| mattoliverau | gmmaha: pudb not working is fine | 00:00 |
|---|---|---|
| gmmaha | mattoliverau: :D it helps dumb ppl like me a lot | 00:00 |
| *** lcurtis has quit IRC | 00:01 | |
| mattoliverau | lol, just use, pudb works if you use the 'python -m testtools.run discover --load-list my-list' technique, and that works good enough :( | 00:01 |
| mattoliverau | s/:(/:) | 00:01 |
| ho | good morning! | 00:02 |
| notmyname | both normal and in-process runs completed on my saio (in-process was faster). both successful | 00:02 |
| notmyname | that's a first | 00:02 |
| mattoliverau | vinsh: The subrequests *LO users to grab the segments don't include the referrer in the request, so adding the *.example.com referrer to the segments container doesn't work. Making it .r:* is working cause it accepts everything. | 00:04 |
| *** dmorita has quit IRC | 00:04 | |
| mattoliverau | notmyname: I know, it scares me too. Looks like logging gives a performance hit | 00:05 |
| *** dmorita has joined #openstack-swift | 00:05 | |
| *** nadeem has quit IRC | 00:06 | |
| gmmaha | mattoliverau: ohh didnt know that.. thanks :D | 00:06 |
| notmyname | gmmaha: where is SWIFT_TEST_DEBUG_LOGS set? | 00:07 |
| notmyname | gmmaha: ah, I see. defaults to off (undefined == off) | 00:08 |
| notmyname | only set it to turn it on | 00:08 |
| notmyname | nm | 00:08 |
| gmmaha | notmyname: yeah :) | 00:08 |
| vinsh | mattoliverau: Where does that leave us then? :) | 00:08 |
| vinsh | Don't want the segments container to have to be public... | 00:09 |
| mattoliverau | notmyname: so *LO | 00:10 |
| vinsh | Here is a summary I sent to the operators list today: http://lists.openstack.org/pipermail/openstack-operators/2015-December/009198.html | 00:10 |
| mattoliverau | notmyname: so *LO's and ACLs, do segment containers have to be global? We can't passing referer along so can't lock em down. | 00:11 |
| mattoliverau | wow my english is awesome. | 00:11 |
| notmyname | mattoliverau: I left my +2 on the testr patch (the second +2). if you like it, can you also add a +2? | 00:11 |
| notmyname | mattoliverau: I'll read it upside down and it will make sense | 00:11 |
| mattoliverau | notmyname: lol, thanks | 00:12 |
| notmyname | mattoliverau: or, some would say you don't speak english anyway ;-) | 00:12 |
| notmyname | mattoliverau: this is where acoles_ comes in to correct us both, right? | 00:12 |
| mattoliverau | I am one of them :P | 00:12 |
| mattoliverau | yup, exactly | 00:12 |
| notmyname | mattoliverau: do you think we should get acoles_ to also +2 the testr patch? or should you +A if you like it? | 00:12 |
| notmyname | hurricanerix: did you see the updates from gmmaha on patch 214206? | 00:13 |
| patchbot | notmyname: https://review.openstack.org/#/c/214206/ - Modify functional tests to use ostestr/testr | 00:13 |
| *** Jeffrey4l has joined #openstack-swift | 00:13 | |
| mattoliverau | notmyname: I'd like to +A, but I might have looked and been too involved in it now | 00:13 |
| notmyname | mattoliverau: vinsh: so yeah. *LO and permissions on the segments containers | 00:13 |
| vinsh | Indeed, what's the deal with those? | 00:13 |
| notmyname | mattoliverau: meh. we're all involved. I just want to make sure we're all ok with the way tests are run now | 00:14 |
| notmyname | vinsh: basically, you've discovered how they work. reading your ops ML post now | 00:14 |
| mattoliverau | notmyname: re: *LOs I wonder if we should add referer headers when making sub requests, so the egment containers can be locked down.. I can write a patch for that.. not that it will solve vinsh's question now :P | 00:15 |
| notmyname | mattoliverau: ah! it looks like there might be a bug in swift where the *LO middleware doesn't pass on the referrer (or the wsgi req.acl) for the subrequests | 00:15 |
| vinsh | When we were in the joint swift-dev/swift-operators meeting in that smaller room in tokyo. Ops were asked "whats tough about swift" I faltered with an answer.. but meant to say.. crazy corner cases like this that users come up with :) | 00:16 |
| notmyname | mattoliverau: are we passing on the wsgi.authorize function? probably, since otherwise we have an info leak (authorized for DLO would allow access to everything | 00:16 |
| mattoliverau | notmyname: I pretty sure we are.. I'll raise a bug, then investigate it. I've done enough debugging to know what referers isn't passed :) | 00:17 |
| notmyname | mattoliverau: cool, thanks | 00:18 |
| vinsh | mattoliverau: cc/add me please "adam.vinsh" | 00:18 |
| vinsh | Thanks for considering this. | 00:18 |
| notmyname | vinsh: is that your Launchpad id? | 00:18 |
| vinsh | oops id is "adam-vinsh" | 00:19 |
| mattoliverau | notmyname: just wanted a confirmation that I wasn't going crazy and there wasn't an obvious reason a segment container needed to be global :P | 00:19 |
| vinsh | ditto | 00:19 |
| notmyname | mattoliverau: no, it should require the same permissions as the manifest | 00:19 |
| mattoliverau | that's what I thought :) | 00:22 |
| mattoliverau | notmyname: good news, I have found a replacement Aussie to send in my place to the mid-cycle, alex will go. Which means you need to make sure she is sarcastic and gives people a hard time in my wake :P | 00:26 |
| notmyname | oh, cool! | 00:26 |
| asettle | o/ hellooooo | 00:26 |
| notmyname | asettle: we'll be glad to see you! | 00:26 |
| asettle | I hope so! I was just going to see if we have set dates? :) | 00:27 |
| mattoliverau | also means we can lock asettle and joel in a room to get swiftclient docs done :P | 00:27 |
| asettle | Yeah, lucky Joel :P I'm going to email him nowish :) | 00:27 |
| notmyname | timburke should likely be there too | 00:27 |
| notmyname | asettle: cc me please :-) | 00:27 |
| mattoliverau | great! | 00:27 |
| asettle | notmyname will do :) | 00:27 |
| mattoliverau | but your all not allowed to have too much fun :P | 00:28 |
| notmyname | asettle: not that joe's doc patch landed, I'm going to see how I can get my outline proposed and landed too | 00:28 |
| asettle | Okay great :) looking forward to it. | 00:28 |
| notmyname | mattoliverau: you'll be having fun too. (look up the work meconium) | 00:29 |
| notmyname | *word | 00:29 |
| asettle | WOW WHY WOULD YOU DO THAT | 00:29 |
| mattoliverau | notmyname: lol.. thanks :P | 00:29 |
| notmyname | it's not that bad. just part of having kids ;-) | 00:30 |
| mattoliverau | notmyname: the baby is due around then, if I went to midcycle it'll come while away, now that I'm not it'll be very late.. cause that's how the world works :P | 00:30 |
| notmyname | mattoliverau: then it will have the same birthday as me ;-) | 00:31 |
| asettle | Haha yeah | 00:31 |
| mattoliverau | notmyname: then notmyname will have to be here middlename | 00:32 |
| mattoliverau | s/here/her/ | 00:32 |
| *** david-lyle has joined #openstack-swift | 00:36 | |
| *** david-lyle_ has quit IRC | 00:36 | |
| timburke | mattoliverau: don't listen to notmyname -- those really weren't so bad | 00:48 |
| timburke | what's bad is later, once she's on solids, and then she gets sick | 00:49 |
| openstackgerrit | Merged openstack/python-swiftclient: Test 'string' behaviour of get_object https://review.openstack.org/258018 | 00:49 |
| mattoliverau | timburke: lol, thanks.. I'm going to pretend I didn't read that #ignoranceisbliss | 00:49 |
| timburke | totally worth it :) | 00:50 |
| *** Jeffrey4l has quit IRC | 00:51 | |
| * mattoliverau is moving again, another doctors appointment, brb | 00:55 | |
| kota_ | good morning | 00:55 |
| *** pgbridge has quit IRC | 00:56 | |
| *** zhill has quit IRC | 00:59 | |
| mattoliverau | kota_: morning | 01:02 |
| *** dmorita has quit IRC | 01:07 | |
| *** ho_ has joined #openstack-swift | 01:10 | |
| *** dmorita has joined #openstack-swift | 01:11 | |
| *** ho has quit IRC | 01:12 | |
| *** chlong has joined #openstack-swift | 01:21 | |
| kota_ | mattoliverau: o/ | 01:24 |
| *** zhill has joined #openstack-swift | 01:36 | |
| *** zhill has quit IRC | 01:40 | |
| *** klrmn has quit IRC | 01:42 | |
| *** lyrrad has quit IRC | 01:42 | |
| *** rvasilets__ has quit IRC | 01:45 | |
| *** ho has joined #openstack-swift | 01:47 | |
| *** ho_ has quit IRC | 01:48 | |
| janonymous_ | mattoliverau , kota_ : morning :) | 01:52 |
| mattoliverau | janonymous_: o/ | 01:52 |
| janonymous_ | i see a lot of work on testr patch in mail, great work mattoliverau, gmmaha and others. | 01:55 |
| mattoliverau | janonymous_: thanks, yeah, that took alot longer then expected, can now finally get back to other reviews :) | 01:55 |
| *** klrmn has joined #openstack-swift | 01:55 | |
| *** km has quit IRC | 01:56 | |
| *** km has joined #openstack-swift | 01:56 | |
| *** zhill has joined #openstack-swift | 01:57 | |
| janonymous_ | janonymous_ : yeah i was in sync in initial commits but many changes from then till now. | 01:57 |
| janonymous_ | *matoliverau | 01:57 |
| janonymous_ | **mattoliverau | 01:57 |
| *** noark9 has joined #openstack-swift | 01:59 | |
| notmyname | jrichli: wbhuber: FYI http://openstacktexas.com | 02:01 |
| *** zhill has quit IRC | 02:01 | |
| notmyname | dfg: redbo: ^ | 02:01 |
| kota_ | janonymous: ! | 02:02 |
| kota_ | I'm now looking at your paste for Reconceiler unit test. | 02:03 |
| *** klrmn has quit IRC | 02:04 | |
| janonymous_ | kota_:hey, thanks | 02:06 |
| kota_ | janonymous: i can see that with direct nosetests so i could look at the reason | 02:06 |
| janonymous_ | kota_: are you getting failure too | 02:07 |
| kota_ | janonymous: right now, "tox -e py27" doesn't fail but "nosetests test/unit/container/test_reconciler.py" fails | 02:07 |
| janonymous_ | yes | 02:07 |
| kota_ | so probably there is a difference between runner environments. | 02:08 |
| janonymous_ | kota_:yes maybe. | 02:08 |
| kota_ | i'm thinking that comes from... time zone? | 02:08 |
| kota_ | not sure | 02:08 |
| kota_ | that is because non-US people seeing it | 02:09 |
| janonymous_ | try tox -e py27 test.unit.container.test_reconciler | 02:09 |
| janonymous_ | that fails too | 02:09 |
| kota_ | oh, really | 02:09 |
| kota_ | will try | 02:09 |
| kota_ | done and looks like reproduced | 02:11 |
| janonymous_ | :) | 02:11 |
| kota_ | but to be able to reproduce is a way to fix it :-) | 02:11 |
| janonymous_ | hehee yes | 02:12 |
| * janonymous_ travelling to work | 02:12 | |
| janonymous_ | kota_ :will be back in 30 mins :) | 02:13 |
| asettle | notmyname you still awake? | 02:14 |
| *** klrmn has joined #openstack-swift | 02:15 | |
| *** janonymous_ has quit IRC | 02:17 | |
| kota_ | janonymous: fyi, it worked well in environ running on UTC | 02:19 |
| Zyric | Hi all. I've installed Slogging and left it going for a few hours today but the only logs I have are access ones. Looking at the logs all the account/container stats have 404'd for some reason. Thoughts? | 02:34 |
| *** Jeffrey4l has joined #openstack-swift | 02:37 | |
| *** mac_ified has joined #openstack-swift | 02:38 | |
| *** zhill has joined #openstack-swift | 02:40 | |
| *** gyee has quit IRC | 02:41 | |
| *** zhill has quit IRC | 02:45 | |
| *** sanchitmalhotra has joined #openstack-swift | 02:49 | |
| kota_ | janonymous: fixed, will push patch soon. | 02:49 |
| janonymous | kota_: great | 02:58 |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Fix reconciler test to calc lastmodified as UTC https://review.openstack.org/258233 | 02:59 |
| kota_ | done | 02:59 |
| kota_ | janonymous: please try it :) | 02:59 |
| janonymous | kota_: i was just fixing ut issues for testr patch , i got these errors then :) | 03:00 |
| janonymous | kota_: i saw the bug with fix . Yes i just tested it , and it passes . yay :) | 03:03 |
| *** zhill has joined #openstack-swift | 03:21 | |
| *** venkat has joined #openstack-swift | 03:25 | |
| kota_ | janonymous: yey :) | 03:25 |
| *** zhill has quit IRC | 03:26 | |
| kota_ | Zyric: did you set up stats-logger and log-uploader? | 03:27 |
| kota_ | they are tools to collect account/container stats | 03:28 |
| Zyric | kota_: I suppose I have, I followed the guide at https://github.com/notmyname/slogging/blob/master/doc/source/overview_stats.rst | 03:29 |
| Zyric | I have one of the storage.error files (number four I think) paste-binned here: http://paste.openstack.org/show/482035/ | 03:30 |
| kota_ | Ah, it looks not a slogging error but just Swift, right? | 03:32 |
| kota_ | Have you create AUTH_test account for your Swift? | 03:33 |
| mattoliverau | sigh, been playing with a patch to fix referers on segment containers.. easy to fix, but it wasn't working and debugged the hell out of it.. just to find I mistyped the referrer (.r <url>) on the segments container. #FAIL | 03:35 |
| mattoliverau | been one of those afternoons | 03:35 |
| Zyric | kota_: Auth_test is the default test account, I believe | 03:36 |
| *** nadeem has joined #openstack-swift | 03:37 | |
| kota_ | Zyric: Did the test work well? | 03:37 |
| Zyric | kota_: When I enter a get command on the account it appears to work | 03:38 |
| kota_ | Curious, swift logged as the account was not there but worked well.... | 03:39 |
| Zyric | kota_: Yes, I can GET /v1/AUTH_test/ using curl -v -H 'X-Auth-Token: <authToken>' http://127.0.0.1:8080/v1/AUTH_test/ and it doesn't 404 me | 03:40 |
| *** dmorita has quit IRC | 03:46 | |
| *** dmorita has joined #openstack-swift | 03:47 | |
| kota_ | sorry not sure. but what i can suggest is that perhaps you need to make sure the swift setup configuration and find it to trace correct route (from proxy to account server) | 03:50 |
| Zyric | kota_: Hm alright, so check the swift configuration files for information? Like swift.conf and /etc/swift/object-server.conf and such? | 03:53 |
| kota_ | yup and espically /etc/swift/account.ring.gz | 03:54 |
| kota_ | swift-get-nodes might help you | 03:55 |
| kota_ | that's a binary command to find the node devices from your supeficied path and ring information. | 03:56 |
| Zyric | Nice, thanks for the information kota_ I'll look into those | 03:57 |
| Zyric | Not sure how Slogging is getting confused with pathing considering the only locational data I set is the account name for uploading logging data. | 03:58 |
| *** km_ has joined #openstack-swift | 04:02 | |
| *** km has quit IRC | 04:04 | |
| *** kei_yama has quit IRC | 04:04 | |
| *** kei_yama has joined #openstack-swift | 04:13 | |
| venkat | Hi all, | 04:13 |
| venkat | One doubt : md5 is not collision resistant. | 04:13 |
| venkat | I am taking examples for creating containers. | 04:14 |
| venkat | How can we ensure that every new PUT request to container, creates a new database file with | 04:15 |
| venkat | hexdigest as filename | 04:15 |
| *** asettle is now known as asettle-afk | 04:19 | |
| *** proteusguy_ has quit IRC | 04:20 | |
| notmyname | asettle-afk: I'm here | 04:26 |
| venkat | Can someone can answer the above question? | 04:27 |
| notmyname | venkat: all hash functions have collisions (by definition). the question is what is the chance of a hash collision given the inputs | 04:29 |
| notmyname | venkat: and while md5 is used for the placement (and the resulting filenames), the probability of collision (especially when looking at eg 2nd preimage attacks) has been seen as very low | 04:31 |
| notmyname | meaning, sure, it could be changed, and probably will at some point, but it's not something that is causing issues such that those running swift are prioritizing that change (including those with billions of objects) | 04:31 |
| notmyname | venkat: and I'm not trying to dismiss the concerns you have about md5! | 04:32 |
| venkat | notmyname: Ok, Thanks, sha-1 is better than md5 right? | 04:33 |
| *** proteusguy_ has joined #openstack-swift | 04:33 | |
| notmyname | venkat: yes, but wise counsel would say sha-1 shouldn't be used for security either | 04:33 |
| notmyname | venkat: for some specifics on the implementation, here's what is hashed (with md5) in swift for placement and on-disk filename | 04:34 |
| venkat | notmyname: Ok, Thnak you | 04:34 |
| venkat | *Thank | 04:35 |
| notmyname | prefix + / + account + / + container + / + object + / + suffix | 04:35 |
| notmyname | you (the user) control account, container, and object strings | 04:35 |
| notmyname | and the total of those three (plus the slashes between them) needs to be less than 1536 bytes, after utf8 encoding and url quoting | 04:36 |
| notmyname | 1536 = 256 + 256 + 1024 | 04:37 |
| notmyname | 256 being the default max size for account and container | 04:37 |
| notmyname | and 1024 being the default max size for object name | 04:38 |
| notmyname | oh, and while you sortof maybe in some situations might have control over the object name, you really don't have much control over it. unless you own the cluster, you have to use what you're given for the account part from the cluster | 04:38 |
| notmyname | so really, you the user only control the container and object part | 04:39 |
| notmyname | so yeah, 256 + 1024 utf8/quoted bytes is much bigger than the 2**128 bits in md5 (so by definition there are collisions there) | 04:40 |
| venkat | notmyname: Thanks, this information is very helpful.. | 04:41 |
| notmyname | venkat: questions for you, now :-) | 04:42 |
| notmyname | venkat: why do you ask? are you being asked about it? | 04:42 |
| notmyname | venkat: do you have a swift cluster where this is an issue? | 04:42 |
| venkat | I am going through the code, I understood how url's are mapped to hexdigest. | 04:43 |
| venkat | later I explored md5 | 04:43 |
| venkat | In the wiki, MD5 is not collision resistant. | 04:44 |
| *** badari has quit IRC | 04:44 | |
| notmyname | right | 04:44 |
| venkat | in the future MD5 should no longer be implemented...where a collision-resistant hash function is required. | 04:44 |
| venkat | above line is from wiki | 04:45 |
| notmyname | what wiki? | 04:45 |
| venkat | https://en.wikipedia.org/wiki/MD5#Collision_vulnerabilities | 04:45 |
| notmyname | venkat: so why are you perusing the swift code? ;-) | 04:46 |
| notmyname | Zyric: did you get your questions answered? | 04:48 |
| venkat | I am learning swift code, just I am interested in open-source projects | 04:49 |
| *** trifon has joined #openstack-swift | 04:49 | |
| notmyname | venkat: awesome! I'd love to help you get involved | 04:49 |
| notmyname | we've got lots of stuff to do, and lots of ways to help out | 04:50 |
| venkat | notmyname: Ok, Thank you, sure i will help | 04:50 |
| notmyname | venkat: so you're already doing 2 things which are super important for anyone contributing to swift. (1) asking questions and being in this IRC channel and (2) looking at the swift code to see what it does | 04:51 |
| venkat | ok | 04:52 |
| notmyname | venkat: if you look in the channel topic, there's a like to a review dashboard. it's a good place to see what changes people are suggesting for swift, and it's our code review tool | 04:52 |
| notmyname | and our bug tracker is at https://bugs.launchpad.net/swift (for swift) and https://bugs.launchpad.net/python-swiftclient (for the official SDK/CLI) | 04:53 |
| notmyname | venkat: are you familiar with openstack projects? have you contributed to any others? | 04:53 |
| notmyname | on the launchpad bug trackers, the "wishlist" items are things we think would be cool to add to the project | 04:53 |
| venkat | notmyname: yes, I worked on cloudkitty-billing project, I submitted 1 patch to nova | 04:53 |
| notmyname | oh, cool | 04:54 |
| notmyname | so what brings you to swift? have you deployed it? used it? just really really like storage? | 04:54 |
| venkat | I tried only SAIO and devstack | 04:54 |
| notmyname | venkat: oh cool! now you've doing something else that's good: you've set up the SAIO :-) | 04:55 |
| venkat | Ok, I have the SAIO setup | 04:56 |
| *** badari has joined #openstack-swift | 04:56 | |
| notmyname | venkat: any chance you work for hgst? | 04:57 |
| venkat | yes, how you know? | 04:58 |
| notmyname | lol. /me is sneaky | 04:59 |
| notmyname | you're connected to IRC from an IP address provisioned to hgst | 05:00 |
| venkat | Ok, you are too smart | 05:00 |
| notmyname | venkat: so anyway... is there something in swift that you're interested in? something way I can help you get involved? | 05:01 |
| notmyname | venkat: so as to have parity in the information sharing, I'm http://not.mn | 05:01 |
| *** kei_yama has quit IRC | 05:02 | |
| venkat | Yes, I want to know how db's are replicated across the conatiner servers | 05:02 |
| *** km has joined #openstack-swift | 05:02 | |
| venkat | when the background replication occurs? | 05:02 |
| notmyname | venkat: http://docs.openstack.org/developer/swift/overview_replication.html#db-replication | 05:02 |
| notmyname | yes | 05:02 |
| *** km_ has quit IRC | 05:03 | |
| *** ppai has joined #openstack-swift | 05:03 | |
| venkat | notmyname: Thank you, I will go through the above link | 05:08 |
| venkat | my github link : https://github.com/venkatesh551 | 05:08 |
| *** dmorita has quit IRC | 05:08 | |
| notmyname | venkat: please feel free to drop by and ask questions as you have them. please feel free to leave reviews on patches, and if you've got some code to share, please feel free to submit it! | 05:10 |
| *** dmorita has joined #openstack-swift | 05:11 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:11 | |
| *** dmorita has quit IRC | 05:15 | |
| *** yatin has joined #openstack-swift | 05:20 | |
| *** zhill has joined #openstack-swift | 05:20 | |
| *** badari has quit IRC | 05:24 | |
| *** kei_yama has joined #openstack-swift | 05:25 | |
| venkat | notmyname: sure, I am very much interested in open-source contribution. | 05:26 |
| *** trifon has quit IRC | 05:31 | |
| openstackgerrit | venkatesh proposed openstack/swift: remove unused parameters in the method https://review.openstack.org/243401 | 05:31 |
| notmyname | venkat: oh wait. is that you? ^ | 05:31 |
| venkat | notmyname: yes | 05:33 |
| *** links has joined #openstack-swift | 05:34 | |
| *** klrmn has quit IRC | 05:34 | |
| *** SkyRocknRoll has quit IRC | 05:37 | |
| *** zhill has quit IRC | 05:40 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:45 | |
| *** lpabon has quit IRC | 05:45 | |
| *** SkyRocknRoll has quit IRC | 06:12 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:26 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift: Pass HTTP_REFERER down to subrequests https://review.openstack.org/258280 | 06:31 |
| *** changbl has joined #openstack-swift | 06:42 | |
| *** asettle-afk is now known as asettle | 06:44 | |
| asettle | oof notmyname are you still around? sorry! Ran off to the docs | 06:44 |
| *** nakagawamsa has joined #openstack-swift | 06:51 | |
| *** nakagawamsa has quit IRC | 06:53 | |
| *** nakagawamsa has joined #openstack-swift | 06:58 | |
| *** bsdkurt has quit IRC | 07:00 | |
| *** rcernin has joined #openstack-swift | 07:01 | |
| *** bsdkurt has joined #openstack-swift | 07:02 | |
| *** km_ has joined #openstack-swift | 07:08 | |
| *** km has quit IRC | 07:08 | |
| *** zaitcev has quit IRC | 07:08 | |
| *** nadeem has quit IRC | 07:09 | |
| *** rcernin_ has joined #openstack-swift | 07:12 | |
| *** wanghua has joined #openstack-swift | 07:12 | |
| *** rcernin has quit IRC | 07:13 | |
| wanghua | Hi all, does swift have the same policy mechanism which can be configured in policy.json as other openstack projects? | 07:14 |
| *** bsdkurt has quit IRC | 07:15 | |
| *** bsdkurt has joined #openstack-swift | 07:17 | |
| *** rcernin_ has quit IRC | 07:21 | |
| *** rcernin has joined #openstack-swift | 07:22 | |
| openstackgerrit | Merged openstack/swift: Modify functional tests to use ostestr/testr https://review.openstack.org/214206 | 07:25 |
| *** Jeffrey4l has quit IRC | 07:31 | |
| *** ntt has joined #openstack-swift | 07:35 | |
| *** Jeffrey4l has joined #openstack-swift | 07:44 | |
| *** arnox has joined #openstack-swift | 07:58 | |
| janonymous | kota_ : just one ques, why was it passing when all ut were run, why doesn't it fail ? | 08:00 |
| *** torgomatic has quit IRC | 08:01 | |
| kota_ | janonymous: sorry, idk | 08:02 |
| kota_ | right now | 08:02 |
| kota_ | perhaps some codes would exist to adjust default timezone to utc? | 08:04 |
| kota_ | not sure. | 08:04 |
| *** Jeffrey4l has quit IRC | 08:08 | |
| *** chlong has quit IRC | 08:17 | |
| *** Jeffrey4l has joined #openstack-swift | 08:21 | |
| *** hseipp has joined #openstack-swift | 08:24 | |
| *** Jeffrey4l has quit IRC | 08:26 | |
| *** ppai has quit IRC | 08:36 | |
| *** rvasilets__ has joined #openstack-swift | 08:36 | |
| *** Jeffrey4l has joined #openstack-swift | 08:38 | |
| *** zhill has joined #openstack-swift | 08:39 | |
| *** jordanP has joined #openstack-swift | 08:44 | |
| *** zhill has quit IRC | 08:45 | |
| *** ppai has joined #openstack-swift | 08:50 | |
| *** rcernin has quit IRC | 08:50 | |
| *** Jeffrey4l has quit IRC | 08:55 | |
| *** mac_ified has quit IRC | 08:58 | |
| *** Jeffrey4l has joined #openstack-swift | 08:59 | |
| kota_ | k, heading for home. | 09:02 |
| kota_ | bye | 09:02 |
| *** venkat has quit IRC | 09:05 | |
| cschwede | vinsh: thx for the info on my puppet-swift patch - i just rebased that one. let me know if this looks ok to you in the review | 09:05 |
| *** rcernin has joined #openstack-swift | 09:08 | |
| *** Jeffrey4l has quit IRC | 09:11 | |
| *** jistr has joined #openstack-swift | 09:13 | |
| janonymous | kota_: thanks :) | 09:15 |
| *** venkat has joined #openstack-swift | 09:19 | |
| *** noark9 has quit IRC | 09:19 | |
| *** jmccarthy1 has joined #openstack-swift | 09:39 | |
| *** jmccarthy1 has quit IRC | 09:39 | |
| *** jmccarthy1 has joined #openstack-swift | 09:40 | |
| *** joeljwright has joined #openstack-swift | 09:40 | |
| *** ChanServ sets mode: +v joeljwright | 09:40 | |
| *** jmccarthy1 has quit IRC | 09:40 | |
| *** venkat has quit IRC | 09:46 | |
| *** acoles_ is now known as acoles | 09:47 | |
| *** venkat has joined #openstack-swift | 09:58 | |
| *** jmccarthy has quit IRC | 09:59 | |
| *** jmccarthy has joined #openstack-swift | 10:00 | |
| *** Jeffrey4l has joined #openstack-swift | 10:04 | |
| eranrom | acoles here? | 10:04 |
| *** rledisez has joined #openstack-swift | 10:07 | |
| acoles | eranrom: yes, hi | 10:08 |
| acoles | wow long scrollback today | 10:09 |
| acoles | mattoliverau: i thought i read that notmyname was going to become middleware, but then saw middlename :) | 10:09 |
| acoles | mattoliverau: which got me thinking that you could use the swift canonical pipeline as middlenames... | 10:10 |
| acoles | ok so the func test patch landed, time to go play with it... | 10:11 |
| mattoliverau | acoles: lol genius! | 10:13 |
| acoles | mattoliverau: hmm, "genius" is a good name too | 10:13 |
| *** Jeffrey4l has quit IRC | 10:15 | |
| *** Jeffrey4l has joined #openstack-swift | 10:15 | |
| mattoliverau | So is Alistair ;) | 10:16 |
| acoles | no no no | 10:17 |
| openstackgerrit | Hisashi Osanai proposed openstack/swift: Fix duplication for headers in Access-Control-Expose-Headers https://review.openstack.org/258392 | 10:17 |
| acoles | so prone to mispeling | 10:17 |
| acoles | (sic) | 10:17 |
| *** jmccarthy1 has joined #openstack-swift | 10:20 | |
| tdasilva | acoles: isn't there a famous "middle..." something or other in the UK? ;) | 10:25 |
| acoles | tdasilva: middleton? | 10:26 |
| tdasilva | if you adtop Acoles Genius Middleware you could almost pass as a royalty | 10:26 |
| tdasilva | swift royalty | 10:27 |
| tdasilva | duh, Alistair ..... | 10:27 |
| acoles | tdasilva: lol | 10:29 |
| acoles | tdasilva: mattoliverau : TypeError: unbound method test_bad_metadata2() must be called with TestAccount instance as first argument (got nothing instead) | 10:30 |
| *** aix has joined #openstack-swift | 10:30 | |
| acoles | :/ | 10:30 |
| acoles | got that using: tox -r -e func -- -n test.functional.test_account.TestAccount.test_bad_metadata2 | 10:30 |
| mattoliverau | What :( it worked on mine, John's and tdasilva's | 10:31 |
| acoles | mattoliverau: but that was *yesterday* :P | 10:31 |
| tdasilva | acoles: instead of -n, use -r for regex | 10:31 |
| mattoliverau | Lol | 10:31 |
| tdasilva | tox -e func -- -r test.functional.test_account.TestAccount.test_bad_metadata2 | 10:31 |
| acoles | tdasilva: yep that works, but in scrollback i thought mtreinish was saying -n should work | 10:33 |
| acoles | this works: tox -r -e func -- -n test.functional.test_account.TestAccount | 10:34 |
| acoles | so must be that -n is only happy with a class, not a test method | 10:34 |
| acoles | mattoliverau: did you see -n work with a method, or just class? | 10:35 |
| tdasilva | acoles: I never got -n to work with a method, only class | 10:35 |
| eranrom | acoles: so I am struggling with the so called redundant call to get_container to allow the changes to get commited. | 10:36 |
| mattoliverau | Yeah only class in the end for me to. But -r works to the class so that's good | 10:36 |
| eranrom | the code basically does: put_container (that is synced), put_object, get_container, and then run-once the sunc daemon | 10:37 |
| eranrom | without the get_container between the put_object and the run-once the broker just do not show the new object | 10:38 |
| eranrom | So I went and printed whatever gets merged as a result of get_info after each of the steps. | 10:38 |
| eranrom | I consistently get: | 10:39 |
| *** ho has quit IRC | 10:39 | |
| eranrom | After the put_object I see: Going to merge items: [{'object_count': '0', 'name': 'dest-container-414d74d3-4904-479d-a86f-4f473f1ff1ce', 'deleted': 0, 'bytes_used': '0', 'put_timestamp': '1450259790.01652', 'storage_policy_index': '0', 'delete_timestamp': '0'}] | 10:40 |
| eranrom | After the get_container I see: Going to merge items: [{'name': 'object-50567314-f745-41fa-bfdc-77442a98c7dd', 'deleted': 0, 'created_at': '1450259790.13798', 'storage_policy_index': 1, 'etag': '70b68ae721a61941a1a62724dde5d5e4', 'content_type': 'application/octet-stream', 'size': 9}] | 10:40 |
| eranrom | the latter is not seen without the get_container. FWIW and not surprisingly head_container has the same effect | 10:41 |
| acoles | tdasilva: mattoliverau this is fun http://paste.openstack.org/show/482053/ | 10:41 |
| acoles | gmmaha: ^^ | 10:41 |
| eranrom | Let me dig dipper... | 10:42 |
| tdasilva | acoles: func: commands succeeded :/ | 10:43 |
| acoles | tdasilva: yeah i love that positivity :) | 10:44 |
| acoles | eranrom: ok, so you only see the object update flushed to the db during the get_container, but not when sync calls get_info? | 10:45 |
| eranrom | acoles: correct | 10:46 |
| eranrom | Can object upload return with async pending from all object nodes? | 10:46 |
| acoles | eranrom: yes - if you mean can the object PUT return before any container has been updated | 10:47 |
| eranrom | yes, in fact before any container replica has been updated. | 10:48 |
| acoles | eranrom: at least i believe that to be so, there was a recent change to how object->container updates are handled | 10:48 |
| acoles | but that just made them potentially more likely to be async | 10:48 |
| acoles | eranrom: so could it just be that the object update has not arrived at container server when you log the get_info call but has arrived when the get_container occurs? | 10:49 |
| eranrom | yes, I think that that's is what hapenning. so the call is redundant, and it just a matter of waiting. | 10:50 |
| eranrom | that is the call to get_container | 10:50 |
| eranrom | I will try | 10:51 |
| acoles | eranrom: your first "Going to merge items" log above is a container->account update, the second is an object->container update | 10:51 |
| eranrom | Right. So the first one is triggered by the creation of the container | 10:53 |
| acoles | tdasilva: i'll file a bug report for the --until-failure problem, looks like the testr.conf command may need working on. | 10:53 |
| eranrom | and the second is by the objecyt upload | 10:53 |
| tdasilva | acoles: ok, i was about to test it here on my end | 10:53 |
| tdasilva | acoles: sorry i forgot to test that, i hadn't used the flag before | 10:54 |
| tdasilva | it actually sounds very useful | 10:54 |
| acoles | tdasilva: yes, although what I will really miss is nosetests -x, which runs tests once stopping on first failure | 10:55 |
| tdasilva | acoles: oh, i thought that's what --until-failure did | 10:55 |
| acoles | tdasilva: well maybe, I don't know yet :), but the os-testr docs say it *repeats* the tests until failure | 10:56 |
| tdasilva | mmm...that sounds weird | 10:56 |
| *** Jeffrey4l has quit IRC | 10:58 | |
| *** aix has quit IRC | 10:59 | |
| *** aix has joined #openstack-swift | 11:00 | |
| acoles | tdasilva: bug 1526725 | 11:00 |
| openstack | bug 1526725 in OpenStack Object Storage (swift) "tox -e func -- --until-failures does not work" [Undecided,New] https://launchpad.net/bugs/1526725 | 11:00 |
| tdasilva | acoles: thanks | 11:04 |
| acoles | tdasilva: I can of course still use nosetests -x but not via tox, so its not so bad. | 11:06 |
| tdasilva | acoles: true true | 11:06 |
| tdasilva | I was about to test removing the env. variable to see if that clears up the error | 11:07 |
| acoles | tdasilva: maybe we could have tox -e func_with_nose :) like the good ol' days | 11:07 |
| tdasilva | ha! | 11:07 |
| tdasilva | tox -e func_old_school | 11:07 |
| acoles | i keep laughing at clayg's comment recently 'every release is a little bit better than the next' | 11:08 |
| acoles | tdasilva: thats even better | 11:08 |
| acoles | eranrom: the object server will wait 1 sec for container update to complete swift/obj/server.py:304 | 11:10 |
| acoles | eranrom: but the thread will wait 3 secs before timing out on the container server request and queueing the update for async re-try | 11:10 |
| acoles | kota_: thanks for re-review | 11:12 |
| acoles | kota_ is probably out partying at this time :) | 11:13 |
| kota_ | Not so a frequent party! | 11:14 |
| kota_ | You are welcome, acoles :) | 11:14 |
| *** joeljwright has quit IRC | 11:15 | |
| eranrom | acoles: Thanks. Apparently, the commit_puts_stale_ok call triggered from get_info from container sync never gets to aquire the lock over the pending flie. | 11:17 |
| eranrom | I will dig more into this. | 11:17 |
| eranrom | gotta go now | 11:17 |
| eranrom | Thanks! | 11:17 |
| tdasilva | mtreinish: are you aware of issues with --until-failure? We getting errors running like this: tox -e func -- --until-failure | 11:29 |
| tdasilva | looks like neutron also has similar issue: https://bugs.launchpad.net/os-testr/+bug/1467209 | 11:30 |
| openstack | Launchpad bug 1467209 in os-testr "--until-failure option does not work" [Undecided,New] | 11:30 |
| ntt | Hi, How the swift proxy point to the object/container and account server? I cannot find any reference in /etc/swift/proxy-server.conf? | 11:36 |
| tdasilva | ntt: that information comes from the rings | 11:37 |
| ntt | so, the proxy parse the ring file and obtains ip of object/container/account server? | 11:37 |
| tdasilva | the proxy gets that information from the Ring class. | 11:40 |
| tdasilva | ntt: checkout some of these links: http://docs.openstack.org/developer/swift/overview_architecture.html | 11:40 |
| tdasilva | https://swiftstack.com/openstack-swift | 11:40 |
| tdasilva | ntt: this is also helpful: http://www.florentflament.com/blog/openstack-swift-ring-made-understandable.html | 11:41 |
| ntt | tdasilva: thank you | 11:42 |
| *** rvasilets__ has quit IRC | 11:45 | |
| *** goodygum has joined #openstack-swift | 11:51 | |
| openstackgerrit | janonymous proposed openstack/swift: Fix reconciler test to calc lastmodified as UTC https://review.openstack.org/258233 | 11:53 |
| *** yatin has quit IRC | 11:55 | |
| *** km_ has quit IRC | 12:06 | |
| *** kei_yama has quit IRC | 12:10 | |
| *** CaioBren_ has joined #openstack-swift | 12:20 | |
| *** CaioBre__ has joined #openstack-swift | 12:22 | |
| *** CaioBren_ has quit IRC | 12:22 | |
| *** CaioBrentano has quit IRC | 12:23 | |
| acoles | eranrom: i think i see the cause of needing the get_container call. the get_info only commits from pending file if pending file exists, and there is no pending file in the sync_containers sub_dir, which is where the sync daemon thinks the db is and therefore where it looks for the pending file. | 12:30 |
| *** joeljwright has joined #openstack-swift | 12:30 | |
| *** ChanServ sets mode: +v joeljwright | 12:30 | |
| acoles | eranrom: a very quick hack...http://paste.openstack.org/show/482058/ | 12:30 |
| acoles | eranrom: left comment on gerrit | 12:37 |
| kota_ | back | 12:38 |
| *** CaioBre__ has quit IRC | 12:47 | |
| eranrom | acoles: Thanks! Great catch! I think the right way to do this is to initiate the broker in container sync from the pointed dir rather then the symlink. I wil try that later and fix accordingly. | 12:52 |
| *** links has quit IRC | 12:57 | |
| *** rcernin has quit IRC | 12:58 | |
| *** lpabon has joined #openstack-swift | 13:00 | |
| *** CaioBrentano has joined #openstack-swift | 13:06 | |
| *** chlong has joined #openstack-swift | 13:10 | |
| *** rcernin has joined #openstack-swift | 13:13 | |
| *** Jeffrey4l has joined #openstack-swift | 13:16 | |
| *** ppai has quit IRC | 13:21 | |
| vinsh | cschwede: for the auditor block.. copy the replicator block above and just change replicator to auditor | 13:22 |
| vinsh | It should be the same form also using the service class. | 13:23 |
| *** jmccarthy1 has quit IRC | 13:24 | |
| cschwede | vinsh: thx - you mean in manifests/storage/generic.pp, right? | 13:25 |
| *** breitz has quit IRC | 13:32 | |
| *** zhill has joined #openstack-swift | 13:32 | |
| vinsh | Yep! | 13:32 |
| *** breitz has joined #openstack-swift | 13:32 | |
| *** zhill has quit IRC | 13:37 | |
| vinsh | cschwede: One more thing to.. in the spec file... copy the replicator block also.. will preserve the "ensure" enable/disable testing in the spec :) needs name passed in all that the same. | 13:37 |
| vinsh | Then, I say its ready to go :) I'm glad to see this merge soon... I'm using it as a starting point for my next review. I'm adding support for a dedicated replication network in it. | 13:38 |
| *** SkyRocknRoll has quit IRC | 13:43 | |
| kota_ | someone could have a time to review my too small 2 patches? | 13:45 |
| kota_ | patch 258233 and patch 255096 | 13:46 |
| patchbot | kota_: https://review.openstack.org/#/c/258233/ - Fix reconciler test to calc lastmodified as UTC | 13:46 |
| patchbot | kota_: https://review.openstack.org/#/c/255096/ - Sleep enough for trampoling | 13:46 |
| kota_ | one is 1 line change and the other is 2 lines :) | 13:47 |
| janonymous | both fixes UT for me :) | 13:47 |
| kota_ | janonymous: me too ;) | 13:48 |
| *** delatte has joined #openstack-swift | 13:48 | |
| *** venkat has quit IRC | 13:51 | |
| *** delattec has quit IRC | 13:52 | |
| *** lcurtis has joined #openstack-swift | 14:02 | |
| *** hseipp has quit IRC | 14:04 | |
| *** rcernin has quit IRC | 14:06 | |
| gmmaha | acoles: tdasilva: i didnt try that piece out. Let me check it out | 14:08 |
| *** eranrom has quit IRC | 14:08 | |
| acoles | kota_: hehe i fixed one of those a year ago in fast-post patch :) https://review.openstack.org/#/c/135380/19/test/unit/container/test_reconciler.py | 14:08 |
| *** delattec has joined #openstack-swift | 14:09 | |
| *** ekarlso has quit IRC | 14:10 | |
| kota_ | acoles: great! ...not landed yet! :/ | 14:10 |
| acoles | kota_: i think your patch will land before mine ;) | 14:11 |
| kota_ | acoles: lol | 14:13 |
| *** delatte has quit IRC | 14:13 | |
| *** changbl has quit IRC | 14:14 | |
| *** ctennis has quit IRC | 14:16 | |
| *** ctennis has joined #openstack-swift | 14:16 | |
| lcurtis | good morning all | 14:17 |
| lcurtis | does anyone know if it is possible to increase xfs inode cache size beyond 32gb? | 14:17 |
| gmmaha | acoles: tdasilva: merging the .testr.conf makes it go fine, but hitting this error now http://paste.openstack.org/show/482076/ | 14:18 |
| gmmaha | http://paste.openstack.org/show/482077/ | 14:19 |
| *** daemontool has joined #openstack-swift | 14:19 | |
| tdasilva | gmmaha: :/ it seems to be like there's an issue with testr | 14:21 |
| gmmaha | tdasilva: yeah.. seems like it. | 14:22 |
| gmmaha | wondering if there are more configs that we need/should set to make the --until-failure work | 14:23 |
| acoles | gmmaha: what did you mean by " merging the .testr.conf" ? | 14:25 |
| acoles | gmmaha: oic your second paste | 14:26 |
| gmmaha | acoles: yeah, not breaking up the line | 14:33 |
| acoles | gmmaha: iirc this http://paste.openstack.org/show/482076/ is what i saw when trying an earlier patchset | 14:33 |
| gmmaha | acoles: yeah, that error seems familiar. think its close to the same we are seeing when you run the test with -n and give it a function against a class | 14:35 |
| *** blmartin has joined #openstack-swift | 14:42 | |
| pdardeau | good morning | 14:44 |
| gmmaha | good morning pdardeau | 14:48 |
| wbhuber | notmyname: Openstack TX sounds great. I'm bookmarking it. | 14:49 |
| *** joeljwright has quit IRC | 14:49 | |
| pdardeau | gmmaha: Guten tag! | 14:49 |
| *** joeljwright has joined #openstack-swift | 14:51 | |
| *** badari has joined #openstack-swift | 14:51 | |
| *** ChanServ sets mode: +v joeljwright | 14:51 | |
| *** dustins|out is now known as dustins | 14:52 | |
| *** daemontool_ has joined #openstack-swift | 14:54 | |
| *** joeljwright has quit IRC | 14:55 | |
| *** daemontool has quit IRC | 14:56 | |
| *** joeljwright has joined #openstack-swift | 14:58 | |
| *** ChanServ sets mode: +v joeljwright | 14:58 | |
| *** daemontool_ has quit IRC | 14:59 | |
| acoles | gmmaha: got it! our functional test classes subclass unittest.TestCase, changing to unittest2.TestCase and --until-failures and -n both work (along with test_command being on single line) | 15:09 |
| acoles | tdasilva: mattoliverau ^^ | 15:09 |
| gmmaha | acoles: ohhh | 15:10 |
| acoles | gmmaha: I am just doing global replace to check complete test suite | 15:10 |
| gmmaha | no way i would have figured that out. :) thanks a ton! | 15:10 |
| tdasilva | neat | 15:10 |
| gmmaha | acoles: awesome | 15:10 |
| acoles | gmmaha: i'll push a patch and then everyone can try it out | 15:10 |
| gmmaha | absolutely | 15:11 |
| tdasilva | acoles: thanks! | 15:11 |
| *** jmccarthy1 has joined #openstack-swift | 15:11 | |
| *** jmccarthy1 has quit IRC | 15:14 | |
| gmmaha | acoles: the unittest2 page says its a backport of unittest features from py2.7 and onwards https://pypi.python.org/pypi/unittest2 | 15:16 |
| *** dmorita has joined #openstack-swift | 15:16 | |
| *** jmccarthy has quit IRC | 15:16 | |
| gmmaha | should we look into updating theunittest package also? | 15:16 |
| acoles | gmmaha: sorry, updating where? | 15:20 |
| gmmaha | acoles: ignore that comment.. i thought there were two seperate packages unittest and unittest2 | 15:21 |
| gmmaha | doesnt seem like it.. | 15:21 |
| acoles | argh, now our skipped tests show as fails! | 15:22 |
| *** tsg has joined #openstack-swift | 15:26 | |
| *** zaitcev has joined #openstack-swift | 15:29 | |
| *** ChanServ sets mode: +v zaitcev | 15:29 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix func test --until-failure and --no-discover options https://review.openstack.org/258578 | 15:35 |
| acoles | gmmaha: i included your fix to .testr.conf and made you co-author ^^ | 15:36 |
| gmmaha | acoles: thanks.. :) | 15:37 |
| * gmmaha goes to test the patch | 15:37 | |
| acoles | mtreinish: would you mind taking a look at this patch https://review.openstack.org/258578 ? It seems to work but I got there by trial and error rather than great understanding! | 15:38 |
| *** daemontool has joined #openstack-swift | 15:40 | |
| *** acoles is now known as acoles_ | 15:43 | |
| gmmaha | acoles_: are you also seeing a setup and teardown for each test with --until-failure | 15:47 |
| gmmaha | ? | 15:47 |
| mtreinish | acoles_: oh, cool. Yeah nose isn't really unittest compatible. We hit that same thing before when we did the initial tempest migration a few years ago | 15:48 |
| *** pgbridge has joined #openstack-swift | 15:59 | |
| *** acoles_ is now known as acoles | 15:59 | |
| acoles | mtreinish: thanks | 16:01 |
| acoles | gmmaha: yes, --until-failure seems to invoke the package setup/teardown for every test method, and with that comes a lot of noise | 16:01 |
| acoles | but at least it works! | 16:02 |
| gmmaha | acoles: ohh yeah.. just wanted to verify that the outcome is consistent | 16:04 |
| *** joeljwright has quit IRC | 16:04 | |
| *** joeljwright has joined #openstack-swift | 16:09 | |
| *** ChanServ sets mode: +v joeljwright | 16:09 | |
| *** joeljwright has quit IRC | 16:14 | |
| *** ekarlso has joined #openstack-swift | 16:17 | |
| *** janonymous_ has joined #openstack-swift | 16:19 | |
| *** daemontool__ has joined #openstack-swift | 16:21 | |
| -openstackstatus- NOTICE: Gerrit will be offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 16:21 | |
| *** daemontool has quit IRC | 16:21 | |
| *** janonymous_ has quit IRC | 16:22 | |
| *** janonymous_ has joined #openstack-swift | 16:23 | |
| *** garthb has joined #openstack-swift | 16:25 | |
| *** zhill has joined #openstack-swift | 16:30 | |
| *** minwoob has joined #openstack-swift | 16:33 | |
| *** alejandrito has joined #openstack-swift | 16:44 | |
| *** esker has joined #openstack-swift | 16:48 | |
| notmyname | good morning | 16:52 |
| notmyname | I'll be in and out today. tons of meetings (including the swift team meeting) | 16:52 |
| janonymous_ | notmyname: o/ | 16:54 |
| *** daemontool__ has quit IRC | 16:56 | |
| janonymous_ | I have added a patch for opinions in https://wiki.openstack.org/wiki/Meetings/Swift, as suggested by Ondrej in the comments. | 16:57 |
| janonymous_ | notmyname: But it gets pretty late for me to attend it, should i remove it from there or is it ok ? | 16:57 |
| notmyname | janonymous_: do you have a specific question that you want addressed? eg "shall we do X or Y in this case" | 16:58 |
| janonymous_ | Oh, not particularly , it about how should we proceed with config values in config files and code | 16:59 |
| janonymous_ | *it's | 16:59 |
| *** diazjf has joined #openstack-swift | 17:00 | |
| notmyname | janonymous_: ok. we've got a lot on the agenda already, so I think it would be good to keep a discussion in gerrit. I can leave the patch reference on the agenda, though, to ask for some reviews on it | 17:01 |
| notmyname | janonymous_: if you've got a specific question, perhaps asking it in gerrit would be a good async way for people to respond | 17:01 |
| *** lcurtis has quit IRC | 17:02 | |
| notmyname | janonymous_: also that means you don't have to be up ridiculously early in the morning :-) | 17:02 |
| janonymous_ | yes, but i was suggested to ask it in meeting if possible in gerrit :) | 17:02 |
| janonymous_ | *gerrit comment | 17:03 |
| *** daemontool__ has joined #openstack-swift | 17:03 | |
| *** arnox has quit IRC | 17:03 | |
| notmyname | seems the questions are from onovy | 17:03 |
| janonymous_ | yup | 17:04 |
| onovy | ? | 17:05 |
| janonymous_ | it would be good if someone could review and land some opinions on the same, then no need to land it in meeting i guess | 17:05 |
| janonymous_ | onovy: o/ | 17:05 |
| notmyname | janonymous_: agreed | 17:05 |
| janonymous_ | onovy: its about patch https://review.openstack.org/#/c/227855/ | 17:06 |
| -openstackstatus- NOTICE: Gerrit is offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 17:06 | |
| *** ChanServ changes topic to "Gerrit is offline for a software upgrade from 17:00 to 21:00 UTC. See: http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html" | 17:06 | |
| *** lcurtis has joined #openstack-swift | 17:06 | |
| onovy | it's long open review so i suggested to discuss it on meeting | 17:07 |
| onovy | my opinion is in gerrit | 17:07 |
| *** janonymous__ has joined #openstack-swift | 17:08 | |
| janonymous_ | onovy: i also agree with what you suggested | 17:09 |
| janonymous_ | onovy: So, i think i would make the changes as per that and some review comments would do the good | 17:10 |
| janonymous_ | onovy: what do you say.. ? | 17:10 |
| *** rledisez has quit IRC | 17:10 | |
| *** daemontool__ has quit IRC | 17:12 | |
| tdasilva | is gerrit down for anybody else? | 17:12 |
| diazjf | tdasilva -> http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 17:13 |
| notmyname | yup | 17:13 |
| notmyname | see the topic | 17:13 |
| *** lcurtis has quit IRC | 17:14 | |
| onovy | janonymous_, yep, that's good idea | 17:14 |
| notmyname | tdasilva: this would be a good time to peruse https://github.com/google/git-appraise ;-) | 17:14 |
| tdasilva | ha! thanks | 17:14 |
| tdasilva | i think I saw that on hacker news the other day | 17:15 |
| notmyname | yeah, that's where I saw it | 17:15 |
| notmyname | looked really interesting | 17:15 |
| notmyname | and I still think we should have something like gerrit that is an interface to it. just not gerrit as the system which must be online for anyone to do anything | 17:15 |
| tdasilva | maybe we can just do it here | 17:15 |
| tdasilva | acoles: your patch works, +2 :D | 17:15 |
| notmyname | heh | 17:16 |
| gmmaha | acoles: +1 from me on the patch. it works for me. Just cant put it in gerrit now :) | 17:16 |
| janonymous_ | Haha (Y) | 17:16 |
| tdasilva | acoles: I also tried --pdb and it now works with the specific test case, so great work, thanks for digging into it | 17:16 |
| *** zhill has quit IRC | 17:17 | |
| tdasilva | notmyname: so..just wondering...what's the next step regarding the tempest tests? | 17:17 |
| janonymous_ | Okay cool | 17:17 |
| janonymous_ | onovy: thanks | 17:18 |
| acoles | gmmaha: tdasilva: great | 17:18 |
| notmyname | tdasilva: next step, I think, is to talk to the tempest team (ie mtreinish) and let them know and hopefully they can pull our tests in now | 17:18 |
| janonymous_ | btw, i was working with testr for unittests ,it's nearly done with some ut's left for proxy to fix. | 17:20 |
| acoles | if gerrit is down i guess i will have to do expense reports | 17:20 |
| *** daemontool__ has joined #openstack-swift | 17:25 | |
| *** lyrrad has joined #openstack-swift | 17:25 | |
| janonymous_ | notmyname: Thanks, i can go now :) | 17:25 |
| janonymous_ | see ya guys. bye | 17:25 |
| notmyname | good night | 17:25 |
| *** gyee has joined #openstack-swift | 17:26 | |
| *** janonymous__ has quit IRC | 17:27 | |
| *** janonymous_ has quit IRC | 17:27 | |
| pchng | Is review.openstack.org down for anyone else? I'm getting a 503 error | 17:29 |
| gmmaha | pchng: i am told there is a known maintanence happening | 17:29 |
| pchng | gmmaha: okay, thanks for the note | 17:29 |
| gmmaha | http://lists.openstack.org/pipermail/openstack-dev/2015-December/081037.html | 17:29 |
| pchng | ah, okay, thanks explains it :) | 17:30 |
| gmmaha | pchng: anytime | 17:30 |
| *** Jeffrey4l has quit IRC | 17:31 | |
| *** zhill has joined #openstack-swift | 17:37 | |
| *** zhill has quit IRC | 17:42 | |
| *** jistr has quit IRC | 17:42 | |
| *** daemontool__ has quit IRC | 17:43 | |
| clayg | heyoh! | 17:43 |
| *** diazjf has quit IRC | 17:43 | |
| *** daemontool__ has joined #openstack-swift | 17:44 | |
| pdardeau | hey clayg! | 17:47 |
| tdasilva | clayg: hello! | 17:47 |
| acoles | clayg: hi | 17:48 |
| clayg | hey fellas | 17:50 |
| clayg | I'm digging out of emails and reading diffs today - i'd like to get a handle on what we have lined up with priority reviews asap | 17:50 |
| *** daemontool__ has quit IRC | 17:51 | |
| *** logan- has quit IRC | 17:51 | |
| acoles | clayg: eranrom and I made some progress with his container sync patch 205803, he's making a change right now i think | 17:53 |
| patchbot | acoles: https://review.openstack.org/#/c/205803/ | 17:53 |
| acoles | clayg: the func test change landed | 17:53 |
| *** dabukalam has quit IRC | 17:54 | |
| acoles | clayg: kota has a +2 on patch 231121 | 17:54 |
| patchbot | acoles: https://review.openstack.org/#/c/231121/ | 17:54 |
| acoles | didn't patchbot used to give us the commit subject? | 17:54 |
| acoles | maybe it can't while gerrit is down | 17:54 |
| *** dabukalam has joined #openstack-swift | 17:54 | |
| clayg | oh, sweet - no gerrit should make it easier to work on emails ;) | 17:55 |
| acoles | true :) | 17:56 |
| *** diazjf has joined #openstack-swift | 17:58 | |
| *** dmorita has quit IRC | 17:59 | |
| *** dmorita has joined #openstack-swift | 18:00 | |
| *** zhill has joined #openstack-swift | 18:02 | |
| *** klrmn has joined #openstack-swift | 18:04 | |
| *** zhill has quit IRC | 18:04 | |
| *** zhill has joined #openstack-swift | 18:05 | |
| *** garthb has quit IRC | 18:07 | |
| pchng | clayg: I submitted a small PR to vagrant SAIO to fix the proxy server config to reflect the trivial_keymaster -> keymaster rename: https://github.com/swiftstack/vagrant-swift-all-in-one/pull/32 | 18:09 |
| clayg | pchng: thanks for the heads up! | 18:12 |
| *** jordanP has quit IRC | 18:15 | |
| *** rohit_ has joined #openstack-swift | 18:17 | |
| pdardeau | any ideas on the highest num. partitions that folks use? | 18:24 |
| *** silor has joined #openstack-swift | 18:25 | |
| clayg | pdardeau: I'd *freak* out if anyone used anything higher than 24 - 22 is a much more likely upper bound | 18:26 |
| zaitcev | I think RAX was 18 all around until the last couple of years. | 18:26 |
| clayg | what time is gerrit supposed to come back? | 18:27 |
| ahale | zaitcev: rax was 20 | 18:27 |
| ahale | well, is i suppose | 18:27 |
| zaitcev | ahale: I see, mine is from defaults in gholt's old master class | 18:27 |
| pdardeau | clayg: thx! | 18:28 |
| *** daemontool__ has joined #openstack-swift | 18:28 | |
| pdardeau | clayg: part of why i asked is i got segfaults doing rebalance on 24 | 18:30 |
| clayg | add more memory to your vm | 18:33 |
| clayg | ;) | 18:33 |
| *** briancurtin has quit IRC | 18:33 | |
| *** briancurtin has joined #openstack-swift | 18:34 | |
| *** aix has quit IRC | 18:35 | |
| pdardeau | clayg: i took the other option (doc, it hurts when i do this. don't do that) | 18:35 |
| *** trifon has joined #openstack-swift | 18:43 | |
| *** zhill has quit IRC | 18:45 | |
| notmyname | hello clayg | 18:46 |
| clayg | notmyname: hey! | 18:47 |
| notmyname | clayg: are you coming in tot he office today, or just digging out from home? | 18:47 |
| clayg | i'm wfh today - i'm planning on coming in tomorrow | 18:47 |
| notmyname | got it | 18:47 |
| notmyname | clayg: (very) short summary of swift is ring stuff landed. testr functests landed. need to get eyes on copy middleware and concurrent reads | 18:48 |
| clayg | got most of the junk email out of the way - only ~100 I want to read, but a lot of the ones I'm interested in point to gerrit | 18:48 |
| notmyname | oh, we did a swiftclient release | 18:48 |
| clayg | notmyname: sounds awesome! | 18:48 |
| clayg | notmyname: great work to all! | 18:48 |
| notmyname | yeah, lots of progress | 18:48 |
| *** david-lyle_ has joined #openstack-swift | 18:49 | |
| *** david-lyle has quit IRC | 18:50 | |
| notmyname | clayg: I made a small PR to vSAIO to add some words around using it as a test target | 18:50 |
| *** david-lyle_ is now known as david-lyle | 18:54 | |
| *** lpabon has quit IRC | 18:58 | |
| *** eranrom has joined #openstack-swift | 19:00 | |
| onovy | notmyname, new swiftclient release? 2.7.0? | 19:02 |
| *** daemontool__ has quit IRC | 19:09 | |
| notmyname | onovy: yeah. happened last week | 19:09 |
| onovy | that's all, already building deb package :) | 19:09 |
| onovy | but good news everyone: swauth 1.1 is out | 19:10 |
| *** daemontool has joined #openstack-swift | 19:10 | |
| notmyname | cool | 19:10 |
| onovy | /all/old/ :] | 19:10 |
| *** zhill has joined #openstack-swift | 19:15 | |
| *** david-lyle has quit IRC | 19:17 | |
| *** david-lyle has joined #openstack-swift | 19:20 | |
| *** silor has quit IRC | 19:21 | |
| *** silor has joined #openstack-swift | 19:23 | |
| *** dmorita has quit IRC | 19:24 | |
| *** david-lyle_ has joined #openstack-swift | 19:24 | |
| *** david-lyle has quit IRC | 19:24 | |
| *** acoles is now known as acoles_ | 19:25 | |
| *** dmorita has joined #openstack-swift | 19:27 | |
| pdardeau | notmyname: any suggestions on how to test ring with large number of devices? | 19:28 |
| pdardeau | 60000-70000 ish | 19:28 |
| notmyname | pdardeau: you don't need any servers to do that. you just need some ram and CPU. probably would be good to talk to torgomatic (not online?) about his ring fuzzing tests | 19:29 |
| onovy | is swiftclient 2.7.0 for liberty or mitaka release? | 19:29 |
| timburke | mitaka | 19:29 |
| onovy | ok, thanks | 19:29 |
| notmyname | there will be others in the mitaka release too | 19:30 |
| onovy | of course :) | 19:30 |
| pdardeau | notmyname: ok, i'll do that | 19:30 |
| *** david-lyle_ has quit IRC | 19:30 | |
| pdardeau | notmyname: so how would i experience the 65535 device limit without using any servers? | 19:31 |
| notmyname | pdardeau: the limits are only in the builder and in the builder operations. you don't need to actually deploy a ring to test that it has more than 2**32 drives. likewise you don't even need to have servers or drives to add it to a ring | 19:32 |
| zaitcev | import swift.common.ring as rm; r = rm.Ring(something=65636) | 19:32 |
| zaitcev | I guess | 19:32 |
| notmyname | yeah :-) | 19:32 |
| zaitcev | and use ulimit(1) to see how much RAM is used | 19:32 |
| pdardeau | notmyname: got it! i'm able to add 70000 devices right now to the ring | 19:32 |
| notmyname | with current master code? | 19:32 |
| pdardeau | notmyname: yes | 19:33 |
| notmyname | so write_ring and make sure you can actually get all of them | 19:33 |
| notmyname | the dev_id serialization has the 2**16 limit | 19:33 |
| pdardeau | but not the way zaitcev has shown above | 19:33 |
| pdardeau | i have rb.add_dev in loop | 19:34 |
| zaitcev | pdardeau: but that's the test after you added and rebalanced it all. I thought you wanted to see what server does without actual server. | 19:34 |
| pdardeau | zaitcev: i see, thx for the explanation | 19:35 |
| *** trifon has quit IRC | 19:35 | |
| zaitcev | pdardeau: Like notmyname says, builder is not run on a server node or in a context of a server process. Builders are run on a management workstation. So that is what you're accelerating. After that, the *.ring.gz goes to servers and then it becomes interesting how much RAM that big a ring uses up. | 19:35 |
| notmyname | zaitcev: it's not just accelerating. also removing any device count limits (really that's the main thing) | 19:36 |
| notmyname | pdardeau: I'm looking for that part of the code now | 19:36 |
| pdardeau | zaitcev: i knew the part of the builder being offline cli mechanism, but thanks for mentioning | 19:38 |
| notmyname | pdardeau: I think you'll be mucking around in the swift/common/ring/ring.py:RingData class | 19:41 |
| notmyname | especially the [de]serialize_v1() methods | 19:41 |
| notmyname | ie you'll make a *_v2 for those | 19:41 |
| notmyname | assuming there is backwards incompatibility | 19:41 |
| *** zhill has quit IRC | 19:41 | |
| ahale | pdardeau: dunno if you found it yet, but doing a rebalance should make it error ? | 19:44 |
| notmyname | pdardeau: also, I'd suggest only using a part power of 17 to make rebalances faster | 19:46 |
| notmyname | ie dont use 24 or something ;-) | 19:47 |
| *** david-lyle has joined #openstack-swift | 19:48 | |
| pdardeau | ahale: thanks! that's what i was looking for, but not asking correctly (the symptom) | 19:49 |
| *** trifon has joined #openstack-swift | 19:50 | |
| *** daemontool has quit IRC | 19:50 | |
| ahale | no worries - ( like https://gist.github.com/racker-andrew-hale/acac080b71df32674a7b ) | 19:50 |
| *** bsdkurt has quit IRC | 19:53 | |
| *** lpabon has joined #openstack-swift | 19:55 | |
| Zyric | Morning | 19:59 |
| notmyname | Zyric: hello | 19:59 |
| notmyname | pdardeau: basically, it comes down to the array.array structure of replica2part2dev_id | 20:05 |
| *** lyrrad has quit IRC | 20:05 | |
| *** zhill has joined #openstack-swift | 20:05 | |
| *** lyrrad has joined #openstack-swift | 20:06 | |
| pdardeau | notmyname: got it, thx | 20:06 |
| *** diogogmt has joined #openstack-swift | 20:10 | |
| *** zhill has quit IRC | 20:10 | |
| *** lyrrad has quit IRC | 20:15 | |
| *** minwoob has quit IRC | 20:17 | |
| *** lyrrad has joined #openstack-swift | 20:19 | |
| *** trifon has quit IRC | 20:20 | |
| *** silor has quit IRC | 20:27 | |
| *** timur___ has joined #openstack-swift | 20:28 | |
| *** tsg has quit IRC | 20:33 | |
| *** changbl has joined #openstack-swift | 20:37 | |
| *** tsg has joined #openstack-swift | 20:41 | |
| *** trifon has joined #openstack-swift | 20:45 | |
| *** tsg has quit IRC | 20:47 | |
| onovy | what the f*ck with gerrit... :) | 20:53 |
| notmyname | onovy: upgrade. should be turned on at the top of the hour. see the topic message | 20:53 |
| onovy | notmyname, it's up 'n' running | 20:53 |
| notmyname | oh cool | 20:53 |
| onovy | and f*cked :) | 20:53 |
| onovy | new fancy look, etc. | 20:54 |
| notmyname | oh. yeah. I've played with that before. I'm told it's better this time | 20:54 |
| *** daddyjoseph97 has joined #openstack-swift | 20:55 | |
| *** ho has joined #openstack-swift | 20:55 | |
| notmyname | onovy: oh wow. yeah this is really different | 20:55 |
| onovy | really :) | 20:55 |
| mattoliverau | Morning | 20:55 |
| hrou | onovy, wow yea, and no way to revert back this time I think :) | 20:56 |
| onovy | hrou, can't found anything | 20:57 |
| mattoliverau | I'm too scared to go look | 20:57 |
| onovy | argh command for check-out is there. tiny download button on top right | 20:57 |
| *** tsg has joined #openstack-swift | 20:57 | |
| ho | good morning! | 20:57 |
| *** CaioBrentano has quit IRC | 20:57 | |
| notmyname | it will take a while to get used to | 20:57 |
| onovy | yep | 20:57 |
| notmyname | I hope it's better than the last time I tried the new look | 20:58 |
| *** acoles_ is now known as acoles | 20:58 | |
| mattoliverau | +1 | 20:58 |
| kota_ | good morning | 20:59 |
| notmyname | meeting time. looks like the previous one is finishing up | 21:00 |
| acoles | kota_: mattoliverau : good morning | 21:00 |
| notmyname | in #openstack-meeting | 21:00 |
| *** bjkeller has joined #openstack-swift | 21:01 | |
| *** ChanServ changes topic to "Review Dashboard: https://goo.gl/gSuaEG | Get an IRC bouncer at http://znc.in | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | DocImpact changes http://lists.openstack.org/pipermail/openstack-dev/2015-November/080294.html" | 21:02 | |
| -openstackstatus- NOTICE: Gerrit has been upgraded to 2.11. Please report any issues in #openstack-infra as soon as possible. | 21:02 | |
| *** trifon has quit IRC | 21:03 | |
| openstackgerrit | Ondřej Nový proposed openstack/swift: Keystone middleware deprecated option is_admin removed https://review.openstack.org/258158 | 21:05 |
| pchng | Mmm new gerrit. Interesting UI changes | 21:05 |
| *** ober37 has joined #openstack-swift | 21:06 | |
| *** bjkeller has quit IRC | 21:07 | |
| *** bjkeller has joined #openstack-swift | 21:08 | |
| *** daemontool has joined #openstack-swift | 21:08 | |
| *** changbl has quit IRC | 21:11 | |
| onovy | pchng, __interesting__ :) | 21:11 |
| pchng | It's... good I guess? Looks like the UI is a little "flatter", some stuff moved around. Not sure what is new. I guess the +/- add/remove indicators is now a green/red bar? | 21:13 |
| mattoliverau | it's going to take me some getting used to, I hope I don't abandon anyones changes by default :P | 21:14 |
| *** zhill has joined #openstack-swift | 21:15 | |
| *** trifon has joined #openstack-swift | 21:18 | |
| *** esker has quit IRC | 21:20 | |
| *** daemontool_ has joined #openstack-swift | 21:22 | |
| *** daemontool__ has joined #openstack-swift | 21:22 | |
| *** daemontool has quit IRC | 21:23 | |
| *** trifon has quit IRC | 21:23 | |
| *** daemontool_ has quit IRC | 21:26 | |
| *** dmorita has quit IRC | 21:27 | |
| *** rcernin has joined #openstack-swift | 21:28 | |
| *** dmorita has joined #openstack-swift | 21:28 | |
| acoles | mattoliverau: you can switch off the coloured bars in settings, and siwtch on patch number in listings which is a + imho | 21:29 |
| *** ober37 has quit IRC | 21:32 | |
| *** ober37 has joined #openstack-swift | 21:34 | |
| *** ober37 has quit IRC | 21:35 | |
| *** ober37 has joined #openstack-swift | 21:35 | |
| acoles | notmyname: do you see all patches in dashboard belong to Anonymous Coward? or is just me? | 21:48 |
| notmyname | just you | 21:48 |
| acoles | hmmm | 21:49 |
| *** daemontool_ has joined #openstack-swift | 21:52 | |
| openstackgerrit | Merged openstack/swift: remove unused parameters in the method https://review.openstack.org/243401 | 21:52 |
| *** daemontool__ has quit IRC | 21:56 | |
| *** daemontool_ has quit IRC | 21:56 | |
| tdasilva | acoles: I saw Anonymous Coward when I was not signed in... | 21:58 |
| *** ober37 has quit IRC | 21:59 | |
| acoles | tdasilva: oic it shows me signed in but maybe i need to refresh. thanks | 22:00 |
| tdasilva | acoles: nvm, it's not that, I saw it once and now can't replicate it | 22:00 |
| notmyname | jrichli: torgomatic seems to be not online (and is out sick). clayg is in, but is catching up | 22:01 |
| jrichli | notmyname: thanks. | 22:01 |
| onovy | trying again here: if someone can get me opinion for patch 251151 ; just say localtime or gmt, that's all :) | 22:02 |
| patchbot | onovy: https://review.openstack.org/#/c/251151/ - Show local time in swift-recon in replication part | 22:02 |
| acoles | notmyname: so, the issue with ranged gets is, afaik a multipart response has unpredicatble number of parts, each of which, for us has an encrypted content-type header which is longer than the plaintext. So I'm not sure how we calculate the content-length sent to the client without reading the entire body. | 22:02 |
| acoles | notmyname: but I'm wondering if EC didn't have to deal with similar | 22:03 |
| *** daemontool has joined #openstack-swift | 22:04 | |
| acoles | tdasilva: i signed out/in and all is good now | 22:04 |
| acoles | tdasilva: "good" being relative :) | 22:04 |
| blmartin | cschwede: onovy: I updated patch 257577 so if you get bored today | 22:05 |
| patchbot | blmartin: https://review.openstack.org/#/c/257577/ - Print min_part_hours lockout time remaining | 22:05 |
| clayg | oh - change spark lines in gerrit now! | 22:05 |
| *** bjkeller has quit IRC | 22:07 | |
| onovy | blmartin, i will look | 22:08 |
| *** ianbrown has quit IRC | 22:08 | |
| blmartin | onovy: much appreciated | 22:08 |
| onovy | and peterlisak will look too, right peter? :) | 22:08 |
| *** ianbrown has joined #openstack-swift | 22:08 | |
| acoles | clayg: looks like the 'f' kbd shortcut has gone from gerrit :( that was SO useful | 22:10 |
| mattoliverau | cool in gerrit settings you can add dash boards (you could probaby do this before but I never looked). So now the swift review dashboard in under the 'My' menu in gerrit. | 22:11 |
| peterlisak | onovy, blmartin , yes, sure but tomorrow ... time to go off :) | 22:11 |
| acoles | and we can edit code in gerrit as well as commit messages. scary, | 22:11 |
| acoles | mattoliverau: wasn't the swift dashboard already your browser home page? ;) | 22:13 |
| clayg | acoles: everything gerrit did that didn't annoy you is gone - all the things you hate are worse - and there's new things you didn't even think of that they invinted just to make you less productive | 22:13 |
| mattoliverau | acoles: yeah.. of course, bookmarked, but now I can add it as apart of my gerrit login | 22:13 |
| blmartin | clayg: but it looks soo much flashier though doesn't it! think of how stylish your monitor will look at conferences with gerrit open now | 22:15 |
| acoles | clayg: precisely | 22:15 |
| clayg | blmartin: fair point - sounds like a net win to me | 22:15 |
| *** diazjf has quit IRC | 22:15 | |
| acoles | mattoliverau: oic and you can make any dashboard be the default page by moving it to top | 22:15 |
| *** lpabon has quit IRC | 22:15 | |
| *** diazjf has joined #openstack-swift | 22:16 | |
| acoles | argh, and 'r' is no longer 'leave review' but does something else!! | 22:16 |
| mattoliverau | lol | 22:16 |
| mattoliverau | R is reload change | 22:17 |
| acoles | don't they realise that some of us are too old to learn new ways! | 22:17 |
| onovy | hey, why is "reply" button at top of page, not near to comments? :/ | 22:17 |
| acoles | little r is mark this file as reviewd | 22:17 |
| mattoliverau | onovy: so you can scroll back up and hit reply :P | 22:17 |
| acoles | its to give you time to reconsider | 22:18 |
| mattoliverau | acoles: oh yes, toggle the review flag | 22:18 |
| onovy | blmartin, just one small thing, but maybe i'm missing something. just read it, will try to test it tomorrow, thank you | 22:18 |
| onovy | mattoliverau, :) that's it! | 22:18 |
| mattoliverau | onovy: 'a' is review and publish comments, I hope not all at once. | 22:19 |
| acoles | I do like that I can have the review number as a column in the dashboard. I must focus on that positive. | 22:19 |
| onovy | mattoliverau, thanks, opens comment dialog | 22:19 |
| onovy | rows in tables are higher. this is not good, because i still have 100+ outgoing reviews and that list is really long :/ | 22:20 |
| onovy | and i'm not sure if that "size" column is good. nobody will open "long" reviews now :) | 22:22 |
| acoles | onovy: agree. you can turn it off in settings. | 22:22 |
| blmartin | onovy: likely it is I who is missing something, I look forward to the comments | 22:23 |
| onovy | acoles, no i like it. i will do review for "short" commits only :) | 22:23 |
| mattoliverau | lol | 22:24 |
| *** zhill has quit IRC | 22:25 | |
| acoles | i wonder, if we have one enormous review will all the others go green? | 22:25 |
| blmartin | well, breaking a patch into 30 parts probably isn't that bad I guess | 22:25 |
| onovy | with gerrit deps it's cool. Commit message: First part of my change, does nothing. Commit message2: Second part of my change, still does nothing.... | 22:26 |
| *** blmartin has quit IRC | 22:27 | |
| acoles | ok i am done whinging, good night. | 22:27 |
| clayg | 'night | 22:28 |
| timburke | acoles: i'm guessing no, but only because the opposite doesn't seem to be true: https://review.openstack.org/#/q/project:openstack/swift3+is:open+-label:Workflow-1+-label:Verified-1 | 22:28 |
| onovy | i'm going too. good night | 22:28 |
| briancline | g'night | 22:29 |
| *** diazjf has quit IRC | 22:31 | |
| *** acoles is now known as acoles_ | 22:32 | |
| *** asettle is now known as asettle-sick | 22:39 | |
| *** daemontool has quit IRC | 22:40 | |
| *** zhill has joined #openstack-swift | 22:45 | |
| *** diazjf has joined #openstack-swift | 22:47 | |
| mattoliverau | acoles_, onovy night | 22:47 |
| *** diazjf has quit IRC | 22:48 | |
| openstackgerrit | Timur Alperovich proposed openstack/swift: Fix IPv6 handling in MemcacheConnPool. https://review.openstack.org/258704 | 22:55 |
| *** daemontool has joined #openstack-swift | 22:55 | |
| kota_ | clayg: !!! | 22:59 |
| *** tsg has quit IRC | 23:00 | |
| *** tsg has joined #openstack-swift | 23:04 | |
| *** km has joined #openstack-swift | 23:06 | |
| *** alejandrito has quit IRC | 23:08 | |
| *** daemontool has quit IRC | 23:09 | |
| *** tsg has quit IRC | 23:13 | |
| *** zhill has quit IRC | 23:15 | |
| *** rohit_ has quit IRC | 23:18 | |
| timur___ | quit | 23:23 |
| *** timur___ has quit IRC | 23:23 | |
| *** dustins is now known as dustins|out | 23:30 | |
| *** kei_yama has joined #openstack-swift | 23:31 | |
| *** zhill has joined #openstack-swift | 23:40 | |
| *** changbl has joined #openstack-swift | 23:48 | |
| openstackgerrit | Tim Burke proposed openstack/swift: Expose token expiration time in tempauth auth response https://review.openstack.org/258720 | 23:56 |
| *** chlong has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!