| *** hezhiqiang has joined #openstack-swift | 00:18 | |
| *** kevinc___ has quit IRC | 00:18 | |
| *** hezhiqiang has quit IRC | 00:24 | |
| *** NM has quit IRC | 00:44 | |
| *** lyrrad has quit IRC | 00:51 | |
| *** ccavanna has joined #openstack-swift | 00:54 | |
| *** haigang has joined #openstack-swift | 01:00 | |
| *** ccavanna_ has joined #openstack-swift | 01:00 | |
| *** ccavanna has quit IRC | 01:01 | |
| *** bapalm has quit IRC | 01:12 | |
| *** bapalm has joined #openstack-swift | 01:14 | |
| *** bapalm has quit IRC | 01:21 | |
| *** wbhuber_ has joined #openstack-swift | 01:33 | |
| *** bapalm has joined #openstack-swift | 01:33 | |
| *** bapalm has quit IRC | 01:38 | |
| *** jkugel has joined #openstack-swift | 01:44 | |
| *** bapalm has joined #openstack-swift | 01:44 | |
| *** bapalm has quit IRC | 01:51 | |
| *** haomaiwang has joined #openstack-swift | 01:55 | |
| *** bapalm has joined #openstack-swift | 01:56 | |
| *** bapalm has quit IRC | 02:03 | |
| *** hezhiqiang has joined #openstack-swift | 02:05 | |
| *** haomaiwang has quit IRC | 02:09 | |
| *** haomaiwa_ has joined #openstack-swift | 02:10 | |
| *** bapalm has joined #openstack-swift | 02:10 | |
| *** bapalm has quit IRC | 02:15 | |
| *** jrichli has joined #openstack-swift | 02:15 | |
| *** jkugel has left #openstack-swift | 02:16 | |
| *** bapalm has joined #openstack-swift | 02:23 | |
| *** bapalm has quit IRC | 02:30 | |
| *** bapalm has joined #openstack-swift | 02:33 | |
| *** wbhuber_ has quit IRC | 02:34 | |
| *** bapalm has quit IRC | 02:40 | |
| *** bapalm has joined #openstack-swift | 02:42 | |
| *** haomaiwa_ has quit IRC | 02:47 | |
| openstackgerrit | Pradeep Kumar Singh proposed openstack/swift: Emit warning log in object replicator https://review.openstack.org/185061 | 02:47 |
|---|---|---|
| *** haomaiwang has joined #openstack-swift | 02:47 | |
| notmyname | good evening | 02:49 |
| *** david__ has quit IRC | 02:51 | |
| *** bapalm has quit IRC | 02:52 | |
| *** bapalm has joined #openstack-swift | 02:55 | |
| *** haomaiwang has quit IRC | 02:58 | |
| *** haomaiwang has joined #openstack-swift | 02:58 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwa_ has joined #openstack-swift | 03:01 | |
| *** baojg has joined #openstack-swift | 03:01 | |
| *** bapalm has quit IRC | 03:04 | |
| *** gyee has quit IRC | 03:09 | |
| *** bapalm has joined #openstack-swift | 03:10 | |
| *** baojg has quit IRC | 03:18 | |
| jrichli | notmyname: good evening. looks like you and Joe had some fun at IDF! | 03:21 |
| notmyname | :-) | 03:21 |
| notmyname | they were giving away these, and I picked one up https://en.wikipedia.org/wiki/Intel_Galileo | 03:22 |
| *** sanchitmalhotra has joined #openstack-swift | 03:23 | |
| *** baojg has joined #openstack-swift | 03:24 | |
| jrichli | sounds neat. any specific plans for it yet? | 03:24 |
| notmyname | nope | 03:25 |
| notmyname | for the time being it will probably sit next to my second raspberry pi (unused) | 03:25 |
| jrichli | lol. :-) too busy with 'real' work i guess | 03:26 |
| notmyname | yeah, I'd probably have to stop doing swift stuff on nights and weekends ;-) | 03:26 |
| jrichli | i hear ya! | 03:26 |
| *** sanchitmalhotra1 has joined #openstack-swift | 03:31 | |
| *** sanchitmalhotra has quit IRC | 03:33 | |
| *** baojg has quit IRC | 03:47 | |
| *** baojg has joined #openstack-swift | 03:48 | |
| *** ppai has joined #openstack-swift | 03:51 | |
| ho | ping kota_, dmorita i would like to know miyahara-san's irc name. | 03:58 |
| *** breitz has quit IRC | 04:10 | |
| *** haigang has quit IRC | 04:27 | |
| zaitcev | ho: I'd poke km and ask... | 04:30 |
| *** jrichli has quit IRC | 04:39 | |
| *** bapalm has quit IRC | 04:39 | |
| *** bapalm has joined #openstack-swift | 04:41 | |
| *** bapalm has quit IRC | 04:46 | |
| *** bapalm has joined #openstack-swift | 04:47 | |
| *** bapalm has quit IRC | 04:54 | |
| *** baojg has quit IRC | 04:55 | |
| *** NM has joined #openstack-swift | 04:55 | |
| *** bapalm has joined #openstack-swift | 04:56 | |
| *** bapalm has quit IRC | 05:03 | |
| *** bapalm has joined #openstack-swift | 05:07 | |
| *** bapalm has quit IRC | 05:16 | |
| *** bapalm has joined #openstack-swift | 05:20 | |
| *** kairo has quit IRC | 05:22 | |
| *** NM has quit IRC | 05:23 | |
| *** kairo has joined #openstack-swift | 05:25 | |
| *** bapalm has quit IRC | 05:34 | |
| *** bapalm has joined #openstack-swift | 05:35 | |
| *** bapalm has quit IRC | 05:44 | |
| ho | zaitcev: thanks! do you have any response? | 05:55 |
| zaitcev | ho: I didn't ask. I thought you would. | 05:55 |
| *** bapalm has joined #openstack-swift | 05:56 | |
| ho | zaitcev: oh, i see. i tried it in japanse, now waiting... | 06:00 |
| *** zaitcev has quit IRC | 06:01 | |
| *** baojg has joined #openstack-swift | 06:03 | |
| *** baojg has quit IRC | 06:09 | |
| *** bapalm has quit IRC | 06:15 | |
| *** bapalm has joined #openstack-swift | 06:16 | |
| *** bapalm has quit IRC | 06:20 | |
| ho | kazuhiro around? | 06:44 |
| *** joeljwright has joined #openstack-swift | 06:52 | |
| *** ChanServ sets mode: +v joeljwright | 06:52 | |
| *** hrou has quit IRC | 07:00 | |
| *** mahatic has joined #openstack-swift | 07:07 | |
| *** baojg has joined #openstack-swift | 07:08 | |
| *** baojg has quit IRC | 07:08 | |
| *** rledisez has joined #openstack-swift | 07:08 | |
| *** geaaru has joined #openstack-swift | 07:46 | |
| *** haomaiwa_ has quit IRC | 07:49 | |
| *** haomaiwang has joined #openstack-swift | 07:49 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:02 | |
| *** joeljwright has quit IRC | 08:02 | |
| *** vinsh has quit IRC | 08:03 | |
| *** jordanP has joined #openstack-swift | 08:04 | |
| *** onovy has quit IRC | 08:08 | |
| *** hezhiqiang has quit IRC | 08:25 | |
| *** jistr has joined #openstack-swift | 08:30 | |
| *** wbhuber_ has joined #openstack-swift | 08:30 | |
| *** wbhuber_ has quit IRC | 08:34 | |
| *** sayali has joined #openstack-swift | 08:47 | |
| *** ho has quit IRC | 09:00 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-swift | 09:02 | |
| *** sayali has quit IRC | 09:04 | |
| *** sayali has joined #openstack-swift | 09:08 | |
| *** jodah has quit IRC | 09:10 | |
| *** jith_ has joined #openstack-swift | 09:11 | |
| jith_ | hi all.. is it mandatory to create swiftOperator role.. what is the use of it? | 09:12 |
| *** aix has joined #openstack-swift | 09:19 | |
| *** baojg has joined #openstack-swift | 09:27 | |
| *** baojg has quit IRC | 09:27 | |
| *** onovy has joined #openstack-swift | 09:37 | |
| *** km has quit IRC | 09:49 | |
| *** haomaiwang has quit IRC | 10:01 | |
| *** izadorozhna has left #openstack-swift | 10:01 | |
| *** haomaiwa_ has joined #openstack-swift | 10:02 | |
| openstackgerrit | Kazuhiro MIYAHARA proposed openstack/swift: Fix dispersion-reports error message https://review.openstack.org/213697 | 10:05 |
| janonymous | jith_ : I think no, Roles could be any name, if used with keystone the entry must be present in its roles on server side also. swiftoperator is just for specification of swift role | 10:06 |
| jith_ | janonymous: thanks.. i am using with keystone only... for swift user i will create the role as admin... may be for ther users like demo i have to assign swiftoperator?? is it like so? | 10:08 |
| *** aix has quit IRC | 10:10 | |
| *** kei_yama has quit IRC | 10:16 | |
| *** madhuri_ has joined #openstack-swift | 10:19 | |
| * madhuri_ slaps janonymous around a bit with a large fishbot | 10:19 | |
| *** aix has joined #openstack-swift | 10:22 | |
| janonymous | jith : yes you can assign swiftoperator role or simply use admin and change that in config of proxy. | 10:22 |
| janonymous | madhuri_ : Hahaa yes Irc is changed and has some new features :) | 10:23 |
| madhuri_ | M sorry was just trying around.. Haha :) | 10:25 |
| *** madhuri_ has left #openstack-swift | 10:25 | |
| janonymous | Okay.. :) | 10:26 |
| *** mahatic has quit IRC | 10:28 | |
| jith_ | janonymous: is it possible for demo user to access swift if he is not admin | 10:30 |
| *** sanchitmalhotra has joined #openstack-swift | 10:44 | |
| *** sanchitmalhotra1 has quit IRC | 10:47 | |
| *** haigang has joined #openstack-swift | 10:52 | |
| janonymous | No. it has to have admin rights or reseller rights. | 10:54 |
| janonymous | Or the swiftoperator might be configured with admin rights as well for it. | 10:55 |
| *** haomaiwa_ has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-swift | 11:02 | |
| *** mahatic has joined #openstack-swift | 11:10 | |
| *** sanchitmalhotra has quit IRC | 11:24 | |
| jith_ | janonymous: thanks | 11:24 |
| jith_ | but i didnt get fully.. any documentation for this? | 11:25 |
| jith_ | default one is swiftoperator? | 11:25 |
| *** joeljwright has joined #openstack-swift | 11:28 | |
| *** ChanServ sets mode: +v joeljwright | 11:28 | |
| janonymous | I think _member is default | 11:28 |
| janonymous | Keystone docs might be helpful for this. | 11:29 |
| janonymous | you could add multiple roles in proxy.conf also | 11:29 |
| janonymous | http://docs.openstack.org/havana/install-guide/install/apt/content/ch_keystone.html | 11:31 |
| *** SkyRocknRoll has joined #openstack-swift | 11:37 | |
| *** haomaiwang has quit IRC | 11:42 | |
| *** sayali has quit IRC | 11:44 | |
| *** sayali has joined #openstack-swift | 11:45 | |
| *** nadeem has joined #openstack-swift | 11:48 | |
| *** nadeem has quit IRC | 11:49 | |
| *** kairo has quit IRC | 12:12 | |
| *** kairo has joined #openstack-swift | 12:12 | |
| *** ahale has quit IRC | 12:12 | |
| *** petertr7_away is now known as petertr7 | 12:12 | |
| *** ahale has joined #openstack-swift | 12:12 | |
| kairo | morning. | 12:14 |
| *** alejandrito has joined #openstack-swift | 12:20 | |
| *** vinsh has joined #openstack-swift | 12:26 | |
| *** nakagawamsa has quit IRC | 12:31 | |
| *** abhirc has joined #openstack-swift | 12:44 | |
| *** SkyRocknRoll has quit IRC | 12:49 | |
| *** sayali has quit IRC | 12:51 | |
| *** joeljwright has quit IRC | 12:57 | |
| *** chlong has joined #openstack-swift | 12:58 | |
| *** wbhuber_ has joined #openstack-swift | 12:58 | |
| *** NM has joined #openstack-swift | 12:59 | |
| *** dustins has joined #openstack-swift | 13:02 | |
| *** chlong has quit IRC | 13:07 | |
| *** haigang has quit IRC | 13:10 | |
| *** chlong has joined #openstack-swift | 13:11 | |
| *** peluse_ is now known as peluse | 13:11 | |
| *** ChanServ sets mode: +v peluse | 13:12 | |
| *** vinsh has quit IRC | 13:12 | |
| *** chlong has quit IRC | 13:15 | |
| *** petertr7 is now known as petertr7_away | 13:21 | |
| *** hrou has joined #openstack-swift | 13:21 | |
| *** ppai has quit IRC | 13:24 | |
| *** wbhuber has quit IRC | 13:25 | |
| *** wbhuber_ is now known as wbhuber | 13:25 | |
| *** jkugel has joined #openstack-swift | 13:25 | |
| *** chlong has joined #openstack-swift | 13:29 | |
| *** jrichli has joined #openstack-swift | 13:33 | |
| *** vinsh has joined #openstack-swift | 13:34 | |
| *** vinsh has quit IRC | 13:35 | |
| *** vinsh_ has joined #openstack-swift | 13:35 | |
| *** haigang has joined #openstack-swift | 13:38 | |
| *** ccavanna_ has quit IRC | 13:39 | |
| *** abhirc has quit IRC | 13:50 | |
| *** lcurtis has joined #openstack-swift | 14:06 | |
| *** ccavanna has joined #openstack-swift | 14:12 | |
| *** petertr7_away is now known as petertr7 | 14:15 | |
| *** dbarman has joined #openstack-swift | 14:16 | |
| *** jlhinson has joined #openstack-swift | 14:17 | |
| *** haigang has quit IRC | 14:21 | |
| *** haomaiwang has joined #openstack-swift | 14:26 | |
| openstackgerrit | Stuart McLaren proposed openstack/python-swiftclient: Add minimal working service token support. https://review.openstack.org/182640 | 14:31 |
| *** mahatic_ has joined #openstack-swift | 14:42 | |
| *** mahatic has quit IRC | 14:42 | |
| *** zul has quit IRC | 14:44 | |
| *** zul has joined #openstack-swift | 14:46 | |
| *** minwoob has joined #openstack-swift | 14:58 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** petertr7 is now known as petertr7_away | 15:02 | |
| *** haomaiwang has joined #openstack-swift | 15:02 | |
| *** nadeem has joined #openstack-swift | 15:04 | |
| *** nadeem has quit IRC | 15:05 | |
| *** nadeem has joined #openstack-swift | 15:05 | |
| *** pgbridge has quit IRC | 15:08 | |
| *** haigang has joined #openstack-swift | 15:18 | |
| *** nadeem has quit IRC | 15:18 | |
| *** bapalm has joined #openstack-swift | 15:21 | |
| *** nadeem has joined #openstack-swift | 15:26 | |
| *** nadeem has quit IRC | 15:26 | |
| *** nadeem has joined #openstack-swift | 15:27 | |
| *** petertr7_away is now known as petertr7 | 15:28 | |
| *** zul has quit IRC | 15:37 | |
| notmyname | good morning | 15:38 |
| ccavanna | torgomatic: Hi. Thanks for the UT code. | 15:46 |
| ccavanna | torgomatic: Can I ask you a question about one function in the change? | 15:46 |
| *** gyee has joined #openstack-swift | 15:46 | |
| notmyname | ccavanna: don't ask to ask. just ask :-) | 15:46 |
| notmyname | (also, torgomatic normally isn't around for another hour or so) | 15:47 |
| ccavanna | notmyname: oh, I thought he was up. I'll wait then. Thanks. | 15:47 |
| ccavanna | notmyname: and good morning ! | 15:47 |
| *** jordanP has quit IRC | 15:47 | |
| notmyname | it's ok to go ahead and ask. he'll see it when he gets on line, and it may also be answered by someone else who is online now | 15:47 |
| ccavanna | Great. | 15:48 |
| ccavanna | I got this diff with code changes for UTs from Sam. https://gist.github.com/smerritt/953023d09b0ee1247af3 | 15:49 |
| ccavanna | In this method: assertUpdateStats, the first line is sorting the statistics values. | 15:50 |
| ccavanna | I'm not sure why. | 15:50 |
| ccavanna | Probably for the comparison at the end, as long as we pass all the expected values in order. | 15:52 |
| *** haigang has quit IRC | 15:55 | |
| *** abhirc has joined #openstack-swift | 15:55 | |
| torgomatic | ccavanna: that's because it doesn't matter what order the updates are in. The ordering of the packets containing object.GET.blahblah and object.policy.2.GET.blahblah is completely irrelevant, so I didn't want the test asserting one came before the other | 16:00 |
| *** haomaiwang has quit IRC | 16:01 | |
| ccavanna | torgomatic: thanks and thanks for the code. I wasn't really sure how you wanted the UTs. Are those changes commited? | 16:02 |
| albertom | do i have to create an account.builder and container builder for each policy ? | 16:02 |
| ccavanna | torogomatic: or do you want me to check them in with my change? | 16:02 |
| *** haomaiwang has joined #openstack-swift | 16:02 | |
| torgomatic | ccavanna: nope; I sent a diff for you to evaluate and take what you liked from it, but that's all I did | 16:02 |
| ccavanna | torgomatic: thank you, it was very helpful. | 16:03 |
| torgomatic | ccavanna: you're welcome | 16:03 |
| *** jistr has quit IRC | 16:05 | |
| torgomatic | albertom: nope, just object rings | 16:07 |
| *** petertr7 is now known as petertr7_away | 16:07 | |
| *** chsc has joined #openstack-swift | 16:09 | |
| *** zul has joined #openstack-swift | 16:15 | |
| *** rledisez has quit IRC | 16:25 | |
| *** lyrrad has joined #openstack-swift | 16:25 | |
| *** NM has quit IRC | 16:37 | |
| *** arringtp_ has joined #openstack-swift | 16:56 | |
| openstackgerrit | Stuart McLaren proposed openstack/python-swiftclient: Add minimal working service token support. https://review.openstack.org/182640 | 16:57 |
| *** zul has quit IRC | 16:59 | |
| *** NM has joined #openstack-swift | 17:00 | |
| *** haomaiwang has quit IRC | 17:01 | |
| *** haomaiwa_ has joined #openstack-swift | 17:02 | |
| *** abhirc has quit IRC | 17:06 | |
| openstackgerrit | Doug Hellmann proposed openstack/python-swiftclient: Update .gitreview for stable/juno https://review.openstack.org/215716 | 17:16 |
| notmyname | peluse: https://www.youtube.com/watch?v=MnTEpvK7LAE | 17:17 |
| openstackgerrit | Tim Burke proposed openstack/swift: Add the ability to specify ranges for SLO segments https://review.openstack.org/211726 | 17:22 |
| *** zul has joined #openstack-swift | 17:24 | |
| *** albertom has quit IRC | 17:26 | |
| notmyname | my thoughts on the hackathon last week https://swiftstack.com/blog/2015/08/21/openstack-swift-hackathon-august2015/ | 17:28 |
| *** albertom has joined #openstack-swift | 17:29 | |
| wbhuber | notmyname: reading.... | 17:30 |
| *** abhirc has joined #openstack-swift | 17:30 | |
| wbhuber | it looks good. covered the most pressing topics. things that can be done pre tokyo and post tokyo. thx for writing it up. | 17:33 |
| haypo | peluse: hi. any news from tsg for the pyeclib 1.0.8 release? | 17:36 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 17:37 |
| hrou | notmyname, great summary, thanks for that ! Will share with folks here in the lab, we had written up a summary (better described as an amalgamation of notes) but this is much more concise. | 17:37 |
| acoles | peluse: FYI - still WIP but making progress: patch 215276 | 17:38 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ | 17:38 |
| *** NM has left #openstack-swift | 17:40 | |
| *** janonymous_ has joined #openstack-swift | 17:41 | |
| *** acoles is now known as acoles_ | 17:42 | |
| minwoob | Does anyone know of a way off the top of their head, a method to get the path of a fragment with its etag? | 17:45 |
| haypo | notmyname: hi. by the way, can you tell me how i can make progress on my "fix tox -e py34" patch? https://review.openstack.org/#/c/199034/ -- thanks | 17:45 |
| haypo | this patch got a +2 from Pete Zaitcev | 17:46 |
| openstackgerrit | Carlos Cavanna proposed openstack/swift: Improving statistics sent to Graphite. https://review.openstack.org/202657 | 17:46 |
| jrichli | notmyname: great summary! | 17:48 |
| *** aix has quit IRC | 17:49 | |
| *** geaaru has quit IRC | 17:56 | |
| *** delattec has quit IRC | 17:58 | |
| *** cdelatte has quit IRC | 17:58 | |
| *** zhill has joined #openstack-swift | 18:00 | |
| *** haomaiwa_ has quit IRC | 18:01 | |
| *** mahatic_ has quit IRC | 18:01 | |
| *** haomaiwang has joined #openstack-swift | 18:02 | |
| *** petertr7_away is now known as petertr7 | 18:12 | |
| openstackgerrit | Carlos Cavanna proposed openstack/swift: New troubleshooting case in documentation. https://review.openstack.org/215744 | 18:16 |
| *** albertom has quit IRC | 18:18 | |
| *** dbarman has quit IRC | 18:19 | |
| *** pgbridge has joined #openstack-swift | 18:22 | |
| *** albertom has joined #openstack-swift | 18:23 | |
| *** albertom has quit IRC | 18:39 | |
| *** albertom has joined #openstack-swift | 18:45 | |
| clayg | heyoh! | 18:45 |
| *** bwall is now known as deez_nutz | 18:46 | |
| *** deez_nutz is now known as bwall | 18:46 | |
| *** bwall is now known as obama | 18:47 | |
| *** obama is now known as bwall | 18:48 | |
| *** abhirc has quit IRC | 18:55 | |
| *** janonymous_ has quit IRC | 18:56 | |
| minwoob | clayg: Are you aware of a method that directly retrieves the fragment's path from its etag? | 18:58 |
| minwoob | (Similar to how you can get the metadata from the fragment body). | 18:58 |
| clayg | minwoob: the fragments *path* like /account/container/body ? | 18:59 |
| clayg | er... /a/c/o | 18:59 |
| clayg | I don't know why i typed body | 18:59 |
| minwoob | Right. | 18:59 |
| clayg | minwoob: no, the etag is from the body - that kind of information is all metadata-ish | 19:00 |
| minwoob | Not sure if that will work with shutil.move() but it would get us closer. | 19:00 |
| minwoob | I see. | 19:00 |
| clayg | oh you mean like /srv/node/sdb/part/suffix/hash/frag#1.data ? | 19:00 |
| *** haomaiwang has quit IRC | 19:01 | |
| clayg | i mean - not that you could get that from the etag :P | 19:01 |
| *** haomaiwang has joined #openstack-swift | 19:02 | |
| clayg | how did you get the etag w/o knowning something else about it - just having the etag is like a weird starting point - it's like a macgyver setup - except even macgyver would be like "wtf am I going to do with an etag? is there anything else in here?" | 19:02 |
| minwoob | Hmm | 19:02 |
| clayg | heh | 19:02 |
| minwoob | From direct_get(objectpart, object node) | 19:02 |
| *** haypo has quit IRC | 19:02 | |
| minwoob | object part, object node which in turn was returned by the ring. | 19:03 |
| clayg | that's in some probetest some where right? it may have even been moved to common or some shiz - i think those tests have some attributes set on them in setup like "self.account, self.container_name, self.object_name" or something | 19:03 |
| minwoob | Yeah, exactly. | 19:03 |
| clayg | yeah go look at what direct_get does - it must have some state that i can pull from | 19:03 |
| minwoob | I'm pretty sure just etag but it may be returning more than just that. | 19:05 |
| minwoob | (I'm looking at test_reconstructor_revert.py btw). | 19:05 |
| clayg | minwoob: ok, yeah that method might just be throwing out TMI - you can get the frag index from the the headers - but it's probably determistic - if you know the acc/cont/o and the node you should be able to figure out the fragment index and the path on disk | 19:05 |
| minwoob | (under tests/probe). | 19:05 |
| minwoob | I see. | 19:06 |
| clayg | minwoob: i'm not sure why but it seems move of the new probetests don't reall go all the way down to disk | 19:09 |
| clayg | test_object_failures has some wonky junk | 19:10 |
| clayg | I think it might be better to try something with DiskFileManager.yeild_hashes(device, part) - maybe? | 19:11 |
| clayg | minwoob: maybe look at how swift-get-nodes works (it'd be good to know ayway) - it might be possible to split up cli.info.print_ring_locations into utils.get_ring_locations and cli.info.display_ring_locations | 19:15 |
| clayg | get_ring_locations might return a list of dicts, but instead of *just* the node/device keys (e.g. 'ip', 'port', 'device', etc.) it would also have like 'hash', 'suffix', 'part', 'storage_dir', 'path', 'handoff' (boolean), 'policy' | 19:17 |
| clayg | minwoob: or you can just crib what's its doing and roll your own | 19:17 |
| minwoob | clayg: I see. Once it's been -- artificially -- moved to a primary node though, will the reconstructor be able to pick up on the artificially moved fragment? | 19:21 |
| minwoob | By artificially, I mean by using a utility such as shutil.move() rather than a ring rebalance. | 19:22 |
| minwoob | From what I see, shutil.move() is not being used anywhere else in the code. | 19:23 |
| minwoob | That's what I was planning to go forward with. | 19:23 |
| clayg | minwoob: yeah we're definately off in crazy town - you're going to do great tho | 19:23 |
| clayg | minwoob: the reconstructor is going to deal with whatever he finds in the directory - it doesn't really know how it got there - so for that part shutil.move is going to be fine | 19:24 |
| clayg | minwoob: there's some chance (if the reconstructor has already run) that it won't find anything in that directory tho because it doesn't know to go looking for it | 19:24 |
| clayg | if it seems like at some point some background process isn't noticing that you changed something out from underneath it you may have to call invalidate_hash (or just rm hashes.pkl) | 19:25 |
| clayg | or maybe just fork lift the whole part - that'd probably be better than surgery | 19:26 |
| clayg | like fail a couple of primaries, then lift the parts off the handoffs - but drop them on the wrong nodes (swapped) - the scenario we're acctually trying to cover for is a node that took a handoff suddenly becoming a primary and finding he's holding the wrong frag | 19:27 |
| peluse | notmyname, cool thanks | 19:27 |
| clayg | so moving handoff frags onto the wrong primary seems like a pretty good proxy I think? | 19:27 |
| clayg | peluse: !!! | 19:27 |
| peluse | mr clayg! | 19:28 |
| clayg | how was the big presentation conference thingy?! | 19:28 |
| clayg | peluse: did you sell a bunch of cpus? | 19:28 |
| peluse | great prese O think, weak attendance :( | 19:28 |
| peluse | we had a crappy timeslot, most of the tech sessions were very light on the last day just before lunch. oh well | 19:28 |
| peluse | clayg, so hey I'm spending most of today getting the cluster ready for next week's fun. I saw your etag buckets comments/updates - are you pushing more stuff or want me to run with what's up there this weekend (try to address thingss) or what are ya thinkin'? | 19:29 |
| clayg | peluse: yes - next week is key | 19:30 |
| minwoob | clayg: I like the idea of moving the part around, rather than the fragment -- can only have one fragment per object, on a part, anyway. | 19:30 |
| clayg | peluse: you should work on getting ready for that | 19:31 |
| clayg | peluse: but we neede etag buckets! | 19:31 |
| clayg | hrm... | 19:31 |
| clayg | minwoob: well - heh - unless your a primary :D | 19:31 |
| *** andrey-mp has joined #openstack-swift | 19:34 | |
| peluse | clayg, OK so keep me posted on etag buckets. for now I'll use master for the test cluster | 19:34 |
| clayg | peluse: i guess I'm thinking the bucket response and spawn extra requests is exactly the right thing to do - so it's all tactical now | 19:34 |
| *** tsg has joined #openstack-swift | 19:35 | |
| clayg | peluse: I thinking handling mixed 206/416 in ranged responses is about the same as handling mixed etags and they should both work the same | 19:35 |
| clayg | peluse: I think we should probably group by timestamp instead of etag | 19:35 |
| peluse | clayg, ahhh.... | 19:35 |
| clayg | peluse: I think there could be more failures to come out of additional testing | 19:35 |
| minwoob | Right. | 19:35 |
| peluse | clayg, ain't that the truth always? | 19:36 |
| clayg | peluse: I think acoles made a comment about spawning as many extra requests as needed to service the difference of min-gets and max-bucket was spot on - and I snuck that in my follow up | 19:36 |
| clayg | ... so ... | 19:37 |
| clayg | well that's it really | 19:37 |
| clayg | i don't know what to do now except keep working on it | 19:37 |
| peluse | clayg, cool | 19:37 |
| peluse | clayg, thanks | 19:37 |
| andrey-mp | Hi, Who is responsible for swift3 plugin? I sent quiestions to mailing list and to all people from swift3-core... but no answer... | 19:38 |
| clayg | i have some bandwidth today - might be able to move it forward - i could start squashing stuff over your change; or doing more diffs and/or dependent patches - i think there's other core's around - we might could try to merge something today if that would be helpful? | 19:38 |
| clayg | andrey-mp: try timburke or kota_ (where is kota_ ?) | 19:38 |
| peluse | clayg, yes that would be HUGE | 19:38 |
| clayg | ctennis: are you guys going to be able to deploy master next week? | 19:39 |
| notmyname | andrey-mp: I'm not exactly sure what your proposed change means | 19:39 |
| ctennis | clayg: he plan is to just put git master on the machines and point the existing directories to them | 19:39 |
| andrey-mp | notmyname: I can describe it :) | 19:40 |
| notmyname | andrey-mp: what does it mean to "...add to swift3 project pipeline tempest job." | 19:40 |
| andrey-mp | swift3 doesn't have any functional tests in gating | 19:40 |
| andrey-mp | i propose to add tempest job to swift3 gating to check functionality in real cloud | 19:41 |
| clayg | andrey-mp: do other ec2 compatibility api shims *gate* changes to openstack? | 19:41 |
| clayg | or you want to gate changes to swift3? against some tempest functional tests? | 19:42 |
| andrey-mp | clayg: i don't fully understand your phrase... but i try to answer - ec2 has three functional gate jobs | 19:42 |
| andrey-mp | i want to gate changes with swift3 against some tempest tests. yes :) | 19:43 |
| andrey-mp | tempest.thirdparty.boto - these tests have s3 requests | 19:44 |
| clayg | yeah ok - kota would probably be down for that - sounds reasonable to me <- timburke | 19:44 |
| clayg | andrey-mp: do they pass currently for you when you run them localy on your setup? | 19:44 |
| timburke | andrey-mp: do we already know that current swift3 has sufficient compatibility with the S3 api to pass? i know there are still some blind spots... | 19:44 |
| andrey-mp | clayg: yes. We have gate job for ec2 that enables swift3 plugin and runs tempest and our ec2 functional tests. | 19:45 |
| andrey-mp | timburke: all tests that we have are passed against cloud with swift3 enabled | 19:46 |
| *** bkopilov has quit IRC | 19:47 | |
| timburke | andrey-mp: then, yeah, i'm all for it! i know we were talking about doing something similar for the swift gate at the hackathon. (non-voting, though; swift3 shouldn't dictate what can land in swift) | 19:47 |
| *** bkopilov has joined #openstack-swift | 19:47 | |
| clayg | notmyname: ^ sounds good to me, does andrey-mp have to do anything special to work with infra to get the non-voting job setup? | 19:47 |
| *** alejandrito has quit IRC | 19:48 | |
| clayg | timburke: but if the tests are reliable and provide useful failures it'd be cool to have them gate changes to swift3 (start non-voting obviously) | 19:48 |
| andrey-mp | yeah, I'll do it :) I made same job (non-voting) for our ec2 project - so it will be simple | 19:48 |
| notmyname | what is the "ec2 project" you're talking about? | 19:48 |
| timburke | clayg: *nod* | 19:49 |
| andrey-mp | this is - https://github.com/stackforge/ec2-api | 19:49 |
| clayg | is it holiday in Japan? I'm guessing either kota didn't see the message or he's OOO | 19:49 |
| notmyname | andrey-mp: and are you wanting to add these checks to swift3 or to swift? | 19:49 |
| andrey-mp | is a standalone ec2 api to openstack | 19:49 |
| andrey-mp | notmyname: swift3 | 19:49 |
| clayg | notmyname: he said swift3 | 19:49 |
| notmyname | clayg: it's saturday in japan. also 4:50 am | 19:49 |
| clayg | notmyname: ah! | 19:49 |
| clayg | andrey-mp: the thing about "no one responded" with a global community is - you need to give it a couple three business days :P | 19:51 |
| clayg | IRC is great tho, good call jumping in here :D | 19:51 |
| andrey-mp | clayg: ok :) I'm worrying because my previous mail to mailing list has no response too, but I sent it four days ago :) | 19:52 |
| clayg | andrey-mp: heh, fair enough - swift3+keystone it probably like a subgroup of a subgroup - good for you! | 19:56 |
| clayg | andrey-mp: is this like a s3 http server that just writes stright to a filesystem -> https://github.com/stackforge/ec2-api/blob/master/ec2api/s3/s3server.py | 19:57 |
| andrey-mp | clayg: yes. this is improved copy-paste from nova. nova team deprecated ec2/s3 in nova project | 19:59 |
| andrey-mp | clayg: but this implementation mostly for testing not for real cloud. also it doesn't have authentication at all | 20:00 |
| *** haomaiwang has quit IRC | 20:01 | |
| *** 7JTAAIXB0 has joined #openstack-swift | 20:02 | |
| clayg | i keep laughing at "real cloud" - like ... idk, i'm not a meteorologist :) | 20:05 |
| *** petertr7 is now known as petertr7_away | 20:05 | |
| minwoob | lol | 20:05 |
| andrey-mp | :) | 20:06 |
| clayg | yay Friday! | 20:07 |
| andrey-mp | I dislike a bit unit tests because it works with ephemeral layers and I call it fake cloud :) | 20:07 |
| clayg | even better! :) | 20:08 |
| andrey-mp | so, something like this - https://review.openstack.org/#/c/215766/ | 20:13 |
| clayg | andrey-mp: I honstly have no idea ^ notmyname timburke | 20:14 |
| *** marcusvrn_ has quit IRC | 20:15 | |
| albertom | hi | 20:17 |
| notmyname | hi | 20:17 |
| albertom | how can i create a container with policy-1 ? | 20:17 |
| albertom | swift -V3 post containername <==== this creates a container with the default policy | 20:18 |
| clayg | albertom: -H 'x-container-policy: <policy-name-from-/info>' | 20:18 |
| clayg | ^ works with curl or swift | 20:18 |
| albertom | swift: error: no such option: -H | 20:19 |
| clayg | whoa - `sudo pip install --upgrade python-swiftclient` maybe? | 20:20 |
| albertom | swift 2.3.1 | 20:20 |
| albertom | I have that one | 20:20 |
| clayg | albertom: mine says python-swiftclient 2.4.1.dev21 | 20:21 |
| clayg | --header <header> Set request headers. This option may be repeated. | 20:21 |
| clayg | Example -H "content-type:text/plain" | 20:21 |
| clayg | albertom: oh is that the version of *swift* or python-swiftclient? you did `swift --version` | 20:21 |
| notmyname | it's swift client | 20:21 |
| albertom | I did swift --version | 20:21 |
| albertom | on ubuntu | 20:21 |
| notmyname | albertom: it hink you put the -H in the wrong place | 20:22 |
| albertom | also, it would be nice if horizon could let me choose the policy | 20:22 |
| *** tsg has quit IRC | 20:24 | |
| albertom | oh right | 20:25 |
| albertom | -H after post | 20:25 |
| notmyname | yeah. just confirmed locally | 20:25 |
| *** bkopilov has quit IRC | 20:27 | |
| albertom | swift -V3 post -H 'X-Storage-Policy: silver' silver | 20:27 |
| albertom | yeah that works | 20:28 |
| albertom | thanks! | 20:28 |
| albertom | do youy know if there is a patch already in horizon ? | 20:28 |
| andrey-mp | thanks all! | 20:28 |
| *** tsg has joined #openstack-swift | 20:30 | |
| notmyname | albertom: nope. I don't know | 20:34 |
| *** albertom has quit IRC | 20:34 | |
| *** bkopilov has joined #openstack-swift | 20:35 | |
| *** andrey-mp has quit IRC | 20:35 | |
| *** bkopilov has quit IRC | 20:41 | |
| *** albertom has joined #openstack-swift | 20:41 | |
| *** tsg has quit IRC | 20:49 | |
| *** bkopilov has joined #openstack-swift | 20:53 | |
| *** jrichli has quit IRC | 20:55 | |
| *** 7JTAAIXB0 has quit IRC | 21:01 | |
| *** lyrrad_ has joined #openstack-swift | 21:01 | |
| *** haomaiwang has joined #openstack-swift | 21:01 | |
| *** lyrrad has quit IRC | 21:03 | |
| *** lyrrad_ is now known as lyrrad | 21:03 | |
| *** pberis has joined #openstack-swift | 21:10 | |
| *** dustins has quit IRC | 21:17 | |
| *** jlhinson has quit IRC | 21:18 | |
| openstackgerrit | Matt Riedemann proposed openstack/python-swiftclient: Update from global requirements https://review.openstack.org/215786 | 21:22 |
| peluse | clayg, you still there? | 21:26 |
| notmyname | peluse: he's not been at his desk in a while | 21:42 |
| peluse | notmyname, OK thanks. Have a cont reconciler question that's no big deal - can wait | 21:45 |
| *** abhirc has joined #openstack-swift | 21:45 | |
| *** ccavanna has quit IRC | 21:48 | |
| *** ccavanna has joined #openstack-swift | 21:48 | |
| *** zaitcev has joined #openstack-swift | 21:50 | |
| *** ChanServ sets mode: +v zaitcev | 21:50 | |
| *** ccavanna_ has joined #openstack-swift | 21:50 | |
| *** ccavanna has quit IRC | 21:50 | |
| *** sungju has joined #openstack-swift | 21:53 | |
| *** sungju has quit IRC | 21:53 | |
| *** hrou has quit IRC | 21:54 | |
| *** ccavanna_ has quit IRC | 21:54 | |
| *** zhill has quit IRC | 21:56 | |
| *** haomaiwang has quit IRC | 22:01 | |
| *** haomaiwang has joined #openstack-swift | 22:02 | |
| *** pberis has quit IRC | 22:03 | |
| *** tsg has joined #openstack-swift | 22:05 | |
| *** chsc has quit IRC | 22:13 | |
| *** jkugel has left #openstack-swift | 22:14 | |
| *** minwoob has quit IRC | 22:30 | |
| *** tsg has quit IRC | 22:46 | |
| *** alejandrito has joined #openstack-swift | 22:46 | |
| *** lyrrad has quit IRC | 22:49 | |
| *** lyrrad has joined #openstack-swift | 22:51 | |
| mattoliverau | Back in Oz \o/ | 22:53 |
| notmyname | yay | 22:53 |
| *** haomaiwang has quit IRC | 23:01 | |
| *** haomaiwang has joined #openstack-swift | 23:02 | |
| clayg | peluse: nope | 23:04 |
| *** zhill has joined #openstack-swift | 23:08 | |
| *** abhirc has quit IRC | 23:12 | |
| *** tsg has joined #openstack-swift | 23:13 | |
| peluse | :) | 23:16 |
| peluse | outta here... later | 23:16 |
| *** abhirc has joined #openstack-swift | 23:24 | |
| *** lcurtis has quit IRC | 23:24 | |
| *** nadeem has quit IRC | 23:37 | |
| openstackgerrit | Tony Breeds proposed openstack/python-swiftclient: Update *requirements.txt from global-requirements https://review.openstack.org/215837 | 23:37 |
| *** tsg has quit IRC | 23:39 | |
| timburke | tdasilva: so i've been playing around with the idea of having delete markers in the versions (rather than versioned) container. how would we distinguish between PUT -> DELETE-with-marker (ie, original and marker in versions, nothing in versioned) and PUT -> DELETE-with-marker -> PUT (original and marker in versions, updated copy in versioned)? | 23:45 |
| timburke | we now need to know two things at once: is there an object in the versioned container? and is the version that i'm trying to restore a delete marker? we can get either one fairly easily (HEAD the object in the versioned container or peek at the COPY headers), but i think both at once will be hard. | 23:45 |
| timburke | if we HEAD before trying to do any COPYs, that info may be stale by the time we get a successful COPY. notmyname suggested If-(None-)Match: *, but those are no good; for a COPY, If-None-Match: * applies to the source (as well as the dest?), while If-Match: * isn't respected on PUTs (so it's meaningless for COPYs). | 23:45 |
| timburke | even if they worked, though, we'd be left needing to make two requests in some cases, which opens this up to race conditions even worse than it already is | 23:45 |
| *** chlong has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!