harlowja | dmsimard lol | 01:11 |
---|---|---|
*** sshnaidm has quit IRC | 02:28 | |
*** sshnaidm has joined #ara | 02:41 | |
*** sshnaidm has quit IRC | 03:53 | |
*** sshnaidm has joined #ara | 04:09 | |
*** jrist has quit IRC | 04:33 | |
openstackgerrit | kavitha h r proposed openstack/ara master: Replace six.iteritems() with .items() https://review.openstack.org/482126 | 05:44 |
*** jparrill has joined #ara | 06:11 | |
*** openstackgerrit has quit IRC | 07:33 | |
*** tbielawa has joined #ara | 11:19 | |
*** resmo_ has joined #ara | 11:44 | |
*** tbielawa has quit IRC | 12:01 | |
*** resmo_ has quit IRC | 12:23 | |
*** jparrill has quit IRC | 15:08 | |
*** sshnaidm is now known as sshnaidm|afk | 16:21 | |
*** sshnaidm|afk has quit IRC | 16:58 | |
harlowja | dmsimard what parts of oslo u need? | 17:43 |
harlowja | oslo.utils is generally nice imho | 17:44 |
dmsimard | harlowja: yeah encodingutils from oslo.utils | 17:44 |
harlowja | i'd stay away from some of the larger (oslo.db and oslo.messaging and oslo.config) though | 17:44 |
dmsimard | harlowja: the rest of oslo is fairly incompatible, mostly due to the expectation that we're running oslo.config in standard config files | 17:44 |
harlowja | imho those are 'infective' (for lack of a better word) | 17:44 |
harlowja | right | 17:44 |
harlowja | ya, incompat, infective, ... | 17:44 |
harlowja | aka, drags in more | 17:45 |
dmsimard | harlowja: I'd have to write a oslo.config driver to read from ansible.cfg like ara does or something | 17:45 |
harlowja | do u like the pain | 17:45 |
harlowja | how much u like the pain | 17:45 |
harlowja | lol | 17:45 |
dmsimard | harlowja: It's not something I'm interested in doing | 17:45 |
harlowja | u do not like the pain | 17:45 |
harlowja | lol | 17:45 |
harlowja | u fail | 17:45 |
harlowja | lol | 17:45 |
dmsimard | harlowja: but encodeutils is convenient for py2 vs py3 | 17:45 |
harlowja | ya | 17:45 |
harlowja | u may or may not like https://github.com/openstack/oslo.utils/blob/master/oslo_utils/excutils.py#L35 and a few others in there to | 17:46 |
* harlowja tried to simulate (at least as much as i could) chained exceptions in py3 | 17:46 | |
harlowja | and timeutils has a nice stopwatch that i made, lol | 17:46 |
harlowja | https://github.com/openstack/oslo.utils/blob/master/oslo_utils/timeutils.py#L390 | 17:47 |
harlowja | that's about it, lol | 17:47 |
harlowja | `with timeutils.StopWatch() as watch: | 17:47 |
harlowja | <some code that takes a while> | 17:47 |
harlowja | print watch.elapsed()` | 17:47 |
harlowja | (not super complicated but nice to have ^) | 17:47 |
dmsimard | sure | 17:48 |
dmsimard | anyway.. next ara version is mostly going to be about py3 and ansible 2.4 | 17:48 |
dmsimard | which is not exactly fun work | 17:48 |
harlowja | fun fun | 17:48 |
harlowja | six and u will be BFFs | 17:48 |
harlowja | lol | 17:48 |
dmsimard | yeah I mean it's not too terrible already | 17:48 |
dmsimard | https://review.openstack.org/#/c/481837/ brought down the amount of py3 failing tests from 125 to 29 | 17:49 |
dmsimard | but there's still some annoying encoding things I need to figure out | 17:49 |
dmsimard | my brain has a hard time with encoding and decoding | 17:49 |
harlowja | ya, u just have to accept that there is some unicode type that has no real value until u make it appear (ie decode it) | 17:50 |
harlowja | and if u get binary types u have to encode them into that magic unicode type (that has no value until u decode it) | 17:50 |
harlowja | lol | 17:50 |
harlowja | something like that helps (maybe?) | 17:51 |
harlowja | lol | 17:51 |
dmsimard | I dunno, one day my brain will click | 17:51 |
harlowja | lol | 17:51 |
dmsimard | until then ihavenoidea.gif | 17:51 |
dmsimard | harlowja: problem is also the matter that we store things in the db | 17:55 |
harlowja | ya just store all the things as utf8 | 17:55 |
dmsimard | harlowja: and those things are compressed/decompressed (zlib) on the fly through the sqlalchemy model | 17:55 |
harlowja | or utf32 | 17:55 |
harlowja | ya, weird, don't do that? :-P | 17:56 |
dmsimard | harlowja: everything ends up in the DB, if things wouldn't be compressed the DB would end up huge | 17:56 |
harlowja | huge? | 17:56 |
harlowja | what is huge now-adays :-P | 17:56 |
dmsimard | harlowja: http://logs.bonnyci.org/ara/ | 17:56 |
dmsimard | harlowja: ara is all text, no binary, it compresses super easily | 17:57 |
harlowja | ya, i was more wondering why sqlalchemy is doing the compression vs say the DB itself | 17:57 |
harlowja | (just feels odd, but idk) | 17:57 |
dmsimard | harlowja: https://github.com/openstack/ara/blob/master/ara/models.py#L135 | 17:57 |
harlowja | i gotta ask clint where http://logs.bonnyci.org/ara/ is being used | 17:57 |
harlowja | i didn't think anyone was using bonnyci yet | 17:57 |
dmsimard | harlowja: https://github.com/openstack/ara/blob/master/ara/models.py#L229-L239 | 17:57 |
harlowja | ya, weird | 17:58 |
dmsimard | harlowja: I guess once ARA has an API we can take that out and abstract it there instead if that's what you mean | 17:58 |
harlowja | i thought the DB itself could do that :-/ | 17:58 |
harlowja | though meh, storage is cheap, whoopie | 17:58 |
harlowja | unless its TB sizes, then maybe not so small | 17:59 |
dmsimard | harlowja: come on, compressing is free :P | 17:59 |
harlowja | ya, idk, i just thought innodb could do that automatically | 18:00 |
harlowja | https://dev.mysql.com/doc/refman/5.5/en/innodb-compression-usage.html | 18:00 |
harlowja | (or other?) | 18:00 |
* harlowja not DB expert, lol | 18:00 | |
dmsimard | harlowja: innodb is not particularly reldb agnostic | 18:02 |
dmsimard | harlowja: unless it's abstracted by sqlalchemy maybe | 18:02 |
harlowja | i would expect so, not sure | 18:02 |
dmsimard | I heard about someone sending ara data to oracle so.. | 18:03 |
harlowja | in a dream? | 18:03 |
dmsimard | nope, 4real | 18:03 |
harlowja | prob in a dream then | 18:03 |
harlowja | lol | 18:03 |
dmsimard | harlowja: exhibit number one: http://eavesdrop.openstack.org/irclogs/%23ara/%23ara.2017-03-04.log.html#t2017-03-04T17:08:35 | 18:04 |
dmsimard | exhibit number two: http://eavesdrop.openstack.org/irclogs/%23ara/%23ara.2017-03-05.log.html | 18:04 |
harlowja | prob another person sleep walking to there computer and typing that | 18:05 |
harlowja | lol | 18:05 |
harlowja | in there dream | 18:05 |
harlowja | lol | 18:05 |
* harlowja always wanted to try http://docs.peewee-orm.com/en/latest/ | 18:08 | |
weshay | FYI https://review.openstack.org/#/q/topic:ara | 18:36 |
dmsimard | weshay: yay | 18:38 |
dmsimard | weshay: gotta start somewhere :) | 18:38 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!