| *** fifieldt has joined #openstack-marconi | 00:24 | |
| *** shakamunyi has joined #openstack-marconi | 00:50 | |
| *** shakamunyi has quit IRC | 00:54 | |
| *** nosnos has joined #openstack-marconi | 01:50 | |
| *** shakamunyi has joined #openstack-marconi | 02:51 | |
| *** ayoung has quit IRC | 02:54 | |
| *** shakamunyi has quit IRC | 02:56 | |
| *** ayoung has joined #openstack-marconi | 03:09 | |
| *** ayoung is now known as ayoung_ZZZ | 03:09 | |
| *** haomaiwang has joined #openstack-marconi | 03:52 | |
| *** shakamunyi has joined #openstack-marconi | 03:52 | |
| *** shakamunyi has quit IRC | 03:57 | |
| *** nosnos has quit IRC | 04:10 | |
| *** shakamunyi has joined #openstack-marconi | 04:53 | |
| *** shakamunyi has quit IRC | 04:57 | |
| *** chandan_kumar has joined #openstack-marconi | 05:06 | |
| *** nosnos has joined #openstack-marconi | 05:13 | |
| *** chandan_kumar has quit IRC | 05:31 | |
| *** chandan_kumar has joined #openstack-marconi | 05:42 | |
| *** shakamunyi has joined #openstack-marconi | 05:53 | |
| *** shakamunyi has quit IRC | 05:58 | |
| *** flaper87|afk is now known as flaper87 | 06:42 | |
| *** chandan_kumar has quit IRC | 06:46 | |
| *** shakamunyi has joined #openstack-marconi | 06:54 | |
| *** chandan_kumar has joined #openstack-marconi | 06:54 | |
| *** shakamunyi has quit IRC | 06:58 | |
| *** shakamunyi has joined #openstack-marconi | 07:55 | |
| *** shakamunyi has quit IRC | 08:00 | |
| *** ykaplan has joined #openstack-marconi | 08:18 | |
| *** reed has joined #openstack-marconi | 08:20 | |
| *** jamie_h has joined #openstack-marconi | 08:23 | |
| *** chandan_kumar has quit IRC | 08:46 | |
| *** shakamunyi has joined #openstack-marconi | 08:56 | |
| *** shakamunyi has quit IRC | 09:01 | |
| *** chandan_kumar has joined #openstack-marconi | 09:04 | |
| *** nosnos has quit IRC | 10:05 | |
| *** ykaplan has quit IRC | 10:06 | |
| *** chandan_kumar has quit IRC | 10:32 | |
| *** jchai has joined #openstack-marconi | 10:38 | |
| *** chandan_kumar has joined #openstack-marconi | 10:49 | |
| *** fifieldt has quit IRC | 11:07 | |
| *** fifieldt has joined #openstack-marconi | 11:08 | |
| *** ykaplan has joined #openstack-marconi | 11:19 | |
| *** haomaiw__ has joined #openstack-marconi | 11:35 | |
| *** haoma____ has joined #openstack-marconi | 11:38 | |
| *** haomaiwang has quit IRC | 11:39 | |
| *** haomaiw__ has quit IRC | 11:40 | |
| *** chandan_kumar has quit IRC | 12:06 | |
| *** chandan_kumar has joined #openstack-marconi | 12:24 | |
| *** prashanthr_ has joined #openstack-marconi | 12:24 | |
| *** tedross has joined #openstack-marconi | 12:32 | |
| *** sriram has joined #openstack-marconi | 12:50 | |
| *** sriram has quit IRC | 12:51 | |
| *** sriram has joined #openstack-marconi | 12:51 | |
| *** amitgandhi has joined #openstack-marconi | 12:51 | |
| *** jamie_h_ has joined #openstack-marconi | 12:56 | |
| *** jamie_h has quit IRC | 12:56 | |
| *** ayoung_ZZZ is now known as ayoung | 13:10 | |
| *** amitgandhi has quit IRC | 13:10 | |
| *** amitgandhi has joined #openstack-marconi | 13:11 | |
| *** Obulpathi has joined #openstack-marconi | 13:25 | |
| *** Obulpathi has quit IRC | 13:35 | |
| *** Obulpathi has joined #openstack-marconi | 13:35 | |
| *** jmckind has joined #openstack-marconi | 13:42 | |
| *** alcabrera|afk is now known as alcabrera | 13:47 | |
| *** malini_afk is now known as malini | 13:48 | |
| alcabrera | good morning. :) | 13:48 |
|---|---|---|
| flaper87 | alcabrera: goood morning :) | 13:49 |
| flaper87 | how are you doing? | 13:49 |
| *** balajiiyer has joined #openstack-marconi | 13:49 | |
| *** vkmc has joined #openstack-marconi | 13:50 | |
| *** vkmc has quit IRC | 13:50 | |
| *** vkmc has joined #openstack-marconi | 13:50 | |
| alcabrera | flaper87: really happy, a little on the groggy side - pretty great! :D | 13:50 |
| alcabrera | flaper87: you? | 13:50 |
| * flaper87 gives alcabrera a cup of just-queued-out coffee \_/? | 13:51 | |
| flaper87 | I'm doing great, just went through a bunch of booking madness | 13:51 |
| * alcabrera drinks and enjoys | 13:52 | |
| malini | hello!! | 13:52 |
| flaper87 | malini: hey hey | 13:52 |
| alcabrera | flaper87: I saw some of that! europython and more. ;) | 13:52 |
| alcabrera | malini: good morning! :D | 13:52 |
| flaper87 | malini: btw, did you read the message I sent on Saturday (or Sunday?)? | 13:52 |
| malini | good morning alcabrera! | 13:52 |
| * flaper87 realizes he should get a life | 13:52 | |
| malini | flaper87: CLI for the client? | 13:52 |
| flaper87 | malini: yup | 13:52 |
| malini | yes!! Tht is awesome!!!!! | 13:53 |
| flaper87 | malini: ok, now, you should watch the client gerrit project too ;) | 13:53 |
| flaper87 | :P | 13:53 |
| malini | I was trying to add CLI tests in tempest when I realized we dont have CLI yet | 13:53 |
| malini | we do true TDD :D | 13:53 |
| malini | good point flaper87 | 13:53 |
| flaper87 | malini: we did that on purpose, FWIW. We wanted to keep marconiclient as a library | 13:53 |
| flaper87 | until the python-openstack-common CLI came out | 13:54 |
| malini | aah..ok | 13:54 |
| malini | is there a python-openstack-common CLI now | 13:54 |
| malini | ? | 13:54 |
| flaper87 | that's a nice way to add CLI support by still keeping the library nature of the project | 13:54 |
| flaper87 | malini: that's what we're using :) | 13:54 |
| flaper87 | it's basically a base class and a bunch of entry_points put together and loaded dynamically | 13:55 |
| vkmc | Hi all! :) | 13:55 |
| flaper87 | vkmc: hellooooooooooooooooooooooooooooooooooooooo | 13:56 |
| malini | hello vkmc!!! | 13:56 |
| sriram | Good Morning :) | 13:56 |
| flaper87 | sriram: gooood morning | 13:56 |
| sriram | \m/ | 13:56 |
| malini | flaper87: I was checking a few other clients on Friday, and all of them have a shell.py -which they use in Tempesteg. https://github.com/openstack/python-glanceclient/blob/master/glanceclient/shell.py will we have the same? | 13:57 |
| malini | good morning sriram! | 13:57 |
| flaper87 | malini: yeah, I really hate that code :P | 13:57 |
| flaper87 | FWIW, we're starting to discuss glanceclient's switch-over to the common client | 13:57 |
| sriram | good morning malini,vkmc,alcabrera :) | 13:57 |
| flaper87 | sriram: hey, no optimizations allowed. you gotta send 1 dedicated message to each one of them | 13:58 |
| flaper87 | >.> | 13:58 |
| malini | :D | 13:58 |
| vkmc | How is the marconi-client going? I always see you discussing about it but the truth is I don't know in which stage you are... I'll appreciate a link | 13:59 |
| malini | flaper87: are there any other projects that already use the common client? | 13:59 |
| flaper87 | vkmc: https://github.com/openstack/python-marconiclient/ | 13:59 |
| flaper87 | we've a library, it supports v1 but there's no CLI support | 14:00 |
| alcabrera | sriram: good morning! | 14:00 |
| sriram | flaper87: heh | 14:00 |
| *** tedross has quit IRC | 14:00 | |
| vkmc | flaper87, Awesome, thanks | 14:00 |
| flaper87 | malini: mmh, I know there's support for nova, glance, keystone, cinder etc | 14:00 |
| flaper87 | but those projects still rely on the old code | 14:01 |
| flaper87 | malini: https://github.com/openstack/python-openstackclient/tree/master/openstackclient | 14:01 |
| flaper87 | I see the benefit of having separate releases for the client libraries but, I still think it's kinda wrong... | 14:02 |
| flaper87 | I'd like us to always implement things in both, the server and the client library | 14:02 |
| malini | yeah..it feels wrong | 14:02 |
| flaper87 | whoever is implementing a blueprint should make sure the client library has support for that | 14:03 |
| alcabrera | we're at the point where we can follow through with that, even work in parallel | 14:03 |
| alcabrera | if a feature is to be added to the server | 14:03 |
| alcabrera | another can tackle the analagous client-side feature | 14:03 |
| *** rwsu has joined #openstack-marconi | 14:08 | |
| *** tedross has joined #openstack-marconi | 14:18 | |
| vkmc | In most projects I saw that clients are not perfectly updated with the server features... it's a bummer but that's the way most interfaces are | 14:23 |
| alcabrera | agreed, vkmc | 14:25 |
| alcabrera | it's hard to keep pace w/o making that a top priority | 14:25 |
| vkmc | The library looks great, btw! | 14:25 |
| vkmc | alcabrera, Yeah... your proposed strategy to tackle that issue is the best we can do | 14:26 |
| alcabrera | yup, yup. :) | 14:27 |
| alcabrera | vkmc: and flaper87 has done an awesome job of leading the development on the client! | 14:28 |
| vkmc | He totally did! | 14:28 |
| *** shakamunyi has joined #openstack-marconi | 14:28 | |
| vkmc | I remember when I tried to work on that... so naive... haha | 14:29 |
| *** tedross has left #openstack-marconi | 14:29 | |
| alcabrera | it gets easier, vkmc. :) | 14:30 |
| alcabrera | focusing on haskell in my free time reminds me that I'm still a novice in many ways, but I'm *so* happy to always have more to learn. <3 | 14:30 |
| -openstackstatus- NOTICE: Gerrit downtime for upgrade begins in 90 minutes. See: https://wiki.openstack.org/wiki/GerritUpgrade | 14:31 | |
| vkmc | Hope so :) | 14:31 |
| prashanthr_ | Good morning All ! :) | 14:39 |
| vkmc | Morning prashanthr_ o/ | 14:39 |
| alcabrera | prashanthr_: good morning! | 14:40 |
| malini | hello prashanthr_! | 14:40 |
| vkmc | alcabrera, There are so many things to learn in our field! Time never is enough -.- | 14:41 |
| alcabrera | es la verdad, vkmc. :P | 14:42 |
| prashanthr_ | vkmc: true that. | 14:42 |
| *** cpallares has joined #openstack-marconi | 14:42 | |
| *** mwagner_lap has quit IRC | 14:43 | |
| *** shakamunyi has quit IRC | 14:44 | |
| vkmc | En fin! *sighs* | 14:44 |
| flaper87 | alcabrera: remeber that Rust lang talk I mentioned? | 15:00 |
| flaper87 | remember* | 15:00 |
| flaper87 | the one scheduled for the 29th | 15:00 |
| alcabrera | yup! | 15:01 |
| alcabrera | how'd it go, flaper87? | 15:01 |
| flaper87 | alcabrera: well, 29 is tomorrow | 15:01 |
| flaper87 | BUT | 15:01 |
| alcabrera | oh | 15:01 |
| flaper87 | IT WAS TODAY!!!!!!!!!!!!!!! | 15:01 |
| alcabrera | yes | 15:01 |
| alcabrera | haha | 15:01 |
| flaper87 | T_T | 15:01 |
| alcabrera | @_@ | 15:01 |
| alcabrera | oh geez | 15:01 |
| *** kgriffs|afk is now known as kgriffs | 15:02 | |
| alcabrera | will it be okay, flaper87? | 15:02 |
| alcabrera | will *you* be okay, flaper87? D: | 15:02 |
| flaper87 | someone pinged me on IRC like: Dude, are you alive? People is waiting for you | 15:02 |
| flaper87 | and I replied: I *was* alive (plop) | 15:02 |
| flaper87 | anyway, just did it | 15:02 |
| flaper87 | it went decently | 15:02 |
| alcabrera | hurray! | 15:02 |
| alcabrera | <3 | 15:02 |
| flaper87 | not well, decently | 15:02 |
| flaper87 | I just gave a very high-level introduction to Rust | 15:03 |
| alcabrera | decently to me means -- information was shared, and people have a starting point. :) | 15:03 |
| *** chandan_kumar has quit IRC | 15:03 | |
| flaper87 | alcabrera: exactly that | 15:03 |
| flaper87 | it could've been better though but anyway | 15:03 |
| *** shakamunyi has joined #openstack-marconi | 15:04 | |
| cpallares | flaper87: :D | 15:04 |
| cpallares | flaper87: Where did you present? Was it recorded? | 15:04 |
| kgriffs | flaper87: so, about Rust | 15:06 |
| kgriffs | does it have async goodness built in or as a library? | 15:07 |
| flaper87 | cpallares: yup, recorded, I was actually looking for you to invite you | 15:10 |
| cpallares | flaper87: link? | 15:10 |
| flaper87 | kgriffs: asyn as in, green threads with nice IO handling? | 15:11 |
| kgriffs | yeah | 15:11 |
| flaper87 | kgriffs: it does, it has 2 threading models. Native threads and green threads | 15:11 |
| flaper87 | the default is native threads | 15:11 |
| flaper87 | (although it used to be the other way around | 15:11 |
| flaper87 | ) | 15:11 |
| flaper87 | and green thread's IO uses libuv under the hood | 15:12 |
| flaper87 | (node.js IO library) | 15:12 |
| kgriffs | yep | 15:12 |
| kgriffs | hmm | 15:12 |
| kgriffs | so, does it use promises or coroutines or ? | 15:12 |
| *** jchai is now known as jchai_afk | 15:12 | |
| alcabrera | it was a spawning method, iirc, kgriffs: http://static.rust-lang.org/doc/master/std/task/index.html | 15:13 |
| alcabrera | so | 15:13 |
| *** jchai_afk is now known as jchai | 15:13 | |
| alcabrera | much like gevent.spawn, or haskell's forkIO, or golang's goroutines | 15:14 |
| kgriffs | interesting | 15:14 |
| kgriffs | next question - how does memory management work in Rust? | 15:14 |
| flaper87 | kgriffs: it's user managed. That is, stack / heap allocated values that are zero'd out when they go out of scope | 15:17 |
| flaper87 | kgriffs: there was support for GC'd values but the syntactic sugar for that was removed and the whole GC code was moved into a library | 15:17 |
| alcabrera | GC is therefore optional. :) | 15:17 |
| flaper87 | so now you can have `Gc` objects that contain a GC managed value | 15:17 |
| flaper87 | alcabrera: yup :) | 15:17 |
| kgriffs | flaper87: final question: is there a kick-a** HTTP client library? | 15:18 |
| alcabrera | hahha, I vaguely remember flaper87 working on such a thing. ;) | 15:18 |
| alcabrera | kgriffs: ^ | 15:18 |
| flaper87 | kgriffs: there's a fair rust-http library but a new one based on the old one and maintained by the same guy was recently announced | 15:18 |
| kgriffs | ok | 15:18 |
| kgriffs | flaper87: oh, one other question - zeromq support? | 15:19 |
| kgriffs | (or something similar) | 15:19 |
| flaper87 | kgriffs: https://github.com/erickt/rust-zmq | 15:20 |
| alcabrera | rust is pretty amazing. It's the best imperative-oriented language available to date, imo. | 15:20 |
| kgriffs | mmm | 15:20 |
| flaper87 | kgriffs: http://www.rust-ci.org/projects/ | 15:20 |
| flaper87 | (there's nanomsg too) | 15:20 |
| kgriffs | is rustpkg any good? | 15:21 |
| kgriffs | I guess I should go play with all this | 15:21 |
| kgriffs | :p | 15:21 |
| kgriffs | I have been learning Go, but I'm not sure that I like it very much | 15:22 |
| malini | before you go play with rust, kgriffs | 15:22 |
| malini | me & sriram are planning to work on benchmarking marconi this week | 15:22 |
| malini | we have an inital plan here https://etherpad.openstack.org/p/marconi-benchmark-plans | 15:23 |
| alcabrera | the lack of generics in Go is a killer downer | 15:23 |
| malini | Can everybody take a look & add your comments? | 15:23 |
| *** prashanthr_ has left #openstack-marconi | 15:23 | |
| alcabrera | as well as the really obtuse separation between built-in types, maps, channels, and every data type a user might define | 15:23 |
| malini | flaper87: Can you add your thoughts to https://etherpad.openstack.org/p/marconi-benchmark-plans ? | 15:26 |
| flaper87 | kgriffs: rustpkg is dead :) | 15:27 |
| flaper87 | kgriffs: there's a new package manager that `wycast` is working on | 15:27 |
| flaper87 | wycast -> author of bundler | 15:27 |
| flaper87 | and ember.js | 15:27 |
| flaper87 | IIRC | 15:27 |
| flaper87 | anyway, enough publicity | 15:27 |
| flaper87 | just wanted to say that the old rustpkg was deprecated and a new one is being developed under a paid contract | 15:28 |
| kgriffs | flaper87: ah, ok. good to know | 15:28 |
| kgriffs | flaper87: so, I was thinking about writing a killer benchmarking tool for HTTP APIs. | 15:29 |
| -openstackstatus- NOTICE: Gerrit downtime for upgrade begins in 30 minutes. See: https://wiki.openstack.org/wiki/GerritUpgrade | 15:30 | |
| alcabrera | Rust/Haskell/Erlang are all pretty amazing for accomplishing just that, kgriffs. ;) | 15:31 |
| kgriffs | I like the idea of using Rust without GC, so I don't have to worry about the runtime kicking in and skewing results | 15:31 |
| kgriffs | Tsung is Erlang... but I'd like a tool that was more optimized for API testing vs. websites, and a tool that gives more straightforward, consistent reports | 15:32 |
| kgriffs | plus, I need an excuse to learn a new language, and that new language is not Erlang. :p | 15:33 |
| alcabrera | lol | 15:33 |
| *** malini is now known as malini_afk | 15:37 | |
| *** malini_afk is now known as malini | 15:39 | |
| *** jamie_h_ has quit IRC | 15:48 | |
| *** jamie_h has joined #openstack-marconi | 15:49 | |
| *** mwagner_lap has joined #openstack-marconi | 15:51 | |
| kgriffs | malini, alcabrera: so, it seems we have a few ways to configure the client behavior | 15:52 |
| kgriffs | 1. a conf file and CLI options ala httperf, ab | 15:52 |
| kgriffs | 2. XML or JSON ala Tsung (https://github.com/rackerlabs/csi-marconi/blob/master/load/tsung.xml) | 15:53 |
| kgriffs | 3. Use the benchmarking language directly ala JMeter | 15:53 |
| * alcabrera googles the JMeter | 15:54 | |
| kgriffs | 4. Host a minimal scripting language like Lua | 15:55 |
| alcabrera | (3) with a FP language could be beautiful. (4) is not a bad place to be at, either. | 15:55 |
| kgriffs | 5. write a DSL which may be a subset or superset of an existing language | 15:56 |
| malini | mpanetta had an interesting idea recently | 15:56 |
| alcabrera | and by (3), I mean (5) | 15:56 |
| alcabrera | :D | 15:56 |
| malini | am trying to find tht mail | 15:56 |
| alcabrera | or hmmm | 15:56 |
| alcabrera | yup, actually I meant (3) | 15:56 |
| alcabrera | but yes | 15:56 |
| alcabrera | mpanetta did have some very interesting ideas | 15:56 |
| malini | kgriffs: I just forwarded you a mail | 15:57 |
| kgriffs | reading | 16:00 |
| kgriffs | interesting | 16:01 |
| kgriffs | so it seems like the heart of his idea is the ability to benchmark end-to-end scenarous | 16:02 |
| kgriffs | which means the tool needs a way to trace the lifetime of a message across multiple clients. | 16:02 |
| * alcabrera imagines Google Dapper; trace-based profiling | 16:04 | |
| *** ykaplan has quit IRC | 16:06 | |
| *** Obulpathi has quit IRC | 16:07 | |
| *** amitgandhi has quit IRC | 16:11 | |
| *** Obulpathi has joined #openstack-marconi | 16:25 | |
| -openstackstatus- NOTICE: Gerrit is unavailable until further notice for a major upgrade. See: https://wiki.openstack.org/wiki/GerritUpgrade | 16:36 | |
| *** ChanServ changes topic to "Gerrit is unavailable until further notice for a major upgrade. See: https://wiki.openstack.org/wiki/GerritUpgrade" | 16:36 | |
| kgriffs | malini: is this bp still valid? https://blueprints.launchpad.net/marconi/+spec/devstack-marconi-tests | 16:44 |
| *** shakamunyi has quit IRC | 16:46 | |
| alcabrera | malini, kgriffs: very relevant to our benchmarking discussions: https://twitter.com/SciencePorn/status/460577159929933824/photo/1 | 16:56 |
| kgriffs | hehe | 17:01 |
| *** jmckind has quit IRC | 17:06 | |
| malini | kgriffs: we still need tht bp implemented | 17:08 |
| kgriffs | malini: ok, should it be Juno-1, or later? | 17:08 |
| malini | let me chk the timeline for Juno-1 | 17:09 |
| malini | I cant find the timelines :( | 17:12 |
| malini | when should we have this done , if it were Juno-1 ? | 17:13 |
| malini | kgriffs: ^ | 17:13 |
| *** shakamunyi has joined #openstack-marconi | 17:13 | |
| kgriffs | the juno schedule probably isn't finalized yet, so we will have to guesstimate | 17:16 |
| kgriffs | malini: probably 6 weeks after the summit | 17:17 |
| malini | Lets make it Juno-2 | 17:17 |
| kgriffs | ok | 17:17 |
| kgriffs | https://bugs.launchpad.net/marconi/+bug/1292136 | 17:18 |
| kgriffs | malini: how about this bug? Juno-1? | 17:18 |
| malini | The last I heard we'll have Trusty at the gate by June | 17:19 |
| malini | So lets keep this for Juno-1 | 17:19 |
| kgriffs | ok | 17:19 |
| kgriffs | malini: this one we can mark as "won't fix" right? https://bugs.launchpad.net/marconi/+bug/1294068 | 17:20 |
| malini | are we officially moving sqlalchemy to the back burner? | 17:21 |
| kgriffs | yes. We won't remove it because it is handy for development not to have to spin up mongo or something, but I don't think we need to gate on MySQL, just sqlite. | 17:23 |
| kgriffs | (imo) | 17:23 |
| malini | in tht case, we can mark it as won't fix | 17:23 |
| kgriffs | kk, thanks! | 17:25 |
| kgriffs | bbl | 17:25 |
| *** chandan_kumar has joined #openstack-marconi | 17:26 | |
| *** jamie_h has quit IRC | 17:27 | |
| *** balajiiyer has quit IRC | 17:29 | |
| *** shakamunyi has quit IRC | 17:35 | |
| *** kgriffs is now known as kgriffs|afk | 17:36 | |
| *** ayoung is now known as ayoung_rrrrmmm | 17:55 | |
| *** kgriffs|afk is now known as kgriffs | 18:02 | |
| *** tjanczuk has joined #openstack-marconi | 18:03 | |
| tjanczuk | Hi folks. I was wondering who is attending the summit in Atlanta and what a best way to meet the Marconi team would be during the event? | 18:05 |
| alcabrera | tjanczuk: I'm attending. As to the best way to meet the Marconi team, I feel that malini is more on top of this than I. :) | 18:07 |
| malini | tjanczuk: we are planning a team outing kinda thing in ATL | 18:08 |
| malini | will you be interested in joining us? | 18:08 |
| tjanczuk | if you don't mind me crashing the party I would love to. | 18:09 |
| malini | the more the merrier | 18:09 |
| tjanczuk | do you already have a pre-agreed time & place? | 18:09 |
| malini | not yet | 18:10 |
| malini | Most probably it'll be on Sunday before the summit | 18:10 |
| malini | from the travel schedules we have seen so far | 18:10 |
| malini | I have added you to the list here https://etherpad.openstack.org/p/marconi-party | 18:11 |
| *** rossk has joined #openstack-marconi | 18:11 | |
| malini | We have the marconi meeting tomorrow & hope to finalize the date/time then | 18:11 |
| flaper87 | tjanczuk: I only know I'll eat A LOT! | 18:12 |
| tjanczuk | thanks for adding me to the list (took me a while to locate it under "humans"; we can fix it later) | 18:13 |
| tjanczuk | I am only flying in at 9pm on Sunday. | 18:13 |
| malini | :( | 18:14 |
| tjanczuk | In general how do you guys plan to stay in touch during the event for coordination? | 18:14 |
| flaper87 | tjanczuk: the italian way. Yelling! | 18:14 |
| tjanczuk | I am told there are 3000 attendees. That may not scale. | 18:15 |
| flaper87 | tjanczuk: jokes apart, we did Twitter, sms and email in HKG | 18:15 |
| flaper87 | tjanczuk: where are you flying from? | 18:15 |
| malini | isn't there an app to track? | 18:15 |
| tjanczuk | Seattle | 18:15 |
| flaper87 | I think I'll buy a local sim card | 18:15 |
| flaper87 | malini: that app is for pets, I'm a robot. Come on! | 18:15 |
| malini | so far, we have two ppl flying in on Sunday night -including tjanczuk | 18:15 |
| malini | wonder if we should just do this on a conf day evening | 18:16 |
| *** amitgandhi has joined #openstack-marconi | 18:16 | |
| flaper87 | malini: that may work too | 18:17 |
| flaper87 | however, there are *many* events during the conf week | 18:17 |
| flaper87 | (in the evening) | 18:17 |
| kgriffs | yeah. Well, let's not forget we have a project pod area we can use during the day | 18:17 |
| kgriffs | for unconferency stuff | 18:18 |
| tjanczuk | What's that area? | 18:18 |
| flaper87 | lets just do something on saturday / sunday and use the dev room for uncon | 18:18 |
| flaper87 | tjanczuk: there's normally a dev room in the conf | 18:18 |
| kgriffs | tjanczuk: every project gets a table. they are located in a few rooms at the conference center | 18:18 |
| flaper87 | it's sort of a lounge bar with redbull and no internet access | 18:18 |
| kgriffs | ROFL | 18:19 |
| flaper87 | ok, I'm kidding, there's no redbull | 18:19 |
| flaper87 | nor internet access | 18:19 |
| sriram | no internet access :O | 18:19 |
| flaper87 | :D | 18:19 |
| * flaper87 is full of bad / evil jokes | 18:19 | |
| tjanczuk | Lack of WIFI must be intentional to foster good discussion. | 18:19 |
| tjanczuk | Is this where folks normally hang out during the day? | 18:19 |
| *** balajiiyer has joined #openstack-marconi | 18:20 | |
| kgriffs | yep, people hang out there during breaks, when there aren't any other interesting sessions to go to, or you want to follow up on a topic that doesn't have a time slot | 18:20 |
| kgriffs | if we are lucky they will drop some access points in those rooms so we don't have to contend with the masses for an IP | 18:21 |
| flaper87 | kgriffs: alcabrera malini btw, what's the cheapest / convenient phone company in ATL? | 18:22 |
| flaper87 | I just need the sim card for a week | 18:23 |
| kgriffs | and if all else fails, I heard flaper87 is bringing his personal hotspot and will let everyone use it for free. ;) | 18:23 |
| alcabrera | lol | 18:23 |
| *** jchai has quit IRC | 18:23 | |
| flaper87 | kgriffs: what do you think the sim card is for? | 18:23 |
| alcabrera | flaper87: I've had great experiences with Ting.com on my Nexus 5. I'm looking at about $15/month | 18:23 |
| flaper87 | :P | 18:23 |
| flaper87 | alcabrera: you own a nexus 5? | 18:23 |
| tjanczuk | I think I may be too late to join on Sunday if that is when you decide to meet but please keep me in the loop on that outing. Otherwise I will try to find you in the pod. | 18:23 |
| * flaper87 owns one too | 18:23 | |
| alcabrera | flaper87: oh yes - as of about... a month ago! | 18:23 |
| alcabrera | I purchased the ting.com SIM card for about $10 | 18:24 |
| flaper87 | alcabrera: does buying the card have an extra.... thanks | 18:24 |
| flaper87 | :P | 18:24 |
| alcabrera | lol | 18:24 |
| alcabrera | it also comes with a $25 credit on their service for getting the SIM card, iirc | 18:24 |
| flaper87 | muahahaha, if they only knew I'll throw it away the week after | 18:24 |
| flaper87 | :P | 18:24 |
| flaper87 | alcabrera: do you know if there's a ting.com store at the airport ? | 18:26 |
| alcabrera | flaper87: I'm not sure. I'm inclined to doubt it. >.> | 18:27 |
| flaper87 | If I get off the plain and realize I don't have internet access, I'll probably freak out... | 18:27 |
| flaper87 | I'm a robot, that's where my food comes from | 18:27 |
| flaper87 | anyway, I'll figure something out | 18:27 |
| *** ayoung_rrrrmmm is now known as ayoung | 18:32 | |
| tjanczuk | On a more serious note, I was following the marconi discussions on openstack-dev and this thread in particular was interesing: http://lists.openstack.org/pipermail/openstack-dev/2014-March/030367.html | 18:34 |
| flaper87 | tjanczuk: what are your thougnts about it? | 18:35 |
| flaper87 | tjanczuk: did you go through the whole thread? | 18:35 |
| flaper87 | I'm really interested in knowing what your opinion is and what your doubts, if any, are | 18:36 |
| tjanczuk | Yes I went through the whole thing. By the time I was done I had 5 cups of coffee in me. It looked to me like the concerns expressed by folks in there were preventing Marconi from graduating. | 18:36 |
| tjanczuk | I actually looked at Marconi before this thread showed up (not sure if you recall but I popped in on this IRC some time ago). | 18:37 |
| kgriffs | I think I remember you popping in a while back. You are with HP? | 18:37 |
| tjanczuk | Yes I am with HP. In general I get the value proposition of Marconi: to deliver a *multi-tenant* messaging system (as opposed to a mechanism to provide single-tenant deployments of existing brokers). | 18:38 |
| kgriffs | ah, ok. Yep I remember now seeing you around. | 18:39 |
| flaper87 | tjanczuk: thanks for taking the time for looking at marconi and reading that thread | 18:40 |
| tjanczuk | One piece of feedback from that thread that resonated with me was the unusual (for a messaging system) modes of accessing messages, virtually enabling random access to any messages. | 18:40 |
| tjanczuk | Despite the fact the storage layer is extensible, having this prescriptive API shape at the transport layer in practice constraints what technology one can use at the storage layer. | 18:41 |
| tjanczuk | That combined with the fact there are folks who don't like Mongo for legal reasons do not leave a lot of options on the table. | 18:42 |
| kgriffs | tjanczuk: do you have a specific backend in mind that you would like to use? | 18:43 |
| flaper87 | tjanczuk: that's a valid concern. We're planing to discuss queues flavors at the summit which will bring a way to categorize the kind of storage driver available / enabled | 18:44 |
| tjanczuk | Given that most OpenStack deployments already require a RabbitMQ cluster (and the know-how of operating one), I think it would be beneficial to allow folks to re-use that expertise. In other words, I would see value in putting a multi-tenant HTTP head on top of RabbitMQ. | 18:44 |
| flaper87 | this definitely doesn't fix your concern but it, at least, gives a nice way to organize them | 18:44 |
| flaper87 | as of MongoDB's license issue. I keep hearing that many folks complain about it but I still haven't heard / read a *real* case where someone *can't* use mongodb because of its license | 18:45 |
| flaper87 | I'm not a layer, so bare with me there. | 18:45 |
| kgriffs | as for AMQP, we are discussing ways to make that work | 18:46 |
| kgriffs | we have some ideas for v2.0 API that will make it easier to support an AMQP backend; flaper87 has been thinking a lot about that | 18:46 |
| flaper87 | kgriffs: speaking of which, I've more things to share on that topic but I don't want to hijack the discussion | 18:47 |
| kgriffs | tjanczuk: another direction we are exploring is Redis. | 18:47 |
| tjanczuk | I am not a lawyer either and I am not judging anyone for using or not using Mongo for whatever reason. I am just observing there are folks for whom it is a turn-off. | 18:47 |
| tjanczuk | Redis is interesting. Did you look at Kafka? | 18:48 |
| kgriffs | tjanczuk: yes, it is a turnoff, and it will take time for people to either decide it is OK after all, or find a concrete reason why they don't want to use it. | 18:48 |
| kgriffs | in the meantime, we need to have more options. | 18:48 |
| kgriffs | tjanczuk: I can't speak for others in the team, but I did look at Kafka around when it first came on the scene | 18:49 |
| tjanczuk | I know this may sound bizzare, but I almost wish Marconi did not give people any "options" for storage | 18:49 |
| flaper87 | tjanczuk: I think that would bring more problems that it would solve, TBH. and it wouldn't help with the "You guys are re-inventing the wheel" comments we keep getting. | 18:50 |
| kgriffs | it's a clever system, but we haven't thought seriously about using it because we wanted to support more semantics than kafka would make possible; I should take another look at it now that it has been baking a while. | 18:50 |
| flaper87 | it's not a terrible idea but it has a huge impact in the community | 18:50 |
| tjanczuk | I think some of the best technologies are so because they are opinionated: they do one thing and do it great. This also enables a lot of optimizations opportunity for which is lost once an extensibility point is in the system. | 18:50 |
| kgriffs | swift basically went that route | 18:51 |
| kgriffs | they wrote their own storage layer | 18:51 |
| notmyname | ? | 18:51 |
| * notmyname sees "Swift" | 18:51 | |
| kgriffs | i mean, built on top of sqlite | 18:51 |
| kgriffs | but swift handles all the replication, etc. of the data | 18:51 |
| flaper87 | I think that, as long as the Marconi team, focuses on a small set of technologies, it's fine | 18:52 |
| tjanczuk | flapper87: I think reinventing the wheel is not necessarily a bad thing, as long as you can show it is a better wheel. | 18:52 |
| * kgriffs like this tjanczuk guy | 18:52 | |
| flaper87 | tjanczuk: agreed, but we need to experiment before we can claim that | 18:52 |
| tjanczuk | I think if Marconi could deliver the perf of, say, Rabbit, but solve the multi-tenancy problem, that would be a better wheel for me. | 18:52 |
| kgriffs | tjanczuk: +1, that phrase is often overused and abused | 18:52 |
| flaper87 | I think we'll get to the point where the team focus will be on 2 drivers and the rest will be left out to the community | 18:53 |
| tjanczuk | (That is why I keep pestering you guys about performance number on openstack-dev) | 18:53 |
| flaper87 | tjanczuk: I like that last statement | 18:53 |
| kgriffs | tjanczuk: fwiw, I've been pestering people too; I think we will finally get at least some basic benchmarks done and posted in the next week | 18:54 |
| * kgriffs crosses fingers | 18:54 | |
| malini | tjanczuk: On tht note, we are finally ready to do some serious benchmarking | 18:54 |
| malini | tjanczuk: this is our initial plan https://etherpad.openstack.org/p/marconi-benchmark-plans . Feel free to add your comments | 18:54 |
| malini | The last benchmark we did was 3-4 months back | 18:55 |
| tjanczuk | I will take a look in the pm, I need to pick up my kid from school shortly. | 18:55 |
| kgriffs | anyway, I think at some point it wouldn't hurt to experiment with more of a domain-specific backend storage engine, but I don't think we have the bandwidth to take that on just yet | 18:55 |
| flaper87 | tjanczuk: before you go, thanks a lot for your feedback. it's very welcome and appreciated. Keep it coming :) | 18:56 |
| tjanczuk | I have a concern that with the current architecture of Marconi it will be hard to get decent perf #s, but I would love to be proved wrong. | 18:56 |
| tjanczuk | The reason I mentioned Kafka is that it is interesting not only from the perspective of a potential backend, but also the entire mindset of approaching a problem. | 18:57 |
| * flaper87 read kafka's paper last week | 18:58 | |
| kgriffs | tjanczuk: totally valid concern. We need to think about some concrete goals, and then find a way to accomplish them. | 18:58 |
| tjanczuk | These guys are really looking at the messaging stack holistically. Having a very specific messaging semantics allows them to milk perf to the extend not possible with a more generic messaging interfaces. | 18:58 |
| tjanczuk | OK, you will not get exactly once delivery. Worse, your consumers must keep track of which messages were consumed. But know what, it gives them 10x the perf of other messaging systems. And it very well addresses specific scenarios. | 19:00 |
| tjanczuk | Perhaps it would make sense to apply a similar mindset to Marconi. Build a very tightly integrated stack. Perhaps some minimal extensibiloty points here and there, to plug in custom auth for example. | 19:00 |
| *** amitgandhi has quit IRC | 19:01 | |
| *** amitgandhi has joined #openstack-marconi | 19:01 | |
| tjanczuk | I will go over the test plan for perf. Otherwise I look forward to chatting more before and during the summit. | 19:02 |
| malini | thanks tjanczuk | 19:02 |
| kgriffs | yep, let's keep thinking about this. | 19:02 |
| *** openstackgerrit has quit IRC | 19:04 | |
| kgriffs | It will be helpful to chat in person with pen and paper and/or whiteboard in hand. | 19:05 |
| *** amitgandhi has quit IRC | 19:06 | |
| flaper87 | Although I agree with the feeling about kafka, I think we're still not ready to take such decision. | 19:06 |
| kgriffs | alcabrera, flaper87: do we have any GNOME OPW interns, or are they all GSoC? | 19:06 |
| flaper87 | kgriffs: Aazz is an OPW intern | 19:07 |
| kgriffs | wait, do we have 4 interns now? Or is cpallares going to work on something different? | 19:07 |
| kgriffs | (4 total, across GSoC, OPW, and Red Hat) | 19:07 |
| kgriffs | Victoria, Prashanth are GSoC | 19:08 |
| kgriffs | Cindy is interning directly with Red Hat, right? | 19:08 |
| kgriffs | cpallares: ^^^ ? | 19:10 |
| flaper87 | kgriffs: yup | 19:10 |
| kgriffs | ok | 19:10 |
| flaper87 | she's wearing a very nice Red Hat now :) | 19:10 |
| kgriffs | rock on | 19:10 |
| alcabrera | sweet! | 19:12 |
| alcabrera | I'm very happy for cpallares. :D | 19:12 |
| alcabrera | I didn't know! | 19:12 |
| cpallares | alcabrera: thanks :) | 19:13 |
| *** tjanczuk has quit IRC | 19:13 | |
| * cpallares is annoyed fedora doesn't have popup notifications for irc | 19:13 | |
| alcabrera | heh | 19:14 |
| alcabrera | I'm been using weechat way too long. I have no notifications, short of looking at my terminal window and seeing my name highlighted. >.> | 19:14 |
| cpallares | alcabrera: haha that would not work for me. | 19:15 |
| cpallares | alcabrera: although IRC notifications are rather distracting | 19:15 |
| * cpallares remembers all those convos about chickens and video cameras | 19:15 | |
| alcabrera | hahaha | 19:16 |
| alcabrera | I participated in one discussion recently | 19:16 |
| alcabrera | where there was a mass exchange of adorable bunny pictures | 19:17 |
| alcabrera | cpallares: :) | 19:17 |
| cpallares | alcabrera: :D | 19:17 |
| cpallares | alcabrera: my reaction to anything bunny related: http://x2.fjcdn.com/comments/A+_143cf603e2473c89f40a7af0c58a16e8.jpg | 19:18 |
| * alcabrera clicks | 19:18 | |
| alcabrera | hahaha | 19:18 |
| alcabrera | cpallares: ~so kawaii~ | 19:19 |
| cpallares | falcon has a pretty cute bunny picture. | 19:19 |
| cpallares | with a pancake. | 19:19 |
| cpallares | that's also how I reacted to it | 19:19 |
| alcabrera | cpallares: http://imgur.com/a/zYZvn/ -- unleash the bunnies | 19:19 |
| alcabrera | that was the winning post | 19:20 |
| cpallares | so cute :3 | 19:20 |
| *** jmckind has joined #openstack-marconi | 19:20 | |
| cpallares | alcabrera: This is my devious hamster. He likes to escape, last time that happened he got his head stuck and I took a picture of him to shame him. https://scontent-b-ord.xx.fbcdn.net/hphotos-frc3/t1.0-9/598836_369653763109461_477964547_n.jpg | 19:22 |
| *** tedross has joined #openstack-marconi | 19:26 | |
| *** amitgandhi has joined #openstack-marconi | 19:26 | |
| kgriffs | lol | 19:27 |
| kgriffs | what kind of hamster is he? | 19:27 |
| alcabrera | cpallares: cuuutee! Sneaky little fellow. :) | 19:29 |
| cpallares | kgriffs: he's a dwarf hamster and his name is Gimli :P :P :P | 19:30 |
| * kgriffs gives Gimli a virtual snack | 19:30 | |
| alcabrera | cpallares: does Gimli have an awesome axe? ;) | 19:30 |
| *** ChanServ changes topic to "OpenStack Queuing and Notification Service || Smile :D || Meetings every Tuesday @ 15:00 UTC || Wiki: https://wiki.openstack.org/wiki/Marconi || Paste: http://paste.openstack.org/ || Send messages and make some noise :D" | 19:31 | |
| -openstackstatus- NOTICE: Gerrit upgrade to 2.8 complete. See: https://wiki.openstack.org/wiki/GerritUpgrade Some cleanup tasks still ongoing; join #openstack-infra if you have any questions. | 19:31 | |
| cpallares | alcabrera: why yes he does, it's made of seeds. | 19:31 |
| cpallares | (and straw) | 19:32 |
| cpallares | yay gerrit works now | 19:32 |
| alcabrera | w00t | 19:34 |
| *** chandan_kumar has quit IRC | 19:40 | |
| cpallares | alcabrera: do you know what hadoop is? | 19:55 |
| alcabrera | cpallares: yup! | 19:57 |
| alcabrera | I've never used it, but I'm familiar with the platform. | 19:57 |
| cpallares | alcabrera: ah okay well I have a hadoop sticker, do you want it? I'm sending them your way tomorrow. | 19:58 |
| cpallares | alcabrera: I thought it was a postgresql elephant, but it looks green and on something :P | 19:59 |
| cpallares | but it's cute | 19:59 |
| cpallares | so maybe it could find a home on your laptop | 19:59 |
| alcabrera | oh! | 20:01 |
| alcabrera | cpallares: yay! That's wonderful news. I'm looking forward to the stickers. | 20:01 |
| alcabrera | And yes - I'll take a Hadoop elephant. :) | 20:01 |
| alcabrera | or an elephant | 20:01 |
| cpallares | alcabrera: http://cdn.dice.com/wp-content/uploads/2013/06/elephant_rgb_sq.png | 20:01 |
| alcabrera | the tech-ephanter | 20:01 |
| cpallares | alcabrera: but it's greenish | 20:02 |
| alcabrera | hahaha | 20:02 |
| alcabrera | I love his little face | 20:02 |
| alcabrera | like | 20:02 |
| cpallares | haha | 20:02 |
| alcabrera | "I'm an elephant YEAAAHH" | 20:02 |
| alcabrera | ";D" | 20:02 |
| cpallares | haha | 20:03 |
| *** amitgandhi has quit IRC | 20:08 | |
| *** amitgandhi has joined #openstack-marconi | 20:09 | |
| alcabrera | I'm out for the night. Take care, all! | 20:16 |
| alcabrera | cpallares: o/ | 20:16 |
| *** shakamunyi has joined #openstack-marconi | 20:17 | |
| *** alcabrera is now known as alcabrera|afk | 20:18 | |
| *** sriram has quit IRC | 20:20 | |
| *** mwagner_lap has quit IRC | 20:27 | |
| *** vkmc has quit IRC | 20:43 | |
| *** tedross has quit IRC | 20:47 | |
| *** cpallares has quit IRC | 20:59 | |
| *** jmckind has quit IRC | 21:03 | |
| *** tedross has joined #openstack-marconi | 21:05 | |
| *** Obulpathi has quit IRC | 21:24 | |
| *** amitgandhi has quit IRC | 21:24 | |
| *** malini is now known as malini_afk | 21:32 | |
| *** tjanczuk has joined #openstack-marconi | 21:35 | |
| *** balajiiyer has quit IRC | 21:44 | |
| *** openstackgerrit has joined #openstack-marconi | 21:54 | |
| *** flaper87 is now known as flaper87|afk | 22:18 | |
| *** mwagner_lap has joined #openstack-marconi | 22:19 | |
| *** tedross has quit IRC | 22:24 | |
| *** reed has quit IRC | 22:26 | |
| *** tjanczuk has quit IRC | 22:54 | |
| *** ayoung is now known as ayoung_DadMode | 23:01 | |
| *** ayoung_DadMode is now known as ayoung_ | 23:56 | |
| *** ayoung_ is now known as ayoung | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
