Tuesday, 2016-09-06

openstackgerritDoron Chen proposed openstack/storlets: replace lxc_device with docker_device  https://review.openstack.org/36584403:56
eranromHi all, any chance we can  move the meeting today?06:41
eranromNow that I am able to edit the wiki I have posted an agenda for today: https://wiki.openstack.org/wiki/Meetings/Storlets#Agenda:07:18
eranromplease edit as needed.07:18
openstackgerritMerged openstack/storlets: Make storlet gateway totally independent from swift  https://review.openstack.org/36521707:36
kota_oh,07:49
kota_eranrom: just now, i noticed you wanted to move the meeting today07:49
eranromonly if possible.07:49
kota_eranrom: do you want to change the time? or day? or skipping?07:50
eranromI can attend only the first 15 mins07:50
kota_anything is ok to me though :-)07:50
eranrombest for me is to change the time, say 2 hours later??07:50
kota_I'm fine :-)07:51
kota_wait a sec, poke takashi with messenger07:51
eranromthanks!07:51
eranromother hours are good as well.07:52
kota_hmm, no response from him :/07:54
eranromok, worse case we can ask him in 6 mins :-)07:55
kota_:-)07:55
*** takashi has joined #openstack-storlets08:01
kota_hello takashi08:01
takashikota_: hi08:01
eranromHi08:02
eranromSorry I am late08:02
takashikota_: I just noticed about your message now08:02
eranromtakashi: are you ok with this?08:02
takashieranrom, kota_: I'm ok to have a meeting 2 hours later08:02
eranrombasically, any hour later is good for me.08:02
eranromtakashi: This is great! Thanks. I need to pick up my doughter from the kindergarder :-)08:03
eranromkindergarden08:03
kota_eranrom: nice08:03
eranromThanks!08:03
takashieranrom: very important job08:03
takashi:-)08:03
eranromindeed :-)08:03
eranromanyway, the agenda is updated.08:03
kota_to tell the truth I also am day off today for taking care of my daughters :P08:03
eranromfeel free to edit08:03
eranromkota_: important job :-)08:03
takashieranrom: OK. will have a look08:04
kota_see you in 2 hours ;-)08:04
eranromsee you. Thanks!08:04
takashisee you later, all08:04
openstackgerritTakashi Kajinami proposed openstack/storlets: Pass storlet language from gateway to daemon factory  https://review.openstack.org/36232608:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet execution  https://review.openstack.org/35998308:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Add Python sdaemon support in daemon_factory  https://review.openstack.org/35622408:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional test for python storlet execution in PUT  https://review.openstack.org/36504608:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement Python SDaemon  https://review.openstack.org/35825208:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Retrieve functest log for daemon-factory and storlets-daemon  https://review.openstack.org/36491108:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Change docs for Python storlet support  https://review.openstack.org/36521308:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Make python storlet daemon load applications  https://review.openstack.org/36330308:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Add range support for python storlet  https://review.openstack.org/36531708:19
openstackgerritTakashi Kajinami proposed openstack/storlets: Pass storlet language from gateway to daemon factory  https://review.openstack.org/36232608:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet execution  https://review.openstack.org/35998308:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Add Python sdaemon support in daemon_factory  https://review.openstack.org/35622408:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional test for python storlet execution in PUT  https://review.openstack.org/36504608:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement Python SDaemon  https://review.openstack.org/35825208:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Retrieve functest log for daemon-factory and storlets-daemon  https://review.openstack.org/36491108:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Change docs for Python storlet support  https://review.openstack.org/36521308:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Make python storlet daemon load applications  https://review.openstack.org/36330308:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Add range support for python storlet  https://review.openstack.org/36531708:23
openstackgerritTakashi Kajinami proposed openstack/storlets: Fix mistake about errno.EBADFD  https://review.openstack.org/36595208:57
openstackgerritMerged openstack/storlets: Prohibit duplicated names about storlet file and dependencies  https://review.openstack.org/36519009:13
takashiping kota_09:17
openstackgerritTakashi Kajinami proposed openstack/storlets: Pass storlet language from gateway to daemon factory  https://review.openstack.org/36232609:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement missing methods in client datagram  https://review.openstack.org/35251309:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add validation logic for python storlets  https://review.openstack.org/35620209:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet registration  https://review.openstack.org/35955309:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet execution  https://review.openstack.org/35998309:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add Python sdaemon support in daemon_factory  https://review.openstack.org/35622409:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional test for python storlet execution in PUT  https://review.openstack.org/36504609:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement Python SDaemon  https://review.openstack.org/35825209:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Retrieve functest log for daemon-factory and storlets-daemon  https://review.openstack.org/36491109:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Change docs for Python storlet support  https://review.openstack.org/36521309:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Make python storlet daemon load applications  https://review.openstack.org/36330309:18
openstackgerritTakashi Kajinami proposed openstack/storlets: Add range support for python storlet  https://review.openstack.org/36531709:18
eranromHi10:00
kota_hello10:00
takashieranrom, kota_: hi10:00
eranromHi, thanks again for delaying the meeting. lets start.10:01
eranromtakashi: wanna update on ptyhon status?10:01
takashieranrom: ok10:01
takashiI just updated the patch set, to address one problem about cancel interface.10:02
takashicurrently I'm hitting a bug, which is brought recently by my patch. I submitted a patch to fix that, so please have a look at it first.10:03
takashipatch 36595210:03
eranromok10:04
takashithis one. just rechecking  https://review.openstack.org/#/c/365952/10:04
eranromsure10:04
takashiso I'll list up current remaining items, including ones eranrom previously pointed out10:05
eranromperfect.10:05
kota_takashi: that looks correct10:05
takashi1. Fix the documentation. mainly we need to update the docs for application developers10:06
kota_why not adding the unit test to go the branch?10:06
takashikota_: I surely needed and need to do that.10:06
takashi:-(10:07
takashi2. Add validation about fds passed throught datagram10:07
takashi3. Refactor func test codes10:08
takashi4. Refactor directory to deal with java and python evenly10:08
takashieranrom: I think the 4th one is addressed in you patch, right?10:08
eranromright10:09
takashi5. add unit tests code for storlet agent10:09
takashi6. and for the last, consider wheather we can improve current interface design10:10
takashiIs there anything missed here, from your side? > eranrom, kota_10:11
* eranrom looking at his notes10:11
takashione more thing ... 7. refactor agent codes (daemon factory and storlet daemon) which has currently many duplicated codes10:12
kota_looks good for now, not sure sorry, i didn' t walk throught the python support patch yet10:12
eranromtakashi: the only thing I have to add is the non-graceful exit of the python storlet daemon10:13
eranromtakashi: you may have it hidden on your list.10:13
takashieranrom: thanks. I need to add that. maybe we need to add them to the etherpad (new one?)10:14
eranromtakashi: good idea.10:14
kota_sounds good.10:15
eranromtakashi: How do you want to proceed. land now, and fix later, define minimum set of fixes before landing, wait to complete everything?10:15
eranrompersonally, I rather not wait till is fixed (last option)10:16
eranromthat is till all is fixed10:16
takashiIMHO, I'd like to land current patch, because I don't want to keep such a long patch set for long time, to avoid possible many merge conflicts10:16
eranromagree.10:17
kota_ok10:17
kota_to be fair enough10:17
takashiOn the other hand, we still need to take some time to get the final decision about interface design. Currently it works, but we need to confirm 'it is the best'.10:17
kota_i wonder if you could add some docs for current work/constraint/remaining tasks we discuessed above10:17
eranromsuggestion: Lets put the list on the etherpad, and start landing according a given order (is the order in the blueprint good?)10:17
kota_maybe in commit message or inline docs?10:18
takashikota_: maybe I need to add some comments to code. will do that10:19
takashiDoes the following order makes sense to you?10:19
takashi1. Land current implementation10:19
takashi2. Freeze the python interface10:19
takashi3. Add docs for python implementation10:19
takashi4. Fix the remaining items10:20
takashiany comments?10:20
eranromtakashi: I am good with this. IMHO, the only thing to do before that is have the etherpad and the landing order.10:21
takashieranrom: absolutely10:21
kota_i don't mind the order through 2-4 because we may want to change any interface designes when making apps :P10:21
takashi0. Create a list for landing/work order10:21
kota_and not sure i could work a lot to freeze the design10:22
takashikota_: OK10:22
eranromkota_: Absolutely. From experience these things takes time to stabilise, and need to start somewhere...10:22
kota_that's not my code, upstream work is going forward always.10:22
eranrom:-)10:22
takashikota_: :-)10:22
takashieranrom: Maybe we need to think about the versioning about the application interface10:23
takashiat some point10:23
kota_ah, one point10:23
eranromtakashi: yes, we have in java a metadata stating the API version it supports.10:23
kota_application versioing or cut a release when changing interface10:23
takashieranrom: As far as I remember, I keep that metadata for Python. maybe we can start with validating it.10:24
takashiI'll list up this idea to the etherpad :-)10:25
eranromkota_: Generally this is a good practice. We need to decide when to start. So far there are no storlets....10:25
eranromtakashi: sounds good. We also need to decide where in the server side we keep the "current" suported set of APIs. This is also true for Java. So far we just ignored this since there is not yet10:26
eranromany storlets "out there"10:26
kota_eranrom: yeah10:26
eranrombut we certainly need to "design" this10:26
takashieranrom: agree10:26
eranromtakashi: I say that at this point lets add it to the list.10:27
eranromnumber 8?10:27
takashieranrom: yes10:27
eranromwe can discuss the details later.10:27
takashiso, I'll create an etherpad to list up tasks to land things and let you know that address10:27
eranromtakashi: great, thanks!10:27
kota_thanks10:27
takashiI'll rebase the current patch set after we fix the problem I mentioned at the first10:28
takashiand then we can start working to land current implementation10:28
eranromtakashi: sounds like a plan10:28
takashiok. will update you later10:28
eranromanything else on python for now?10:28
takashieranrom: nothing from my side10:29
kota_nothing10:29
takashifor now ;-)10:29
eranrom:)10:29
eranromok, big tent!10:29
eranromkota_: Are you comfortable with landing the py3 patch?10:29
kota_eranrom: sure, that's good to step at first10:30
eranromkota_: great.10:30
takashiI'll have a look and put +2 soon10:30
takashi+2+A10:30
eranromtakashi: thanks.10:30
kota_eranrom: at first, i'm trying to run all tests but almost doesn't work because swift doesn't support py3 yet10:30
kota_i hit a lot of swift bugs for py3 support :/10:31
eranromkota_: yes I saw that you skipped some. I think this is really ok10:31
eranromclearly we cannot support py3 before swift does.10:31
kota_yes, at least we need to update internal client to run unit test10:32
eranromkota_: you mean from the swift side?10:33
kota_yes10:33
kota_e.g. https://review.openstack.org/#/c/348272/10:33
kota_that is in my lader10:33
kota_ah, no10:34
eranromI see and Alistair's :-)10:34
kota_not from swift side10:34
kota_just run storlet unit tests with master swift10:34
kota_i mean10:35
eranromBTW - did we land the patch that takes master swift?10:36
eranromI think I did10:36
takashikota_: I suppose we need patch 364246, right? We need to test against swift master.10:36
eranromlet's land it then10:36
kota_maybe?10:37
eranromoh, its in WIP. I guess this is why I did not land it10:37
eranromshould I?10:37
kota_the reason i set WIP to it is just for testing if we can pass the gates10:37
kota_if it's reasonable for you all, okay let's land with updating the commit message (removing WIP ;-))10:38
takashikota_: looks good to me10:38
eranromagreed.10:39
kota_i think takashi knows the status supporting upstream swift (or not) than me.10:39
kota_more than10:39
eranromwell, for func testing we are still taking 2.7.010:39
kota_yes10:39
eranromI guess we should bump  this up as part of the "big-tent efforts"10:40
eranromwhich brings me to the next question:10:40
kota_that might break if we run func with swift master10:40
eranromright, but need to test :-)10:40
kota_that's a remaining task we are trying to use testr IMO10:40
kota_testr instead of nosetests10:40
takashikota_: testr is just a framework to run tests10:41
takashiwhat we need to work is about installation, I think10:41
eranromtakashi: right10:41
kota_ok, you know more than me about testr10:41
kota_takash:^^10:41
takashiwe need a installation mechanism, which can install swift master, keystone master and storlet master(or required patch version)10:42
eranromtakashi: right.10:42
takashilike we do in other projects, using devstack10:42
takashikota_: I previously had a look about that :-)10:42
kota_nice10:42
eranromtakashi: I believe that your installation remark ^^ is in the big-tent etherpad right?10:43
takashieranrom: meybe I only wrote about keystone version, but this surely should be addressed10:43
takashis/meybe/maybe/10:44
eranromtakashi: ok.10:44
takashieranrom: will update it10:44
eranromtakashi: thanks10:44
eranromAt any rate the first thing I would like to do now is get rid of build.xml where possible, and use setup.py10:45
eranromI will post a patch in the coming days10:45
kota_so great eranrom110:45
kota_110:45
kota_!10:45
eranromWe can decide what's next afterwards.10:46
takashieranrom: +10:46
takashi+110:46
eranromnext topic?10:46
kota_my keyboard touch was tippy10:46
kota_yup10:46
takashieranrom: yes10:47
eranromkota_: any update on the 'crazy stuff' patch?10:47
eranromkota_: we do not have to discuss, just if you want to bring up something...10:48
kota_eranrom: sorry no progress right now10:48
eranromkota_: ok np.10:48
eranromany other patches I should be looking at10:48
kota_I'd like to update some docs according to your comment10:48
kota_but I didn't have the time to do that10:49
eranromkota_: which docs?10:49
kota_crazy thing10:49
eranromkota_: right! forgot. I guess its mainly the commit message10:50
kota_yup10:51
kota_that adds an API so we definately describe it somewhere10:51
eranromkota_: ok. whenever you get to it.10:51
kota_ok10:51
eranromnext?10:51
kota_please10:52
takashiok10:52
eranromI have created an etherpad for the design summit. We still do not have space, but ttx was optimistic about it.10:52
eranromhttps://etherpad.openstack.org/p/storlets-otaca-design-summit10:52
eranromfeel free to add stuff. I guess that at this point itsmainly the fishbowl to worry about10:53
eranromthe working sessions can wait till we gwt closer to the date10:53
eranroms/gwt/get10:53
eranromkota_: btw I am still working on being able to update the IRC topic...10:54
eranromI do not have the permission10:55
eranromyet10:55
kota_oh10:55
takashieranrom: thanks! will dump some ideas.10:55
kota_who knows/is able to permit it?10:55
eranromkota_: infra guys10:56
eranromI got some help from fungi, but I guess I need more.10:56
eranromI will poke them later today10:56
kota_eranrom: ok10:56
eranromanything else for today?10:57
kota_thanks for trying!10:57
takashieranrom: after we get the permission we can add etherpad about big tent, and maybe design summit?10:57
takashiin the topic10:57
eranromtakashi: yeah that's the idea :-)10:57
takashieramrom: ok. thank you for working about that10:57
takashijust one small (?) thing10:58
eranromtakashi: sure10:58
takashiI remember that there are one thing mentioned in the rc meeting, 'the meeting log is not recorded'10:58
takashis/are/is10:58
eranromtakashi: yes10:58
takashishould we move our meeting to the openstack-meeting channel, to use meeting bot?10:59
eranromtakashi: item 10 in the etherpad. Its on my todo list :-)10:59
eranromsame hour - right?10:59
takashieranrom: if possible11:00
eranromtakashi: sure, I am on this.11:00
takashiI know the existing projects already 'booked' the time space11:00
eranromtakashi: right, but I guess our usual hour is not very popular :-)11:00
takashieranrom: agree11:00
kota_absolutely11:00
eranromI will update as soon as I make progress on this11:01
takashieranrom: I know it is not so urgent thing, but still seems better.11:01
eranromagree.11:01
eranromalso, should not be a big deal11:02
takashieranrom: I hope so :-)11:02
eranrom:-)11:02
kota_:-)11:02
eranromwe managed to cover all the agenda. Kudos to us :-)11:02
eranromanything else?11:02
takashithat's all from my side today. let us know if you decide to change the meeting room! (That's the reason why I asked this)11:02
eranromtakashi: sure11:02
kota_enough to me too.11:03
eranromok, thanks for joining. talk to you later. Its a great week for storelts!11:04
eranromstorlets :-)11:04
kota_thanks eranrom!11:04
takashieranrom: thanks!11:04
takashieranrom, kota_: I'm just drafting the etherpad for python storlet https://etherpad.openstack.org/p/storlets-python-support11:20
eranromtakashi: great. please ping when EBADFD: https://review.openstack.org/#/c/365952/ is ready.11:45
takashieranrom: I think you can land that now, if you are OK about adding unit test codes later.11:55
eranromtakashi: Am ok11:55
takashieranrom: can you put +2+A from your side?11:58
takashiI'll rebase my python support patches, after it get merged to master11:58
eranromsec...11:58
takashieranrom: thx!11:59
eranromtakashi: great. let me know when this is done, and I will start landing.11:59
eranromalso, if the order is different from the blueprint pls put the order in the etherpad. thanks!12:00
takashieranrom: I just update the note in blueprint (place range support in front of doc patch)12:02
takashieranrom: That shows current dependency order12:02
eranromtakashi: greart12:02
eranromso I will just need to know when the rebase are done.12:02
takashiI'll remove the -2 on the top patch, after I got +2s for all implementation patches (above doc patch).12:03
takashieranrom: yes. I'll ping you12:03
eranromtakashi: thanks!12:03
takashiafter removing -2, jenkins will start to merge all the changes to master12:03
*** takashi has quit IRC12:39
openstackgerritMerged openstack/storlets: Fix mistake about errno.EBADFD  https://review.openstack.org/36595213:24
*** takashi has joined #openstack-storlets13:47
openstackgerritTakashi Kajinami proposed openstack/storlets: Pass storlet language from gateway to daemon factory  https://review.openstack.org/36232613:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement missing methods in client datagram  https://review.openstack.org/35251313:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add validation logic for python storlets  https://review.openstack.org/35620213:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet registration  https://review.openstack.org/35955313:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet execution  https://review.openstack.org/35998313:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add Python sdaemon support in daemon_factory  https://review.openstack.org/35622413:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional test for python storlet execution in PUT  https://review.openstack.org/36504613:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement Python SDaemon  https://review.openstack.org/35825213:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Retrieve functest log for daemon-factory and storlets-daemon  https://review.openstack.org/36491113:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Change docs for Python storlet support  https://review.openstack.org/36521313:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Make python storlet daemon load applications  https://review.openstack.org/36330313:50
openstackgerritTakashi Kajinami proposed openstack/storlets: Add range support for python storlet  https://review.openstack.org/36531713:50
*** takashi has quit IRC14:24
*** takashi has joined #openstack-storlets15:04
openstackgerritTakashi Kajinami proposed openstack/storlets: Pass storlet language from gateway to daemon factory  https://review.openstack.org/36232615:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional tests for python storlet execution  https://review.openstack.org/35998315:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Add Python sdaemon support in daemon_factory  https://review.openstack.org/35622415:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Add functional test for python storlet execution in PUT  https://review.openstack.org/36504615:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Implement Python SDaemon  https://review.openstack.org/35825215:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Retrieve functest log for daemon-factory and storlets-daemon  https://review.openstack.org/36491115:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Change docs for Python storlet support  https://review.openstack.org/36521315:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Make python storlet daemon load applications  https://review.openstack.org/36330315:14
openstackgerritTakashi Kajinami proposed openstack/storlets: Add range support for python storlet  https://review.openstack.org/36531715:14
takashieranrom, kota_: I finished rebasing my codes about python storlet :-)15:47
takashiwhile rebasing, I found one new thing and added it to the python-storlet etherpad. That is about fd closing.15:48
takashiAlthough the current code works, it causes 'duplicated closing.' We can ignore inside _safe_close, but there still be some possible improvements15:49
takashis/ignore/ignore the errors caused by duplicated closing/15:49
takashiI also added one thing at the bottom of the list. need to check bashate linting about .functest and maybe .unittest file15:50
takashiand also check if there are other bash scripts currently we skip linting15:51
takashibut they can still be done as follow-up work. I think we can keep going ahead to land current implementation15:52
*** eranrom_ has joined #openstack-storlets16:58
*** eranrom has quit IRC17:00
eranrom_takashi: I am afraid you forgot to remove your -218:56

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