tests/unit/benchmark/scenarios/vm/test_vmtasks.py
15:08 openstackgerrit joined #openstack-rally
15:10 yfried_ joined #openstack-rally
15:21 openstackgerrit Roman Vasilets proposed stackforge/rally: Add functional test for task samples  https://review.openstack.org/139220
15:21 openstackgerrit Nir Magnezi proposed stackforge/rally: Modifies VMScenario._run_action function, to handle single bash cmd  https://review.openstack.org/154825
15:23 nmagnezi amaretskiy, ok i will write my questions here so you can just address them when you are done with your curr review :)
15:23 nmagnezi amaretskiy, the code: https://review.openstack.org/#/c/154825/15/tests/unit/benchmark/scenarios/vm/test_utils.py
15:24 nmagnezi amaretskiy, so following to this patch changes (removed _run_command as you suggested and now using _run_action instead)
15:24 amaretskiy nmagnezi i think it is more convenient to add questions as comments to the patch
15:24 nmagnezi amaretskiy, ok will do so
15:27 nmagnezi amaretskiy, done :) https://review.openstack.org/#/c/154825/15/tests/unit/benchmark/scenarios/vm/test_utils.py
15:33 amaretskiy nmagnezi, ok
15:45 kairat_kushaev Guys, what is happening with Rally CI the whole day? I am getting a strange errors "./rally/cmd/cliutils.py:105:28: Q000 Remove Single quotes                          {'labels': field_labels, 'fields': fields})" but I could not break this with my patch.
15:46 kairat_kushaev Would you please help me with the problem?
15:47 rook_ joined #openstack-rally
15:47 wtakase joined #openstack-rally
15:48 abhirc joined #openstack-rally
15:48 openstackgerrit Vitaly Gusev proposed stackforge/rally: Add Ceilometer scenario for nova notifications  https://review.openstack.org/154803
15:48 wtakase boris-42: Hi.
15:59 neeti joined #openstack-rally
16:00 psd joined #openstack-rally
16:01 exploreshaifali joined #openstack-rally
16:05 andreykurilin kairat_kushaev: hi
16:05 kairat_kushaev andreykurilin: hi
16:06 andreykurilin kairat_kushaev: "Q000 Remove Single quotes" is not related to your patch and you can ignore it now.
16:06 kairat_kushaev andreykurilin: Thanks for response. Have any estimations when it is going to be fixed?
16:07 andreykurilin kairat_kushaev: or you can send separate patch with fix of this issue:)
16:08 andreykurilin kairat_kushaev: If you send a patch with fix, I'll try to find another core to merge it as soon as possible:)
16:28 amaretskiy nmagnezi take a look at my comments
16:30 amaretskiy nmagnezi: I think that this new feature requires a json/yaml example, please add a customized boot-runcommand-delete.(json|yaml) with simple inline script example
16:33 exploreshaifali joined #openstack-rally
16:44 nmagnezi amaretskiy, I've wrote a scenario that uses that capability, yet it involves additional work that i've submitted in other patches
16:44 nmagnezi amaretskiy, i will commit this as WIP patch so you can have a look at it
16:44 amaretskiy ok
16:44 nmagnezi :)
16:45 nmagnezi amaretskiy, thanks for you help. i'll get to your comments first thing tomorrow
16:45 nmagnezi amaretskiy, bye for now :)
16:46 amaretskiy but it seems that new feature does not require extra code to have a working example - just add something like this: boot-runcommand-delete-inline-script.json, with script="df -h" and script_is_inline=True
16:46 amaretskiy ok :)
16:48 nmagnezi amaretskiy, yes but boot_runcommands)_delete expects a json output.. which "df -h" will not produce
16:48 amaretskiy yes, sure
16:48 amaretskiy another inline script is required :)
16:50 nmagnezi amaretskiy, sort of.. when inline_script=False it's just a bash cmd :D
16:50 amaretskiy take a look t my comment
16:50 nmagnezi ack
16:50 amaretskiy generally `inline' means the opposite
16:50 psd joined #openstack-rally
16:51 amaretskiy ok, will discuss tomorrow :)
16:56 rook_ joined #openstack-rally
16:58 openstackgerrit Andrey Kurilin proposed stackforge/rally: try to found py27 interpreter for verification  https://review.openstack.org/158490
17:23 openstackgerrit Wataru Takase proposed stackforge/rally: Add name pattern filter for resource cleanup  https://review.openstack.org/139643
17:25 openstackgerrit Olga Kopylova proposed stackforge/rally: Mark 'rally use {deployment|task|verification}' deprecated  https://review.openstack.org/159192
17:31 meteorfox boris-42: hey, saw your message from yesterday. How do you plane to solve the common context problem for multiple scenarios?
17:41 meteorfox boris-42: also, I opened the bug for the cached token issue from yesterday. https://bugs.launchpad.net/rally/+bug/1425615
17:52 boris-42 meteorfox: what I should solve there?)
17:52 boris-42 meteorfox: I will just pass 1 instance of context to all runners that we have
17:53 meteorfox boris-42: I don't know. Yesterday, you told me that people might complain if you combine the contexts, but then you said you found a way to do it
17:54 boris-42 meteorfox: I was talking about "input file format"
17:54 boris-42 meteorfox: btw I am going to simplify it a bit
17:55 boris-42 meteorfox: did you see that patch ?
17:55 meteorfox boris-42: do you have a prototype, or sample file, to see how it looks?
17:55 meteorfox no
17:56 boris-42 meteorfox: https://review.openstack.org/#/c/159065/ so here is proposal for new format
17:56 boris-42 jlk: btw are you around?
17:59 meteorfox boris-42: looks good, so if no context is specified for a scenario, then it takes the global context?
17:59 boris-42 meteorfox: so I am going to remove "scenario" and keep only "scenarios"
17:59 psd joined #openstack-rally
17:59 meteorfox boris-42: hm.. i see
18:00 boris-42 meteorfox: so on top level we have "scenarios", "context", "sla"
18:00 boris-42 meteorfox: inside scenarios we can specify 1-N scenarios with their args and runner
18:00 boris-42 meteorfox: and even context and sla
18:00 meteorfox boris-42: cool
18:00 boris-42 meteorfox: if context and sla are not specified for scenario it will take global one
18:01 boris-42 meteorfox: so if you have 5 scenarios without context they will work all together in one context
18:01 meteorfox boris-42: well I understand that they can inherit the context, but isn't the SLA specific to a scenario?
18:02 boris-42 meteorfox: so actually yes
18:03 boris-42 meteorfox: maybe it makes sense to allow to specify on top level only "scenarios" and "context"
18:03 boris-42 meteorfox: we should make as simple as possible config
18:03 meteorfox boris-42: right
18:03 boris-42 meteorfox: okay let me refactor it a bit
18:05 meteorfox boris-42: so, tasks is a list, and each task can contain multiple scenarios, tasks are executed sequentially, but scenarios are run concurrently. Is that correct?
18:06 boris-42 meteorfox yep
18:06 boris-42 meteorfox: so HA testing is running 1 task that turn-on turn-off nodes, another one does load on API
18:07 boris-42 meteorfox: so you can see how HA works under load
18:07 meteorfox boris-42: cool cool
18:07 boris-42 meteorfox: yep I think it's simpler to understand even then current task format=)
18:08 meteorfox boris-42: once we have persistent-context, how will this file be modified? another key to a DICT to specify the context?
18:09 boris-42 meteorfox: so you will have special context
18:09 meteorfox boris-42: or do we want to have a persisted context across tasks instead of scenarios
18:09 meteorfox ?
18:09 boris-42 meteorfox: called persistance context
18:10 boris-42 meteorfox: so in section context you will add something like
18:10 boris-42 persistance_context: UUID
18:10 boris-42 persistance_context: UUID|NAME
18:10 meteorfox boris-42: I see
18:11 meteorfox boris-42: what about quotas? and rate-limits?
18:15 himangi joined #openstack-rally
18:16 boris-42 meteorfox: hm?
18:16 himangi joined #openstack-rally
18:16 himangi_ joined #openstack-rally
18:17 meteorfox boris-42: well, I assume they can be specified at the global context
18:17 boris-42 meteorfox: yep
18:17 boris-42 meteorfox: as usually just use "quotas"
18:17 boris-42 meteorfox: context to do that
18:18 meteorfox boris-42: cool, yeah, I was thinking about someone that might have scenarios that combined can exhaust the quota
18:19 boris-42 meteorfox: hm but you can set any
18:19 meteorfox boris-42: yeah.
18:19 boris-42 meteorfox: btw what do you think about improving this directory
18:19 boris-42 meteorfox: I would prefer to split to 2 directories at least
18:20 boris-42 meteorfox: using_existing_clouds and deploy_new_clouds
18:20 boris-42 meteorfox: or something like that
18:21 meteorfox boris-42: which directory?
18:23 meteorfox boris-42: wouldn't that cause a lot of duplication? I feel we can come up with a way to re-use the tasks for both cases
18:24 boris-42 meteorfox: https://github.com/stackforge/rally/tree/master/samples/deployments
18:25 boris-42 meteorfox: it's about deployment not tasks
18:25 meteorfox boris-42: oh ok
18:30 meteorfox boris-42: well, maybe they could be name, "provisioned" and "static"
18:31 boris-42 meteorfox: maybe longer name?)
18:32 meteorfox boris-42: i'm a fan of short names. I think it's implied that they will be a cloud deployments
18:33 meteorfox boris-42: anyway, if you think the number of files within samples/deployments will continue to increase, I think you can split into 2, but if it's going to remain how it is currently, then I wouldn't split it
18:34 meteorfox there aren't that many
18:34 meteorfox boris-42: perhaps a convention on the name, might help
18:39 openstackgerrit Boris Pavlovic proposed stackforge/rally: Proposal for new task input file format  https://review.openstack.org/159065
18:40 boris-42 meteorfox: ^
18:41 * meteorfox looking
18:47 meteorfox boris-42: done. I reviewed your patch and left some comments
18:48 boris-42 meteorfox: nice
18:48 boris-42 meteorfox: please assign this bug https://bugs.launchpad.net/rally/+bug/1425615
18:48 meteorfox boris-42: oh right
18:49 meteorfox :)
19:10 garyk joined #openstack-rally
19:15 andreykurilin_ joined #openstack-rally
19:16 psd joined #openstack-rally
19:17 openstackgerrit Andrey Kurilin proposed stackforge/rally: try to found py27 interpreter for verification  https://review.openstack.org/158490
19:23 psd boris-42, Hey! did you find time to sub divide the cloud info task?
19:27 exploreshaifali joined #openstack-rally
19:29 openstackgerrit Jan Klare proposed stackforge/rally: **kwargs in create_image_and_boot_instances only for server  https://review.openstack.org/159233
19:48 psd_ joined #openstack-rally
20:11 mwagner_lap joined #openstack-rally
20:28 openstackgerrit Merged stackforge/rally: Fix db configuration in rally_install.sh  https://review.openstack.org/159141
20:58 jlk boris-42: I am, but really swamped today
21:38 openstackgerrit John Wu proposed stackforge/rally: Add base for Swift API Benchmarks - Patch-2  https://review.openstack.org/159258
22:14 nmagnezi joined #openstack-rally
23:17 openstackgerrit Merged stackforge/rally: **kwargs in create_image_and_boot_instances only for server  https://review.openstack.org/159233

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