| *** logan2 has quit IRC | 00:01 | |
| *** logan2 has joined #openstack-sahara | 00:08 | |
| *** raildo_ has joined #openstack-sahara | 00:44 | |
| *** openstackgerrit has quit IRC | 00:51 | |
| *** openstackgerrit has joined #openstack-sahara | 00:51 | |
| *** raildo_ has quit IRC | 01:08 | |
| *** raildo has quit IRC | 01:10 | |
| *** chen12 has quit IRC | 01:25 | |
| *** witlessb has joined #openstack-sahara | 01:56 | |
| *** witlessb has quit IRC | 02:01 | |
| *** hdd has joined #openstack-sahara | 02:03 | |
| *** hdd has quit IRC | 03:15 | |
| *** coolsvap|away is now known as coolsvap | 03:50 | |
| *** Nikolay_St has quit IRC | 03:56 | |
| *** sgotliv has joined #openstack-sahara | 04:06 | |
| openstackgerrit | lu huichun proposed openstack/sahara-specs: [EDP]upgrade oozie web service api version in sahara oozie engine https://review.openstack.org/200993 | 04:52 |
|---|---|---|
| openstackgerrit | lu huichun proposed openstack/sahara-specs: [EDP]upgrade oozie web service api version in sahara oozie engine https://review.openstack.org/200993 | 04:53 |
| *** pcaruana has quit IRC | 05:08 | |
| *** Longgeek has quit IRC | 05:18 | |
| *** Longgeek has joined #openstack-sahara | 05:20 | |
| *** Nikolay_St has joined #openstack-sahara | 05:46 | |
| *** Nikolay_St has quit IRC | 05:59 | |
| *** Nikolay_St has joined #openstack-sahara | 06:00 | |
| *** nkrinner has joined #openstack-sahara | 06:04 | |
| *** hogepodge has quit IRC | 06:19 | |
| *** hogepodge has joined #openstack-sahara | 06:26 | |
| *** sgotliv has quit IRC | 06:39 | |
| *** pcaruana has joined #openstack-sahara | 06:42 | |
| *** pino|work has joined #openstack-sahara | 07:05 | |
| *** witlessb has joined #openstack-sahara | 07:19 | |
| *** coolsvap is now known as coolsvap|afk | 07:32 | |
| *** chlong-weekend is now known as chlong-afk | 07:33 | |
| *** coolsvap|afk is now known as coolsvap | 07:56 | |
| *** sgotliv has joined #openstack-sahara | 08:08 | |
| openstackgerrit | Sergey Gotliv proposed openstack/sahara: Modify launch_command to support global variables https://review.openstack.org/197315 | 08:10 |
| openstackgerrit | Sergey Gotliv proposed openstack/sahara: Adding support for the Spark Shell job https://review.openstack.org/197333 | 08:10 |
| *** chen12 has joined #openstack-sahara | 08:32 | |
| *** chen12 has quit IRC | 08:43 | |
| *** chen12 has joined #openstack-sahara | 08:44 | |
| *** chen12 has quit IRC | 08:44 | |
| *** Poornima has joined #openstack-sahara | 08:46 | |
| *** jamielennox has quit IRC | 08:46 | |
| *** tosky has joined #openstack-sahara | 09:14 | |
| tosky | degorenko: ping https://review.openstack.org/#/c/200949/ (and https://review.openstack.org/#/c/200948/) - do I have an easy way to test the sahara-ci-config in production, apart from rebuilding the CI? | 09:26 |
| tosky | If the patches breaks by accident the gate, it will be a bit complicated | 09:26 |
| *** Poornima has quit IRC | 09:39 | |
| *** jamielennox has joined #openstack-sahara | 09:46 | |
| *** esikachev has joined #openstack-sahara | 09:50 | |
| *** coolsvap is now known as coolsvap|away | 10:25 | |
| *** coolsvap|away is now known as coolsvap | 10:28 | |
| openstackgerrit | Merged openstack/sahara: Modify launch_command to support global variables https://review.openstack.org/197315 | 10:36 |
| *** coolsvap is now known as coolsvap|away | 10:43 | |
| openstackgerrit | Merged openstack/sahara-image-elements: Rewrite/improve start-mysql.sh https://review.openstack.org/199062 | 10:43 |
| openstackgerrit | Merged openstack/sahara-image-elements: hadoop: consolidate /run creation https://review.openstack.org/199125 | 10:54 |
| openstackgerrit | Sergey Gotliv proposed openstack/sahara: Adding support for the Spark Shell job https://review.openstack.org/197333 | 10:59 |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara-image-elements: Add elements for sync time on VM https://review.openstack.org/129611 | 11:10 |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Add cluster verification https://review.openstack.org/196713 | 11:15 |
| *** _crobertsrh is now known as crobertsrh | 11:18 | |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara: Configure NTP service on cluster instances https://review.openstack.org/200136 | 11:21 |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara: Configure NTP service on cluster instances https://review.openstack.org/200136 | 11:22 |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Improvement check scale in scenario tests https://review.openstack.org/195513 | 11:26 |
| openstackgerrit | Merged openstack/sahara: add unit test covering cancel_job in job_manager https://review.openstack.org/198739 | 11:27 |
| openstackgerrit | Merged openstack/sahara: Updated from global requirements https://review.openstack.org/200913 | 11:27 |
| *** sgotliv has quit IRC | 11:28 | |
| *** coolsvap|away is now known as coolsvap | 11:44 | |
| *** witlessb has quit IRC | 12:05 | |
| *** witlessb has joined #openstack-sahara | 12:05 | |
| *** sgotliv has joined #openstack-sahara | 12:05 | |
| openstackgerrit | Sergey Reshetnyak proposed openstack/sahara: [WIP] Add put data in HDFS for EDP testcase https://review.openstack.org/201102 | 12:05 |
| *** sgotliv has quit IRC | 12:07 | |
| *** sgotliv has joined #openstack-sahara | 12:07 | |
| *** hdd has joined #openstack-sahara | 12:11 | |
| openstackgerrit | Sergey Reshetnyak proposed openstack/sahara: [WIP] Add put data in HDFS for EDP testcase https://review.openstack.org/201102 | 12:11 |
| *** esikachev has quit IRC | 12:29 | |
| *** esikachev has joined #openstack-sahara | 12:33 | |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Add cluster verification https://review.openstack.org/196713 | 12:39 |
| *** sgotliv has quit IRC | 12:47 | |
| *** elmiko has joined #openstack-sahara | 12:50 | |
| *** sgotliv has joined #openstack-sahara | 12:53 | |
| openstackgerrit | Nikita Konovalov proposed openstack/sahara: EDP engine Swift fix for HDP 2.2 https://review.openstack.org/200504 | 13:14 |
| *** sgotliv has quit IRC | 13:28 | |
| *** raildo has joined #openstack-sahara | 13:33 | |
| *** raildo has quit IRC | 13:40 | |
| *** raildo has joined #openstack-sahara | 13:44 | |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Improvement check scale in scenario tests https://review.openstack.org/195513 | 13:44 |
| *** vgridnev has joined #openstack-sahara | 13:49 | |
| *** shakamunyi has joined #openstack-sahara | 13:59 | |
| *** barra204 has joined #openstack-sahara | 13:59 | |
| *** egafford has joined #openstack-sahara | 14:00 | |
| *** tellesnobrega has joined #openstack-sahara | 14:00 | |
| tellesnobrega | alazarev, ping | 14:02 |
| *** tmckay has joined #openstack-sahara | 14:21 | |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Add cluster verification https://review.openstack.org/196713 | 14:22 |
| *** barra204 has quit IRC | 14:25 | |
| *** whydidyoustealmy has joined #openstack-sahara | 14:25 | |
| *** shakamunyi has quit IRC | 14:25 | |
| *** shakamunyi has joined #openstack-sahara | 14:26 | |
| vgridnev | tmckay, are you around? | 14:30 |
| tmckay | hi | 14:31 |
| tmckay | vgridnev ^^ | 14:31 |
| vgridnev | I have question about launching spark jobs using swift | 14:31 |
| vgridnev | How it can be done? | 14:31 |
| vgridnev | Do we have some sample job with some instructions? | 14:32 |
| openstackgerrit | Merged openstack/sahara: Resolve 500 error during simultaneously deletion https://review.openstack.org/198304 | 14:32 |
| *** whydidyoustealmy has quit IRC | 14:34 | |
| *** shakamunyi has quit IRC | 14:34 | |
| tmckay | vgridnev, let me double-check, I haven't run that in a while. There is a wrapper jar that should be bundled with Sahara to make this work. And you need to set a configuration flag (I think it is edp.spark.adapt_for_swift=True) | 14:34 |
| openstackgerrit | Sergey Reshetnyak proposed openstack/sahara: [WIP] Add put data in HDFS for EDP testcase https://review.openstack.org/201102 | 14:34 |
| tmckay | vgridnev, the normal hadoop jar should allow the swift:// paths to be dereferenced (just like Hadoop j obs). The wrapper puts the credentials for swift access into the Hadoop context. | 14:35 |
| vgridnev | tmckay, not only eep.spark.adapt_for_swift should be provided, password and login for swift should be added too, I think | 14:37 |
| openstackgerrit | Merged openstack/sahara: Spark job for Cloudera 5.3.0 and 5.4.0 added https://review.openstack.org/190128 | 14:37 |
| tmckay | vgridnev, yes | 14:39 |
| tmckay | vgridnev, I'll check and see if we have anything about this in edp examples. I know for Liberty crobertsrh is adding check boxes for a few booleans like edp.spark.adapt_for_swift | 14:40 |
| tmckay | vgridnev, and of course the proxy user stuff should work for spark as well | 14:40 |
| crobertsrh | Yes, those patches are up...since late May or early June | 14:41 |
| openstackgerrit | Ethan Gafford proposed openstack/sahara-specs: API to mount manila shares to clusters https://review.openstack.org/200633 | 14:46 |
| *** sgotliv has joined #openstack-sahara | 14:47 | |
| *** nkrinner is now known as nkrinner_afk | 14:48 | |
| *** htruta has joined #openstack-sahara | 14:49 | |
| *** Nikolay_St has quit IRC | 14:50 | |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara: Resolve 500 error during simultaneously deletion https://review.openstack.org/201156 | 14:50 |
| vgridnev | crobertsrh, could you please add your changes for that to etherpad / sahara-reviews-in-horizon | 14:55 |
| vgridnev | ? | 14:55 |
| crobertsrh | will do | 14:55 |
| alazarev | tellesnobrega, pong | 14:58 |
| openstackgerrit | Telles Mota Vidal Nóbrega proposed openstack/sahara: Allow multiple clusters creation https://review.openstack.org/193586 | 15:01 |
| tellesnobrega | ^ | 15:01 |
| tellesnobrega | did a little refactoring on the code above | 15:02 |
| tellesnobrega | can you take a look? | 15:02 |
| tellesnobrega | elmiko, SergeyLukjanov can you guys take a look as well | 15:02 |
| *** sgotliv has quit IRC | 15:02 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-saharaclient: Updated from global requirements https://review.openstack.org/201208 | 15:03 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/sahara: Updated from global requirements https://review.openstack.org/201209 | 15:03 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/sahara-extra: Updated from global requirements https://review.openstack.org/201210 | 15:03 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/sahara-image-elements: Updated from global requirements https://review.openstack.org/201211 | 15:03 |
| elmiko | tellesnobrega: ack | 15:03 |
| *** sgotliv has joined #openstack-sahara | 15:09 | |
| openstackgerrit | Sergey Reshetnyak proposed openstack/sahara: [WIP] Print error stacktrace from Ambari errors https://review.openstack.org/199615 | 15:15 |
| tosky | oh, right, stacktraces for ambari would help | 15:18 |
| tosky | egafford: would they have helped with that HA issue, I guess ^^ | 15:18 |
| tosky | ? | 15:18 |
| egafford | tosky: In that specific case, not really; the error there only manifested in Oozie. | 15:20 |
| tosky | oh, right | 15:20 |
| egafford | tosky: There are definitely other issues we've run into that this would've helped, though. | 15:20 |
| egafford | (Would've taken us one layer deeper into that DNS issue, for instance.) Trouble with the ambari server logs is that they usually just point you to the next layer of the rabbit hole. Still, this is what can easily/reasonably be done, so it makes sense. | 15:22 |
| vgridnev | It's only for new HDP 2.2 plugin, actually | 15:22 |
| *** sgotliv has quit IRC | 15:30 | |
| *** dmitryme has quit IRC | 15:36 | |
| *** crobertsrh has quit IRC | 15:36 | |
| *** DuncanT has quit IRC | 15:36 | |
| *** dmitryme has joined #openstack-sahara | 15:37 | |
| vgridnev | tmckay, could you please send me an email about double-checking of spark with swift? | 15:37 |
| *** crobertsrh has joined #openstack-sahara | 15:38 | |
| *** pcaruana has quit IRC | 15:38 | |
| tmckay | vgridnev, will do | 15:38 |
| vgridnev | vgridnev at mirantis dot com for reference | 15:38 |
| tmckay | ok. otp now but I'll check | 15:38 |
| openstackgerrit | Evgeny Sikachev proposed openstack/sahara: Improvement check scale in scenario tests https://review.openstack.org/195513 | 15:41 |
| *** esikachev has quit IRC | 15:46 | |
| *** esikachev has joined #openstack-sahara | 15:48 | |
| openstackgerrit | Pino Toscano proposed openstack/sahara-image-elements: hadoop: mark our rc.local replacement as executable https://review.openstack.org/201232 | 15:51 |
| *** esikachev has quit IRC | 15:54 | |
| *** DuncanT has joined #openstack-sahara | 15:57 | |
| tellesnobrega | elmiko, thanks | 16:14 |
| elmiko | np | 16:15 |
| *** esikachev has joined #openstack-sahara | 16:22 | |
| *** shakamunyi has joined #openstack-sahara | 16:25 | |
| *** whydidyoustealmy has joined #openstack-sahara | 16:25 | |
| *** tmckay1 has joined #openstack-sahara | 16:25 | |
| *** tmckay has quit IRC | 16:27 | |
| *** esikachev has quit IRC | 16:29 | |
| tellesnobrega | elmiko, i forgot sergeys comment, i'm working on it and uploading in a bit | 16:29 |
| elmiko | tellesnobrega: no worries, just seemed like a nice suggestion | 16:29 |
| tellesnobrega | yeah, i liked it too | 16:30 |
| *** whydidyoustealmy has quit IRC | 16:37 | |
| *** shakamunyi has quit IRC | 16:37 | |
| *** raildo has quit IRC | 16:38 | |
| openstackgerrit | Telles Mota Vidal Nóbrega proposed openstack/sahara: Allow multiple clusters creation https://review.openstack.org/193586 | 16:47 |
| tellesnobrega | elmiko, alazarev, egafford ^ | 16:47 |
| tellesnobrega | sreshetnyak, ^ | 16:47 |
| *** shakamunyi has joined #openstack-sahara | 16:51 | |
| *** whydidyoustealmy has joined #openstack-sahara | 16:51 | |
| *** raildo has joined #openstack-sahara | 16:52 | |
| openstackgerrit | Luigi Toscano proposed stackforge/sahara-ci-config: Directly use the scenario configuration templates https://review.openstack.org/200948 | 16:53 |
| tosky | degorenko: fixed patch ^^, sorry for the mess | 16:54 |
| tosky | I forgot a "--" | 16:54 |
| tmckay1 | vgridnev, are you still there? | 17:00 |
| vgridnev | yes | 17:01 |
| *** tmckay1 is now known as tmckay | 17:01 | |
| tmckay | vgridnev, yay! :) | 17:01 |
| tmckay | vgridnev, okay, I looked at the code in service/edp/spark/engine.py | 17:01 |
| tmckay | I remember now :) If edp.spark.adapt_for_swift is True, the spark EDP engine will upload an extra wrapper jar and a spark.xml file that contains credentials for hadoop. It will generate a different command. | 17:03 |
| tmckay | vgridnev, it handles proxy credentials too | 17:03 |
| tmckay | So, the wrapper reads the spark.xml, updates the hadoop config for the job, and launches the job | 17:03 |
| tmckay | vgridnev, so, all you should need is to set the boolean, and provide the credentials. The rest is transparent. | 17:04 |
| tmckay | vgridenv, except for one small thing -- Spark jobs are like java jobs, so you need to include the swift:// path by hand as an argument (or use the data source substitution stuff) | 17:04 |
| tmckay | vgridnev, make sense? Do you have other questions? | 17:05 |
| tmckay | (in other words, spark jobs reference swift paths the same way java jobs do) | 17:06 |
| vgridnev | tmckay, I will retry it, and will said what exact problem I got | 17:06 |
| tmckay | vgridnev, okay, sounds good. Let me know | 17:06 |
| *** shakamunyi has quit IRC | 17:09 | |
| *** whydidyoustealmy has quit IRC | 17:09 | |
| *** sgotliv has joined #openstack-sahara | 17:09 | |
| *** shakamunyi has joined #openstack-sahara | 17:09 | |
| *** stanchan has joined #openstack-sahara | 17:11 | |
| *** melaks has joined #openstack-sahara | 17:12 | |
| *** hdd has quit IRC | 17:13 | |
| tellesnobrega | have you guys experienced problems when executing a MapReduce job, the jobs ends fine, but the output is not copied to swift so the job is marked as failed? | 17:21 |
| elmiko | tellesnobrega: haven't seen that, but i also haven't tested an MR job recently | 17:22 |
| tellesnobrega | elmiko, i'm running an experiment, and weird things happens... i'm killing one of the nodes during the job processing, but hadoop is supposed to recover from that | 17:23 |
| elmiko | hmm | 17:23 |
| tellesnobrega | sometimes i get the output, sometimes it doesn't work | 17:23 |
| elmiko | weird, i wonder if its a race condition or something? | 17:23 |
| tellesnobrega | even though the jobs is marked as successful in hadoop | 17:23 |
| elmiko | yea, seems odd. | 17:25 |
| *** Longgeek has quit IRC | 17:29 | |
| tellesnobrega | the worse (for me in this case) the important variable i'm measuring is the job completion time, when if fails to copy to swift, sahara gives me a really high completion time | 17:34 |
| *** esikachev has joined #openstack-sahara | 17:34 | |
| elmiko | =( | 17:36 |
| pino|work | "big data, big times" | 17:37 |
| * pino|work runs | 17:37 | |
| *** chlong-afk has quit IRC | 17:38 | |
| elmiko | lol | 17:39 |
| *** melaks has quit IRC | 17:40 | |
| tellesnobrega | lol | 17:43 |
| tellesnobrega | pino|work, i dont mind big times, if they are real | 17:47 |
| *** melaks has joined #openstack-sahara | 17:49 | |
| *** witlessb has quit IRC | 17:52 | |
| *** witlessb has joined #openstack-sahara | 17:52 | |
| *** tosky has quit IRC | 17:56 | |
| *** pino|work has quit IRC | 17:56 | |
| openstackgerrit | lu huichun proposed openstack/sahara-specs: [EDP]upgrade oozie web service api version in sahara oozie engine https://review.openstack.org/200993 | 17:56 |
| *** vgridnev_ has joined #openstack-sahara | 17:57 | |
| vgridnev_ | tmckay, do you have sample spark job? | 17:58 |
| vgridnev_ | that operating with swift | 17:58 |
| *** vgridnev has quit IRC | 17:59 | |
| tmckay | vgridnev_, otp again, I think I have something simple I played with when I was using the wrapper. I'll take a look. | 17:59 |
| tmckay | there is also a spark wordcount I believe in the spark examples (bundled with the spark install) | 17:59 |
| *** esikachev has quit IRC | 18:01 | |
| *** hdd has joined #openstack-sahara | 18:05 | |
| *** tellesmvn has joined #openstack-sahara | 18:19 | |
| *** tellesnobrega has quit IRC | 18:20 | |
| *** tellesmvn is now known as tellesnobrega | 18:20 | |
| vgridnev_ | tmckay, hadoop-swift.jar will be preinstalled on image, right? | 18:25 |
| tmckay | yes | 18:25 |
| *** shakamunyi has quit IRC | 18:36 | |
| *** Nikolay_St has joined #openstack-sahara | 18:41 | |
| *** tmckay has quit IRC | 18:45 | |
| *** shakamunyi has joined #openstack-sahara | 18:47 | |
| *** esikachev has joined #openstack-sahara | 18:50 | |
| *** ekarlso has quit IRC | 18:56 | |
| *** ekarlso has joined #openstack-sahara | 18:58 | |
| *** tmckay has joined #openstack-sahara | 18:59 | |
| *** Nikolay_St has quit IRC | 19:14 | |
| vgridnev_ | egafford, could you please merge this change in stable/kilo branch ? https://review.openstack.org/#/c/200472/ | 19:31 |
| vgridnev_ | this change unlock our gate | 19:31 |
| egafford | vgridnev_: Looking now. | 19:31 |
| egafford | Ah, yup; this. | 19:31 |
| egafford | vgridnev_: Done; thanks. | 19:33 |
| *** melaks has quit IRC | 19:58 | |
| tmckay | vgridnev_, okay, just sent you a simple Scala app for spark. It should take 1 argument (a file path) | 20:01 |
| tmckay | output is just to standard out | 20:01 |
| tmckay | but it's good enough to read from Swift, to prove it works | 20:01 |
| tmckay | vgridnev_, blah, mirantis mail doesn't like the jar file I think. | 20:03 |
| tmckay | I'll send just the source | 20:04 |
| tmckay | well, or put the jar in the tgz | 20:04 |
| vgridnev_ | hm, intresting | 20:08 |
| tmckay | bah, again. just a tgz this time | 20:08 |
| tmckay | ah, maybe it's an archive in an archive | 20:09 |
| vgridnev_ | Seems that all errors was caused by old spark image on lab | 20:09 |
| tmckay | vgridnev_, oh, you made it work? | 20:09 |
| tmckay | Yay :) | 20:09 |
| tmckay | then I don't have to send. Actually, I could send through a draft CR in gerrit ;-) | 20:10 |
| vgridnev_ | tmckay, yes, that works | 20:16 |
| vgridnev_ | thanks for your help | 20:17 |
| tmckay | excellent, great to hear | 20:18 |
| openstackgerrit | Ethan Gafford proposed openstack/sahara: Cluster creation with trust https://review.openstack.org/200719 | 20:22 |
| *** melaks has joined #openstack-sahara | 20:23 | |
| openstackgerrit | Merged openstack/sahara: Fix failed unit tests https://review.openstack.org/200472 | 20:31 |
| *** hdd has quit IRC | 20:31 | |
| *** hdd has joined #openstack-sahara | 20:34 | |
| *** esikachev has quit IRC | 20:34 | |
| *** esikachev has joined #openstack-sahara | 20:42 | |
| *** esikachev has quit IRC | 20:47 | |
| *** tellesnobrega has quit IRC | 20:50 | |
| *** tellesmvn has joined #openstack-sahara | 20:50 | |
| *** tellesmvn is now known as tellesnobrega | 20:52 | |
| *** tellesnobrega has left #openstack-sahara | 20:56 | |
| *** tellesnobrega has joined #openstack-sahara | 20:56 | |
| *** Nikolay_St has joined #openstack-sahara | 21:00 | |
| *** tmckay has quit IRC | 21:01 | |
| *** tellesnobrega has quit IRC | 21:04 | |
| *** tellesnobrega has joined #openstack-sahara | 21:05 | |
| *** crobertsrh is now known as _crobertsrh | 21:07 | |
| *** raildo has quit IRC | 21:08 | |
| egafford | elmiko: Looking at your keystone sessions spec. I haven't studied how sessions interact with trusts (which we're moving to use for all clusters in order to perform long operations and allow automated cluster cleanup.) Have you already poked that bear? | 21:09 |
| *** hdd has quit IRC | 21:11 | |
| elmiko | egafford: hmm, i'll look specifically at the trusts code. it might require some minor adjustments in the client instantiation code, but i think it would be limited to those areas. | 21:13 |
| elmiko | egafford: we would just need to be aware of the trust, and use it during the session scoping. | 21:13 |
| elmiko | egafford: we will still be using whatever credentials are supplied, so in the case of a trust we would just need to ensure that the user with the trust will be generating the session object | 21:14 |
| egafford | Yeah; it looks as though your SessionCache object can allow this to happen. The key structure of the cache is (maybe intentionally) vague atm, but it doesn't look like a conflict. | 21:14 |
| elmiko | i don't think so, the cache should be blind to whatever is requested. the real details will come from the specific client you are requesting and the auth plugin. | 21:15 |
| elmiko | but, if you pass a trust id to the client, we can ensure that the session is created with a trust | 21:15 |
| egafford | Sure. Are the sessions created and cached per-user, though? Our trusts are currently scoped per cluster. | 21:16 |
| elmiko | they will be, so we might need a mechanism to observe a user who wants 2 different sessions for a single service (ie one with a trust and one without) | 21:17 |
| elmiko | but, these sessions will not persist between requests. so it might be a non-issue. we'll see, i suppose lol | 21:17 |
| egafford | I see; if the sessions are non-persistent between requests, I believe it is in face a non-issue. I think I got caught up in the SessionCache-ness of it all and imagined persistent sessions across requests. | 21:18 |
| egafford | s/face/fact/ | 21:19 |
| elmiko | yea, i don't think it will be used that way. this is mainly to avoid us having to call, for example, a nova client create several times within the same request | 21:19 |
| elmiko | we can reuse the client | 21:19 |
| egafford | Sure; that should be fine and good. Although, Telles' multi-cluster change. Hmm... | 21:20 |
| elmiko | it should really be useful there, as the same heat/neutron/nova clients will get reused instead of recreated each time | 21:20 |
| elmiko | and the session object will automagically handle renewing tokens and whatnot | 21:20 |
| egafford | Right, but if we're using trusts per cluster, in theory we'll have a logical breakdown (clusters will use one anothers' trusts.) In practice, I don't think we care even a little; there's nothing about the trusts that's actually scoped to a cluster beyond the fact that we store the ids on them because in the general case that works for us. | 21:22 |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara: Resolve 500 error during simultaneously deletion https://review.openstack.org/201156 | 21:27 |
| elmiko | egafford: i'll look more closely at the trust mechanics for the periodic stuff. my gut feeling is that you are correct, it will be a wash in the end, but just to make sure. | 21:28 |
| *** coolsvap has quit IRC | 21:28 | |
| elmiko | also, lol at whoever put username:sam password:gamgee into our tests ;) | 21:29 |
| *** coolsvap|away has joined #openstack-sahara | 21:33 | |
| *** esikachev has joined #openstack-sahara | 21:38 | |
| elmiko | egafford: yes, i can see how we will adjust the context for the trust work. it's doable | 21:38 |
| elmiko | egafford: its a little easy because the trusts all seem to use the admin context for their operations | 21:40 |
| egafford | elmiko: Yup, they do use the admin user as the trustee, and the tenant as the trustor. | 21:41 |
| elmiko | egafford: yea, should not be a problem. | 21:43 |
| *** witlessb has quit IRC | 22:23 | |
| *** esikachev has quit IRC | 22:30 | |
| *** esikachev has joined #openstack-sahara | 22:31 | |
| *** esikachev has quit IRC | 22:37 | |
| *** shakamunyi has quit IRC | 22:45 | |
| *** egafford has quit IRC | 22:47 | |
| openstackgerrit | Vitaly Gridnev proposed openstack/sahara-image-elements: Add hadoop-swift element on cloudera images https://review.openstack.org/201362 | 22:54 |
| *** sgotliv has quit IRC | 22:58 | |
| openstackgerrit | Merged openstack/sahara-image-elements: openjdk: revert to a semi-manual installation https://review.openstack.org/198034 | 23:03 |
| *** stanchan_ has joined #openstack-sahara | 23:10 | |
| *** stanchan has quit IRC | 23:12 | |
| openstackgerrit | Merged openstack/sahara-image-elements: openjdk: add mappings for openjdk-8 https://review.openstack.org/198035 | 23:16 |
| openstackgerrit | Merged openstack/sahara-image-elements: openjdk: install openjdk 8 for fedora >= 21 https://review.openstack.org/198036 | 23:17 |
| *** stanchan_ has quit IRC | 23:17 | |
| *** chlong-afk has joined #openstack-sahara | 23:18 | |
| *** egafford has joined #openstack-sahara | 23:33 | |
| *** chlong-afk has quit IRC | 23:36 | |
| *** egafford has quit IRC | 23:40 | |
| openstackgerrit | Merged openstack/sahara: [EDP][Oozie] external hdfs missed for URLs in job_configs https://review.openstack.org/192038 | 23:48 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!