| *** amrith has joined #openstack-trove | 00:00 | |
| *** flaper87 is now known as flaper87|afk | 00:07 | |
| *** jcru has quit IRC | 00:08 | |
| *** yidclare has quit IRC | 00:27 | |
| *** cweid has quit IRC | 00:37 | |
| *** plodronio has joined #openstack-trove | 00:39 | |
| *** plodronio has left #openstack-trove | 00:40 | |
| *** yidclare has joined #openstack-trove | 00:40 | |
| *** tanisdl has quit IRC | 00:41 | |
| *** grapex has quit IRC | 00:57 | |
| *** yidclare has quit IRC | 00:57 | |
| *** edmund has quit IRC | 01:00 | |
| *** achampion has joined #openstack-trove | 01:18 | |
| *** ViswaV has quit IRC | 01:32 | |
| *** nosnos has joined #openstack-trove | 01:39 | |
| *** demorris has joined #openstack-trove | 01:44 | |
| *** demorris has quit IRC | 02:11 | |
| *** demorris has joined #openstack-trove | 02:12 | |
| *** grapex has joined #openstack-trove | 02:12 | |
| *** demorris has quit IRC | 02:12 | |
| *** grapex has joined #openstack-trove | 02:12 | |
| *** erkules_ has joined #openstack-trove | 02:23 | |
| *** michael-yu has joined #openstack-trove | 02:24 | |
| *** mattgriffin has quit IRC | 02:24 | |
| *** erkules has quit IRC | 02:26 | |
| *** rongze has joined #openstack-trove | 02:28 | |
| *** robertmyers has joined #openstack-trove | 02:39 | |
| *** krow has quit IRC | 02:40 | |
| *** krow has joined #openstack-trove | 02:45 | |
| *** krow has quit IRC | 02:50 | |
| *** krow has joined #openstack-trove | 02:50 | |
| *** rongze has quit IRC | 02:52 | |
| *** jasonb365 has joined #openstack-trove | 02:54 | |
| *** krow has quit IRC | 02:55 | |
| *** krow has joined #openstack-trove | 02:56 | |
| *** khyati_ has quit IRC | 03:01 | |
| *** michael-yu has quit IRC | 03:05 | |
| *** michael-yu has joined #openstack-trove | 03:08 | |
| *** SushilKM has quit IRC | 03:13 | |
| *** krow has quit IRC | 03:20 | |
| *** krow has joined #openstack-trove | 03:25 | |
| *** jcooley_ has joined #openstack-trove | 03:26 | |
| *** krow1 has joined #openstack-trove | 03:29 | |
| *** krow has quit IRC | 03:30 | |
| *** jasonb365 has quit IRC | 03:32 | |
| *** michael-yu has quit IRC | 03:39 | |
| *** krow1 has quit IRC | 03:41 | |
| *** krow has joined #openstack-trove | 03:46 | |
| *** flaper87|afk has quit IRC | 03:48 | |
| *** flaper87|afk has joined #openstack-trove | 03:50 | |
| *** flaper87|afk is now known as flaper87 | 03:51 | |
| *** krow has quit IRC | 03:51 | |
| *** mattgriffin has joined #openstack-trove | 03:53 | |
| *** edmund has joined #openstack-trove | 03:59 | |
| *** robertmyers has quit IRC | 04:03 | |
| *** krow has joined #openstack-trove | 04:05 | |
| *** mattgriffin has quit IRC | 04:07 | |
| *** mattgriffin has joined #openstack-trove | 04:08 | |
| *** jasonb365 has joined #openstack-trove | 04:08 | |
| *** amcrn_ has quit IRC | 04:08 | |
| *** amcrn has joined #openstack-trove | 04:12 | |
| *** jcooley_ has quit IRC | 04:24 | |
| *** harlowja_ is now known as harlowja_away | 04:31 | |
| *** amcrn has quit IRC | 04:37 | |
| *** amcrn has joined #openstack-trove | 04:38 | |
| *** saju_m has joined #openstack-trove | 04:39 | |
| *** krow has quit IRC | 04:39 | |
| *** krow has joined #openstack-trove | 04:44 | |
| *** radez is now known as radez_g0n3 | 04:48 | |
| *** krow has quit IRC | 04:48 | |
| *** krow has joined #openstack-trove | 04:49 | |
| *** demorris has joined #openstack-trove | 04:50 | |
| *** jasonb365 has quit IRC | 04:53 | |
| *** SushilKM has joined #openstack-trove | 05:03 | |
| *** jcooley_ has joined #openstack-trove | 05:13 | |
| *** demorris has quit IRC | 05:13 | |
| *** krow1 has joined #openstack-trove | 05:20 | |
| *** krow has quit IRC | 05:21 | |
| *** SushilKM has quit IRC | 05:22 | |
| *** krow1 has quit IRC | 05:25 | |
| *** krow has joined #openstack-trove | 05:26 | |
| *** krow1 has joined #openstack-trove | 05:29 | |
| *** krow has quit IRC | 05:31 | |
| *** mattgriffin has quit IRC | 05:51 | |
| *** SushilKM has joined #openstack-trove | 05:54 | |
| *** mattgriffin has joined #openstack-trove | 05:57 | |
| *** saju_m has quit IRC | 05:58 | |
| *** krow1 has quit IRC | 06:00 | |
| *** krow has joined #openstack-trove | 06:07 | |
| *** sballe has quit IRC | 06:11 | |
| *** jcooley_ has quit IRC | 06:23 | |
| *** yogesh has joined #openstack-trove | 06:33 | |
| *** grapex has quit IRC | 06:35 | |
| *** mattgriffin has quit IRC | 06:39 | |
| *** michael-yu has joined #openstack-trove | 06:44 | |
| *** mattgriffin has joined #openstack-trove | 06:45 | |
| *** michael-yu has quit IRC | 06:49 | |
| *** michael-yu has joined #openstack-trove | 06:49 | |
| *** yogesh_ has joined #openstack-trove | 06:49 | |
| *** yogesh has quit IRC | 06:52 | |
| *** yogesh_ has quit IRC | 06:54 | |
| *** grapex has joined #openstack-trove | 07:06 | |
| shivamshukla | s | 07:07 |
|---|---|---|
| *** krow has quit IRC | 07:08 | |
| *** shivamshukla has quit IRC | 07:08 | |
| *** grapex has quit IRC | 07:15 | |
| *** sballe has joined #openstack-trove | 07:24 | |
| *** saju_m has joined #openstack-trove | 07:25 | |
| *** denis_makogon has joined #openstack-trove | 07:33 | |
| *** narmitag has quit IRC | 07:38 | |
| *** grapex has joined #openstack-trove | 07:41 | |
| *** grapex has quit IRC | 07:45 | |
| *** yogesh has joined #openstack-trove | 07:54 | |
| *** michael-yu has quit IRC | 07:54 | |
| *** yogesh_ has joined #openstack-trove | 08:02 | |
| *** tattabbum has joined #openstack-trove | 08:04 | |
| *** yogesh has quit IRC | 08:04 | |
| tattabbum | Hi all, I'm trying to install manually TROVE following this unique guide: http://docs.openstack.org/developer/trove/dev/manual_install.html | 08:07 |
| tattabbum | When I run this command: #trove-manage --config-file=<PathToTroveConf> db_wipe trove_test.sqlite mysql fake | 08:08 |
| tattabbum | (where --config-file=/root/trove/etc/trove/trove.conf.sample) | 08:08 |
| tattabbum | I obtain this output: usage: trove-manage [-h] [--config-dir DIR] [--config-file PATH] [--debug] [--log-config-append PATH] [--log-date-format DATE_FORMAT] [--log-dir LOG_DIR] [--log-file PATH] [--log-format FORMAT] [--nodebug] [--nouse-syslog] [--noverbose] [--s | 08:09 |
| tattabbum | {db_sync,db_upgrade,db_downgrade,datastore_update,datastore_version_update,db_wipe} | 08:10 |
| tattabbum | trove-manage: error: unrecognized arguments: mysql fake | 08:10 |
| tattabbum | So, if trove-manage doesn't manage 'mysql' and 'fake' arguments, there is something of wrong in the guide. | 08:12 |
| *** flaper87 has quit IRC | 08:18 | |
| *** flaper87 has joined #openstack-trove | 08:18 | |
| *** tattabbum has quit IRC | 08:41 | |
| *** grapex has joined #openstack-trove | 08:41 | |
| *** tattabum has joined #openstack-trove | 08:42 | |
| *** grapex has quit IRC | 08:46 | |
| *** tattabum has quit IRC | 08:48 | |
| *** SushilKM has quit IRC | 08:55 | |
| *** tattabbum has joined #openstack-trove | 08:58 | |
| tattabbum | Can someone suggest to me a valid guide to install and configure correctly Trove? | 08:58 |
| tattabbum | This guide (http://docs.openstack.org/developer/trove/dev/manual_install.html#) seems to be wrong :-( | 09:04 |
| denis_makogon | tattabbum, https://github.com/openstack/trove-integration here you can see trove manuals | 09:31 |
| denis_makogon | tattabbum, about that manual, its been outdated since big changes in trove codebase | 09:31 |
| *** amcrn has quit IRC | 09:34 | |
| denis_makogon | tattabbum, if you would need any help, ping me and i'll try to help you | 09:34 |
| tattabbum | denis_makogon, is this guide (https://github.com/openstack/trove-integration) useful to install TROVE in a DEVSTACK environment? | 09:35 |
| denis_makogon | follow guide at trove-integration page | 09:35 |
| denis_makogon | spin up the VM | 09:36 |
| denis_makogon | add user ubuntu | 09:36 |
| denis_makogon | add to visudo | 09:36 |
| tattabbum | denis_makogon, I would install TROVE in an existing OpenStack (HAVANA) environment.. | 09:36 |
| denis_makogon | log in as ubuntu user | 09:36 |
| denis_makogon | yes | 09:36 |
| denis_makogon | i get it | 09:37 |
| tattabbum | ok... | 09:37 |
| denis_makogon | tattabbum, https://github.com/openstack-dev/devstack/blob/master/lib/trove | 09:37 |
| denis_makogon | there you would find required parameters for each Trove service to integrate it into env | 09:38 |
| denis_makogon | but, if you are using clean devstack installation | 09:38 |
| denis_makogon | i would suggest you to use trove-integration manual | 09:39 |
| tattabbum | I have an Openstack environment in which I have installed every single service manually, by means of this guide http://docs.openstack.org/trunk/install-guide/install/apt/content/ | 09:41 |
| tattabbum | And now I want to add the service TROVE | 09:41 |
| *** grapex has joined #openstack-trove | 09:42 | |
| tattabbum | denis_makogon, I have an Openstack environment in which I have installed every single service manually, by means of this guide http://docs.openstack.org/trunk/install-guide/install/apt/content/ | 09:45 |
| tattabbum | denis_makogon, and now I want to add the service TROVE | 09:45 |
| tattabbum | So, do you say to use this (https://github.com/openstack/trove-integration) guide? | 09:46 |
| *** grapex has quit IRC | 09:47 | |
| denis_makogon | yes | 09:47 |
| denis_makogon | its easiest way | 09:47 |
| tattabbum | denis_makogon, but that guide installs all services other than TROVE! | 09:48 |
| tattabbum | *other than --> in addition to | 09:50 |
| *** saju_m has quit IRC | 09:51 | |
| denis_makogon | it installs devstack and integrates trove into it | 09:51 |
| denis_makogon | but if you need just trove | 09:52 |
| denis_makogon | i would suggest to install trove virtualenv, add required parameters to trove-taskmanager.conf, trove.conf, trove-conductor.conf, trove-guestagent.conf | 09:53 |
| tattabbum | denis_makogon, using the trove virtualenv guide (http://docs.openstack.org/developer/trove/dev/manual_install.html#) I obtain errors when I try to run the trove-manage command: | 09:54 |
| tattabbum | in these steps: http://docs.openstack.org/developer/trove/dev/manual_install.html#prepare-database | 09:55 |
| denis_makogon | trove-manage until was udated | 09:55 |
| denis_makogon | *updated | 09:55 |
| *** ikhudoshyn has joined #openstack-trove | 09:55 | |
| denis_makogon | take a look at redstack script at trove-integration | 09:56 |
| denis_makogon | and search for rd_manager | 09:56 |
| denis_makogon | then you'll see how to work with trove-manage util | 09:56 |
| tattabbum | denis_makogon, OK. However here (http://www.gossamer-threads.com/lists/openstack/dev/35860?page=last) you can see an emails that I have posted related to this problem | 10:00 |
| *** denis_makogon has quit IRC | 10:03 | |
| *** nosnos has quit IRC | 10:21 | |
| *** yogesh_ has quit IRC | 10:26 | |
| *** yogesh has joined #openstack-trove | 10:27 | |
| *** yogesh has quit IRC | 10:31 | |
| *** grapex has joined #openstack-trove | 10:42 | |
| *** grapex has quit IRC | 10:48 | |
| *** dmakogon_ is now known as denis_makogon | 10:51 | |
| tattabbum | 10:53 | |
| denis_makogon | tattabbum, i saw your mail, i'll take a look at manual and i'll try to update in asap | 10:57 |
| tattabbum | Wow! Thank you denis! | 11:00 |
| tattabbum | I suggest to say in the new manual that the virtualenv installation is optional | 11:01 |
| tattabbum | denis_makogon, can you send me an email when you will have finished the manual review? | 11:04 |
| denis_makogon | tattabbum, yes, of course | 11:06 |
| *** Barker has quit IRC | 11:40 | |
| *** grapex has joined #openstack-trove | 11:43 | |
| *** grapex has quit IRC | 11:48 | |
| *** radez_g0n3 is now known as radez | 12:41 | |
| *** Barker has joined #openstack-trove | 12:43 | |
| *** grapex has joined #openstack-trove | 12:43 | |
| *** grapex has quit IRC | 12:48 | |
| *** Barker has quit IRC | 12:49 | |
| *** pdmars has joined #openstack-trove | 13:16 | |
| openstackgerrit | Denis M. proposed a change to openstack/trove: WIP: Cassandra backup/restore https://review.openstack.org/73365 | 13:27 |
| *** radez is now known as radez_g0n3 | 13:36 | |
| *** grapex has joined #openstack-trove | 13:44 | |
| *** grapex has quit IRC | 13:48 | |
| *** achampion has quit IRC | 13:58 | |
| *** jimbobhickville has joined #openstack-trove | 14:01 | |
| *** demorris has joined #openstack-trove | 14:15 | |
| *** kevinconway has joined #openstack-trove | 14:20 | |
| *** radez_g0n3 is now known as radez | 14:20 | |
| *** jcru has joined #openstack-trove | 14:26 | |
| denis_makogon | Gate is broken due to broken Cinder inside the DevStack | 14:38 |
| denis_makogon | https://gist.github.com/denismakogon/9002108 | 14:41 |
| *** michael-yu has joined #openstack-trove | 14:43 | |
| *** achampion has joined #openstack-trove | 14:48 | |
| *** grapex has joined #openstack-trove | 14:56 | |
| *** edmund has quit IRC | 14:58 | |
| *** citizengrey has joined #openstack-trove | 15:01 | |
| *** michael-yu has quit IRC | 15:04 | |
| *** erkules_ is now known as erkules | 15:14 | |
| *** grapex has quit IRC | 15:18 | |
| *** grapex has joined #openstack-trove | 15:18 | |
| openstackgerrit | Denis M. proposed a change to openstack/trove: Security groups workflow update https://review.openstack.org/50944 | 15:19 |
| *** robertmyers has joined #openstack-trove | 15:24 | |
| *** jmontemayor has joined #openstack-trove | 15:24 | |
| *** jmontemayor has quit IRC | 15:25 | |
| *** jmontemayor has joined #openstack-trove | 15:26 | |
| *** Ranjitha has quit IRC | 15:30 | |
| *** datsun180b has joined #openstack-trove | 15:34 | |
| *** amytron has joined #openstack-trove | 15:42 | |
| hub_cap | hey tattabbum, guiseppe, ya? | 15:43 |
| demorris | hub_cap: did you see my question last night about python-troveclient pypi? | 15:43 |
| hub_cap | tattabbum: are you trying to develop against an existing cluster, or are you trying to install for productionish use | 15:43 |
| hub_cap | demorris: you didnt post anythign to this channel | 15:44 |
| demorris | wondering how ofter that gets updated. Seems like the last time it was updated was in Oct of 2013 | 15:44 |
| demorris | yeah, it was ST message cause I did not see you in the channel | 15:44 |
| hub_cap | we will push an update for icehouse | 15:44 |
| hub_cap | im always in the channel :) | 15:44 |
| hub_cap | just not here sometimes hehe | 15:44 |
| demorris | well…not always, but close | 15:45 |
| hub_cap | so ya, the openstack's clis are not updated a ton | 15:45 |
| hub_cap | only if there is a problem | 15:45 |
| hub_cap | but again, they have full features in a v2 api :) | 15:45 |
| *** tattabbum has quit IRC | 15:45 | |
| *** kevinconway has quit IRC | 15:45 | |
| hub_cap | we will likely push 1x or 2x for icehouse and we can chat about the relesae cadance for that at the mid cycle sprint nxt wk | 15:45 |
| demorris | right, I was just trying to get access to some of the types and versions stuff, I can just pull from github though | 15:46 |
| demorris | for the latest | 15:46 |
| *** ViswaV has joined #openstack-trove | 15:47 | |
| hub_cap | yea id _always_ recommend a git clone / sudo python setup.py develop/install (whichever u want) | 15:47 |
| *** ViswaV_ has joined #openstack-trove | 15:50 | |
| *** ViswaV has quit IRC | 15:52 | |
| *** thedodd has joined #openstack-trove | 15:56 | |
| *** ViswaV_ is now known as ViswaV | 15:57 | |
| *** edmund has joined #openstack-trove | 16:00 | |
| *** kevinconway has joined #openstack-trove | 16:03 | |
| *** ikhudoshyn has left #openstack-trove | 16:03 | |
| *** dukhlov__ has quit IRC | 16:07 | |
| denis_makogon | hub_cap, nice to see you here | 16:08 |
| denis_makogon | hub_cap, gate is failing due to cinder stuff | 16:09 |
| denis_makogon | hub_cap, and could you please take a look at mongo patches | 16:10 |
| hub_cap | denis_makogon: im not here yet really :) | 16:11 |
| hub_cap | still makin coffee | 16:11 |
| denis_makogon | hub_cap, ok | 16:11 |
| hub_cap | :) | 16:12 |
| denis_makogon | i'm gonna be out for a while, ping me, please if you'll need me | 16:12 |
| *** jasonb365 has joined #openstack-trove | 16:12 | |
| denis_makogon | btw, do you know normal (free or paid) IRC client for an iPad ? | 16:12 |
| datsun180b | denis_makogon: ssh to a decent machine running irssi in a screen session | 16:15 |
| *** russellb is now known as rustlebee | 16:16 | |
| hub_cap | denis_makogon: haha ssh on ipad?!?!? | 16:18 |
| hub_cap | only konetzed does that | 16:18 |
| hub_cap | id try colloquy or linkinus denis_makogon | 16:18 |
| hub_cap | u will have to do sasl | 16:18 |
| *** jcru has quit IRC | 16:21 | |
| *** krow has joined #openstack-trove | 16:21 | |
| *** jcru has joined #openstack-trove | 16:22 | |
| *** demorris_ has joined #openstack-trove | 16:31 | |
| *** demorris has quit IRC | 16:33 | |
| *** demorris_ is now known as demorris | 16:33 | |
| *** mrsnivvel has quit IRC | 16:34 | |
| *** tanisdl has joined #openstack-trove | 16:36 | |
| *** jasonb365_ has joined #openstack-trove | 16:37 | |
| *** jasonb365 has quit IRC | 16:37 | |
| *** jasonb365_ is now known as jasonb365 | 16:37 | |
| cp16net | datsun180b: are you looking at my setup? | 16:38 |
| ViswaV | Good morning all. Is there a sure fire way of getting the trove guest agent's root path (wherever it is installed on the guest VM….eg /home/<user>/trove) programmatically from within , say a method inside manager.py or backup_impl.py ? | 16:39 |
| datsun180b | cp16net: i'd do it too if i wasn't lazy | 16:39 |
| cp16net | lol | 16:39 |
| cp16net | its easy as well | 16:39 |
| *** tanisdl has quit IRC | 16:39 | |
| cp16net | i type ssh irc and then screen -x | 16:39 |
| datsun180b | cp16net: isn't configuring irssi like a huge pain? i remember it was when i was an intern that once i had it done i never wanted to do it again | 16:40 |
| cp16net | nope | 16:40 |
| imsplitbit | nah it's super easy | 16:40 |
| cp16net | datsun180b: hub_cap passed me his config and i just roll with it | 16:40 |
| cp16net | there are just a few things to configure | 16:40 |
| cp16net | bring up a small cs and install it | 16:40 |
| datsun180b | ugh | 16:40 |
| *** tanisdl has joined #openstack-trove | 16:41 | |
| ViswaV | I was planning on using "/home/"+getpass.getuser()+"/trove" …… Is there a better way? | 16:41 |
| datsun180b | but what about the learning experience? | 16:41 |
| datsun180b | ViswaV: does ~ not expand properly? | 16:41 |
| hub_cap | datsun180b: u will have enough to learn to use it well :) | 16:41 |
| hub_cap | im running mine on a 512m server | 16:42 |
| hub_cap | and its more than enough | 16:42 |
| ViswaV | datsun180b: Have to give that a try. | 16:42 |
| datsun180b | hub_cap: says the guy who just snagged all of tim pope's vim configs before you started | 16:42 |
| kevinconway | datsun180b: ViswaV: http://docs.python.org/2/library/os.path.html#os.path.expanduser | 16:43 |
| *** michael-yu has joined #openstack-trove | 16:43 | |
| cp16net | lolz | 16:43 |
| datsun180b | wait, do we install the guest agent in anyone's home? i thought it was /trove/bin/ | 16:44 |
| datsun180b | or something like that | 16:44 |
| kevinconway | tilde expansion should give you the home of the user that ran the agent | 16:44 |
| ViswaV | kevinconway: thx. | 16:45 |
| datsun180b | but he's looking for the ga's root path, not man-about-town's home dir | 16:45 |
| datsun180b | or so i thought | 16:45 |
| ViswaV | datsun180b: redstack still seems to install it on /home/<user> …. Has this changed recently? | 16:45 |
| hub_cap | i think u mean says the guy who traded in tpopes stuff for some emacs-fu | 16:45 |
| hub_cap | datsun180b: ^ ^ | 16:45 |
| datsun180b | hub_cap: you did both, i'm surprised tpope didn't have you excommunicated | 16:45 |
| hub_cap | and hey, the way i figure it, is if i used to host a ruby meetup that tpope came to, i could steal his configs :) | 16:45 |
| datsun180b | ViswaV: you're probably right if you've got it running right in front of you | 16:46 |
| kevinconway | if you want the path to the agent then you should be able to use some combination of os.realpath(__file__) and os.dirname or os.split | 16:46 |
| kevinconway | should get you close to the path of the file from which the code is being called | 16:46 |
| datsun180b | see i keep trying to go to bash and i'm thinking "which nova-guest" would suffice | 16:46 |
| cp16net | it rsyncs the trove driectory to the home folder in the guest instance | 16:47 |
| cp16net | part of the firstboot script before apt-get update and starting the service | 16:47 |
| datsun180b | yeah i'm thinking of my dev vm, not an instance itself | 16:47 |
| cp16net | oh yeah with your vagrant script. :) | 16:47 |
| ViswaV | kevinconway: I think the os.realpath(__file__) is probably more reliable …and agnostic of where trove guest agent will be installed (home dir or /trove or whatever). I am leaning towards that too. | 16:48 |
| datsun180b | hey vagrant's save me boats of time | 16:48 |
| kevinconway | ViswaV: what are you doing that you need the location of the agent vs the location of a datastore? | 16:49 |
| cp16net | ViswaV: hmm why do you need the path? i might have missed that part of this convo? | 16:49 |
| datsun180b | kevinconway ViswaV yeah i was trying to word that question without sounding like a detective | 16:49 |
| ViswaV | kevinconway: was gonna just explain that .. :) here it is.. | 16:49 |
| ViswaV | I need to get to a file/script that will be bundled within the trove guestagent code. Specifically a wrapper backup/restore script I am creating for mongo backup/restore ... | 16:50 |
| ViswaV | the default mongodump, mongorestore do not work with streams properly… and the swift backup/restore strategies seem to work nicely with stream…at least thats how I see mysql backup/restore implemented | 16:50 |
| cp16net | oh so you want to call a bash script for example to run your backup | 16:51 |
| ViswaV | ye | 16:51 |
| cp16net | gotcha | 16:51 |
| datsun180b | neat | 16:51 |
| kevinconway | i think you could do that with a custom backup manager | 16:51 |
| kevinconway | they should be interchangeable in the config for the agent | 16:51 |
| cp16net | umm i think robertmyers did something similar with the mysl backup | 16:51 |
| ViswaV | kevinconway: it is a custom backup manager… just like the mysql_impl.py …. But eventually it has to call mongodump ….and not directly since mongodump has some deficiencies around working with streams... | 16:52 |
| robertmyers | ViswaV: you can create the backup then tar -czf the results which can be streams | 16:52 |
| ViswaV | It's a bit more than tar -czf | 16:53 |
| robertmyers | well, yes | 16:53 |
| robertmyers | just saying tar can do streaming | 16:53 |
| kevinconway | ViswaV: could you not simply implement this property: https://github.com/openstack/trove/blob/master/trove/guestagent/strategies/backup/mysql_impl.py#L30 | 16:54 |
| *** saju_m has joined #openstack-trove | 16:54 | |
| ViswaV | Yes.. that is the plan. I want to be able to set that cmd to "sudo <path to my custom script> <options> " where the custom script is something like | 16:55 |
| ViswaV | https://gist.github.com/vvutharkar/9004711 | 16:55 |
| kevinconway | ok cool. i guess i just misunderstood something | 16:56 |
| kevinconway | it sounded like you wanted to do some magic loading of something using the agent path | 16:56 |
| robertmyers | ViswaV: you could still do most of that in python | 16:56 |
| robertmyers | basically you need a pre_dump cmd | 16:57 |
| robertmyers | and put that logic there | 16:57 |
| *** michael-yu has quit IRC | 16:57 | |
| robertmyers | then the 'dump' command will be tar | 16:57 |
| robertmyers | post_dump can be the cleanup | 16:57 |
| ViswaV | robertmyers: That is definitely worth exploring, if I can get away from embedding a wrapper script. Thx. Will think in that angle. I did notice that the base backup class does not have pre_backup & post_backup methods like the base restore has pre and post restore methods which allow to do some prep and cleanup … I can explore adding those to the base class. | 16:59 |
| robertmyers | see the restore classes for examples | 16:59 |
| ViswaV | Yup | 16:59 |
| robertmyers | yeah, that would be good to add those hooks | 16:59 |
| robertmyers | for exactly this reason | 16:59 |
| ViswaV | Yes. | 17:00 |
| robertmyers | basically the run method should call pre, dump, post | 17:01 |
| robertmyers | where by default pre and post are pass | 17:01 |
| ViswaV | exactly what i was thinking… and the mongo_impl can override those to do what I plan to do in that wrapper script | 17:02 |
| robertmyers | cool | 17:02 |
| *** jasonb365_ has joined #openstack-trove | 17:03 | |
| ViswaV | Thanks for all the pointers, guys! Off to a good start today morning ! Hopefully I can wrap up the day faster and go grab some flowers/chocolates for my wife :) | 17:03 |
| robertmyers | nice | 17:04 |
| *** jasonb365 has quit IRC | 17:04 | |
| *** jasonb365_ is now known as jasonb365 | 17:04 | |
| robertmyers | http://www.nbc.com/saturday-night-live/video/valentines-day-commercial/n45814 | 17:05 |
| *** datsun180b has quit IRC | 17:09 | |
| *** citizengrey has quit IRC | 17:09 | |
| cp16net | robertmyers: hahaha yeah its pretty funny | 17:09 |
| ViswaV | robertmyers: LOL ! | 17:10 |
| robertmyers | I have to run ;) ... to cvs guys | 17:10 |
| *** datsun180b has joined #openstack-trove | 17:10 | |
| cp16net | yeah i can stop by on the way home :_P | 17:10 |
| *** citizengrey has joined #openstack-trove | 17:14 | |
| ViswaV | If only it were ok to send a lovely valentine message to wife in a gist rather than a hallmark card, i would save enough be able to buy all those items from CVS ! | 17:14 |
| cp16net | lolz | 17:15 |
| datsun180b | my wife and i observe holidays like valentine's on a one-week offset either way so restaurants are easier to get into | 17:15 |
| ViswaV | nice strategy! | 17:16 |
| ViswaV | but how did you convince your wife? ! | 17:16 |
| datsun180b | we are both practical and analytical people to a fault | 17:18 |
| jimbobhickville | I have a similar situation, although not one-week, just a day somewhere nearby that is more convenient for us. | 17:18 |
| robertmyers | well, all the cool kids are heading out tonight | 17:19 |
| * robertmyers is cool | 17:19 | |
| hub_cap | hah my wife said to me | 17:26 |
| hub_cap | do NOT get me flowers | 17:26 |
| hub_cap | cuz i do like every other year, and she explicitly told me this yer not to ;) | 17:26 |
| hub_cap | my valentise day? packing for austin | 17:26 |
| *** amytron has quit IRC | 17:30 | |
| kevinconway | hub_cap: it's a trap | 17:31 |
| kevinconway | she totally wants flowers | 17:31 |
| *** shivamshukla_ has joined #openstack-trove | 17:37 | |
| ViswaV | (https://github.com/openstack/trove/blob/master/trove/guestagent/backup/backupagent.py#L160 versus https://github.com/openstack/trove/blob/master/trove/guestagent/backup/backupagent.py#L160 ) , The 'storage' strategy passed is into restore_runner, which keeps the implementation much cleaner in my opinion and the logic flow contained is within restore_runner class which allows for clean sequencing of pre_restore, restore & | 17:45 |
| ViswaV | post_restore. | 17:45 |
| ViswaV | Whereas in the case of backup_runner the concerns of storage leak into the backup agent class, so the runner is not completely in control of sequencing pre_backup, backup, post_backup . | 17:46 |
| hub_cap | kevinconway: hehe, shes not in to vday :) | 17:46 |
| hub_cap | never has been, corporate holiday and all that | 17:47 |
| ViswaV | (is passed .. is contained …. mistyped the words) | 17:47 |
| ViswaV | There is no pre_Backup, post_backup yet…but I plan on implementing those since mongo might require those. | 17:48 |
| ViswaV | The way backupagent.py logic intertwines with the backup strategy base class (and impl class), it is a bit messy to find the correct insertion points for pre_backup() and post_backup()…. restore logic is much cleaner. | 17:49 |
| *** michael-yu has joined #openstack-trove | 17:52 | |
| *** yidclare has joined #openstack-trove | 17:55 | |
| *** datsun180b has quit IRC | 18:04 | |
| *** yidclare has quit IRC | 18:06 | |
| ViswaV | quick question guys: May be I am not looking in the right place… but I don't see a trove CLI for restore. | 18:08 |
| ViswaV | trove --help does not show any restore command… | 18:09 |
| *** michael-yu has quit IRC | 18:11 | |
| *** shivamshukla_ has quit IRC | 18:12 | |
| *** michael-yu has joined #openstack-trove | 18:15 | |
| *** SushilKM has joined #openstack-trove | 18:27 | |
| cp16net | ViswaV: restore requires a new instance to be created | 18:31 |
| cp16net | so its under create and there is an option to pass the backup_id | 18:32 |
| ViswaV | cp16net: thx. hub_cap pointed me to the same. | 18:33 |
| *** yidclare has joined #openstack-trove | 18:41 | |
| *** thedodd has quit IRC | 18:46 | |
| cp16net | np | 18:48 |
| *** michael-yu has quit IRC | 18:53 | |
| *** yogesh has joined #openstack-trove | 18:53 | |
| *** arborism has joined #openstack-trove | 19:06 | |
| *** michael-yu has joined #openstack-trove | 19:08 | |
| openstackgerrit | Greg Hill proposed a change to openstack/python-troveclient: add scheduled tasks https://review.openstack.org/73693 | 19:08 |
| *** saju_m has quit IRC | 19:11 | |
| jimbobhickville | anyone else get a problem with config.template blowing up a rebase? | 19:11 |
| *** krow has quit IRC | 19:12 | |
| jimbobhickville | it's odd because I didn't modify that file at all locally | 19:12 |
| *** SushilKM has quit IRC | 19:13 | |
| *** datsun180b has joined #openstack-trove | 19:13 | |
| *** krow has joined #openstack-trove | 19:14 | |
| jimbobhickville | something with "Externalization of heat template" - denis_makogon, that was you right? | 19:18 |
| *** amcrn_ has joined #openstack-trove | 19:18 | |
| *** amcrn_ is now known as amcrn | 19:18 | |
| *** arborism has quit IRC | 19:19 | |
| demorris | hub_cap: u there? | 19:29 |
| *** thedodd has joined #openstack-trove | 19:32 | |
| *** tanisdl has quit IRC | 19:35 | |
| *** michael-yu has quit IRC | 19:39 | |
| demorris | fyi I just registered a new BP for modifying an instance name. Heard it was pulled out of the Configuration Edits work, which is fine, but we need the ability to rename instances - https://blueprints.launchpad.net/trove/+spec/modify-instance-name | 19:39 |
| demorris | cp16net: already added an instance PUT call with the config edits work, so should be relatively straightforward to add an instance change command | 19:42 |
| *** michael-yu has joined #openstack-trove | 19:44 | |
| *** krow has quit IRC | 19:45 | |
| *** radez is now known as radez_g0n3 | 19:47 | |
| *** krow has joined #openstack-trove | 19:50 | |
| *** denis_makogon_ has joined #openstack-trove | 19:50 | |
| *** yidclare has quit IRC | 19:51 | |
| openstackgerrit | A change was merged to openstack/trove: Raises BadRequest for grant_access on ignore_dbs https://review.openstack.org/72595 | 19:51 |
| *** datsun180b has quit IRC | 19:52 | |
| *** denis_makogon has quit IRC | 19:52 | |
| *** denis_makogon_ is now known as denis_makogon | 19:52 | |
| *** dmakogon_ has joined #openstack-trove | 19:52 | |
| openstackgerrit | Greg Hill proposed a change to openstack/trove: Add scheduled tasks to the API https://review.openstack.org/73702 | 19:53 |
| *** krow1 has joined #openstack-trove | 19:54 | |
| *** krow has quit IRC | 19:54 | |
| *** datsun180b has joined #openstack-trove | 19:55 | |
| denis_makogon | ViswaV, ping | 19:55 |
| ViswaV | Hi denis_makogon | 19:56 |
| denis_makogon | ViswaV, about mongo backup/restore - ikhudoshyn doint this task | 19:56 |
| denis_makogon | ViswaV, i wrote it in your BP | 19:56 |
| denis_makogon | ViswaV, so, please talk to him, and then decide who will do this | 19:57 |
| denis_makogon | ViswaV, since this task booked on ikhudoshyn | 19:57 |
| ViswaV | Oh.. didn't notice that. I am half way through it… I thought I assigned that BP to my self… let me check | 19:57 |
| denis_makogon | ViswaV, i pretty much sure that its not the good idea to steal tasks =( | 19:58 |
| hub_cap | denis_makogon: no one knew ilya was working on it but u | 19:58 |
| hub_cap | so its not fair to say that | 19:58 |
| denis_makogon | hub_cap, blueprint about mongo says that | 19:58 |
| hub_cap | plz be nice ot our new friends :) | 19:58 |
| hub_cap | link? | 19:58 |
| hub_cap | if u mean the _original_ blueprint, then thats not a strong arguement | 19:59 |
| denis_makogon | https://blueprints.launchpad.net/trove/+spec/single-instance-mongodb-ga | 19:59 |
| hub_cap | dude thas the same blueprint | 19:59 |
| hub_cap | as soon as we merge mongo | 19:59 |
| hub_cap | that blueprint is done | 19:59 |
| hub_cap | thats why his code says "implements blueprint ..." | 19:59 |
| hub_cap | u cant say that he will contineu to wokr on everything forever for one blueprint | 20:00 |
| denis_makogon | so, you say that this task now assigned to ViswaV ? | 20:00 |
| ViswaV | denis_makogon: https://review.openstack.org/#/c/50597/ , it specifically says that it implements https://blueprints.launchpad.net/trove/+spec/single-instance-mongodb-ga and that only CRUD is being implemented. | 20:00 |
| hub_cap | yup ViswaV | 20:01 |
| ViswaV | Hence I created https://blueprints.launchpad.net/trove/+spec/backups-single-instance-mongodb and assigned to myself... | 20:01 |
| denis_makogon | ViswaV, last few words | 20:01 |
| hub_cap | still, you and ilya should talk ViswaV | 20:01 |
| denis_makogon | Will use mongodump/mongorestore for backup | 20:01 |
| hub_cap | but its not like u did anythign wrong | 20:01 |
| hub_cap | so dont think u did ViswaV :) | 20:01 |
| ViswaV | hub_cap: Thx. Will definitely talk to ilya. | 20:02 |
| hub_cap | word | 20:02 |
| * hub_cap dissapears again | 20:02 | |
| denis_makogon | i thought that mongo and cassandra features stays for me and Illya | 20:02 |
| hub_cap | denis_makogon: if you are owrking on a feature | 20:04 |
| hub_cap | u first blueprint it | 20:04 |
| demorris | probably more than enough work to do on those to have tasks/bp's shared around | 20:04 |
| hub_cap | u dont finish the feature and then put up a bp | 20:04 |
| hub_cap | if u put up a blueprint for c* backups | 20:04 |
| hub_cap | and assing to yourself, then u are working on it | 20:04 |
| hub_cap | u cant say ill alwaus work on every c* feature | 20:05 |
| hub_cap | if datastax caomes in and says hey, we wanna help, im not denying them ;) | 20:05 |
| denis_makogon | get it | 20:05 |
| demorris | hub_cap: and they just might | 20:05 |
| hub_cap | so yes denis_makogon u can work on c* as long as u want | 20:06 |
| hub_cap | just bp things first :) tell ilya to do the same.. thems the rules, since we are so distributed | 20:06 |
| denis_makogon | i thinks, after chatting with Illya we'll came into common agreement that'll satisfy all of us | 20:06 |
| hub_cap | cool thx denis_makogon | 20:06 |
| * hub_cap dissapears for real | 20:07 | |
| ViswaV | denis_makogon: That would be best. Thx. I don't see Illya online here. Do you want me to send him an email or you will sort out and let us know? | 20:07 |
| *** michael-yu has quit IRC | 20:08 | |
| denis_makogon | ViswaV, would be better to talk via emails - mine, dmakogon@mirantis.com, Illya's ikhudoshyn@mirantis.com | 20:08 |
| denis_makogon | emails would be better, since we are in different timezones | 20:08 |
| *** michael-yu has joined #openstack-trove | 20:09 | |
| denis_makogon | hub_cap, so, what about mongo initial patch ? | 20:10 |
| denis_makogon | ViswaV, sorry about doing hidden work with mongo, probably, it's my falt. | 20:12 |
| *** krow1 has quit IRC | 20:12 | |
| hub_cap | denis_makogon: dont worry about it, we will get it squared away | 20:13 |
| denis_makogon | ViswaV, now it'll be widely open as hell | 20:13 |
| hub_cap | wrt the work | 20:13 |
| hub_cap | wrt the patch, i have to test it, but im packing for austin, so itll be tested tonight or weekend | 20:13 |
| kevinconway | denis_makogon: that is very open | 20:13 |
| hub_cap | if it runs wel im sure itll be merged b4 the trove mid cycle sprint | 20:13 |
| hub_cap | and i know amcrn has tested it a ton | 20:14 |
| hub_cap | so im confident ill be good to go soon denis_makogon | 20:14 |
| *** amcrn has quit IRC | 20:15 | |
| *** krow has joined #openstack-trove | 20:15 | |
| denis_makogon | i submitted final patch for mongo, now it does everything correctly | 20:15 |
| denis_makogon | ViswaV, about your questions (setting strategy to guest conf) and sending it to the VM | 20:16 |
| ViswaV | denis_makogon: No problem. but yeah I too would like your guest agent code (https://review.openstack.org/#/c/50597/) to be merged as well So I can test the whole thing without having to cherry pick your changes first. | 20:16 |
| *** amcrn has joined #openstack-trove | 20:16 | |
| denis_makogon | ViswaV, i'm working on patch that'll give an ability to keep all backup strategies in guest conf | 20:17 |
| denis_makogon | ViswaV, you can see how i did it here https://review.openstack.org/#/c/73365/ | 20:17 |
| ViswaV | denis_makogon: Yes, I guess that would be done in the prepare() method of the mongodb manager…right? I am not touching that part yet. | 20:17 |
| ViswaV | I will take a look at that 73365 as well. | 20:18 |
| denis_makogon | i simply added oslo groups, that allows to separate option sets | 20:25 |
| denis_makogon | each group maned as datastore_manager | 20:25 |
| denis_makogon | [mysql], [cassandra], [redis], etc | 20:26 |
| denis_makogon | ViswaV, https://review.openstack.org/#/c/73365/2/trove/guestagent/backup/backupagent.py | 20:26 |
| denis_makogon | vipul, SlickNik, grapex, amcrn hub_cap, what do you think about updating defaults in trove, but not in devstack | 20:28 |
| SlickNik | denis_makogon: not sure I understand your question. | 20:29 |
| SlickNik | Defaults for what? | 20:29 |
| denis_makogon | SlickNik, https://review.openstack.org/#/c/71870/ | 20:29 |
| denis_makogon | sorry, forgot the link | 20:29 |
| amcrn | i agree with sean | 20:29 |
| denis_makogon | amcrn, so, you that this changes shoud be set in trove, right ? | 20:30 |
| demorris | anyone got a sec for a question on the configuration edits code? | 20:31 |
| denis_makogon | amcrn, and if we'll set update_status_on_fail = True in cfg.py, it would mean that this feature should not be optional at all | 20:31 |
| amcrn | well, let me explain: using devstack as a means to set values so that something *works* (in the case of the review above, increasing timeouts so cassandra doesn't init fail) is a lackluster option, because the sample confs should generally be a good guide for how to get trove up and running. | 20:31 |
| SlickNik | denis_makogon: The set of defaults in Trove should give you a working installation. | 20:31 |
| denis_makogon | SlickNik, it will keep it as it is, almost | 20:32 |
| denis_makogon | SlickNik, just will increase timeout and allow always mark instance with ERROR if timeout | 20:33 |
| denis_makogon | happed | 20:33 |
| SlickNik | amcrn: +1 | 20:33 |
| demorris | given that we are now keeping track of changes to the configuration via config options, does it still make sense to maintain a configuration template (for defaults)? | 20:34 |
| denis_makogon | so, i'll make some changes in cfg.py and will add the descripion in taskmanager.conf.sample where i'll say that this option should have at least _this_ value to support this feature | 20:35 |
| amcrn | demorris: are you referring to https://github.com/openstack/trove/blob/master/trove/templates/mysql/config.template (as an example)? | 20:35 |
| SlickNik | denis_makogon: There's a bigger issue here, I think. What might be a good timeout for one datastore, might not be good for another. | 20:35 |
| denis_makogon | SlickNik, timeout per datastore ? | 20:36 |
| demorris | amcrn: yeah | 20:36 |
| denis_makogon | SlickNik, ok, i'll take a look at it | 20:37 |
| amcrn | demorris: so cp16net can do a better job at elaborating on this, but those templates are moving to a db table | 20:37 |
| demorris | why would we not just store these defaults in the db per flavor size and pull from there... | 20:37 |
| demorris | oh | 20:37 |
| demorris | I thought that was for the available parameters | 20:37 |
| amcrn | and to explain the defaults: there are still some situations in which we (as trove) want to have a default value for a parameter that differs from a vanilla-install, but we don't want to allow users to manipulate/set it | 20:37 |
| demorris | not the defaults | 20:38 |
| amcrn | that's why defaults still has to exist, and as far as i know, that's moving to a table as well | 20:38 |
| demorris | amcrn: right so wouldn't you just flag those in the db with a boolean value such as "configurable" | 20:38 |
| amcrn | something to that effect, yes | 20:38 |
| demorris | additionally you could add "visible" | 20:38 |
| demorris | so you could hide once you don't want the user to see | 20:38 |
| amcrn | right | 20:39 |
| demorris | right now it just seems like things are a bit disparate, so with the current impl you have to do some merging to see 1) what are the defaults 2) what can I configure | 20:39 |
| amcrn | agreed, so once it moves to the table, the general picture of (1) what is available, and (2) what is configurable, should become crystal clear | 20:40 |
| amcrn | hopefully ;) | 20:40 |
| *** radez_g0n3 is now known as radez | 20:40 | |
| demorris | amcrn: okay cool, as long as were thinking about it :) | 20:40 |
| demorris | the original BP had this concept of storing the default and configurable fields | 20:41 |
| demorris | but it did not make it in to the final deliverable | 20:41 |
| amcrn | SlickNik denis_makogon: agreed on usage_timeout and usage_sleep_time per datastore | 20:41 |
| denis_makogon | amcrn, actually it was your idea (while cassandra testing), so, thanks | 20:42 |
| amcrn | np | 20:42 |
| SlickNik | What does this do: "update status on FAIL"? | 20:42 |
| SlickNik | And why do we need it? | 20:42 |
| amcrn | SlickNik: in short, if something goes wrong in the workflow (i.e. security-group creation fails, volume creation fails, etc.), the status is flipped to ERROR and the assets are cleaned up | 20:43 |
| amcrn | (after a timeout) | 20:44 |
| denis_makogon | SlickNik, before that instance will stuck in BUILD status | 20:44 |
| SlickNik | gotcha, it's for a cleanup of previously allocated resources on failure/error. | 20:46 |
| denis_makogon | amcrn, SlickNik, take a look at https://review.openstack.org/#/c/50944/ | 20:46 |
| denis_makogon | i rewrote it | 20:46 |
| denis_makogon | unfortunalety, gate is broken due to cinder, so rd posted -1 | 20:47 |
| denis_makogon | hub_cap, what about releasing new client version ? | 20:47 |
| *** michael-yu has quit IRC | 20:55 | |
| *** michael-yu has joined #openstack-trove | 20:57 | |
| *** michael-yu has quit IRC | 20:58 | |
| *** datsun180b has quit IRC | 21:00 | |
| *** datsun180b has joined #openstack-trove | 21:02 | |
| *** yidclare has joined #openstack-trove | 21:06 | |
| *** jimbobhickville has quit IRC | 21:06 | |
| *** denis_makogon has quit IRC | 21:13 | |
| mat-lowery | question regarding trove-gate is anyone is around | 21:24 |
| mat-lowery | Do we know why it's breaking? Does the investigation need assistance? | 21:25 |
| *** Barker has joined #openstack-trove | 21:26 | |
| mat-lowery | denis_makogon said, "gate is broken due to cinder". How did we arrive at this conclusion? (I'm trying to learn for future debugging purposes.) | 21:33 |
| *** Barker has quit IRC | 21:48 | |
| cp16net | mat-lowery: i'd say either by looking at the console output in the rdkenkins run | 21:50 |
| cp16net | mat-lowery: or by creating a new vm and running through the tests again | 21:51 |
| mat-lowery | cp16net: Thanks. Regarding the console output from Jenkins, the STDERR BEGIN lines (for example here: https://rdjenkins.dyndns.org/job/Trove-Gate/2529/console) don't seem helpful...in fact, it looks like a bug in the test configuration that it's echo'ing like that. | 21:52 |
| mat-lowery | Regarding a new VM, you mean a personal VM? Or do you mean some parameter in Jenkins? | 21:53 |
| cp16net | personal vm | 21:55 |
| cp16net | mat-lowery: check out at the top there is a link to *Full Log* | 21:55 |
| cp16net | https://rdjenkins.dyndns.org/job/Trove-Gate/2529/consoleFull | 21:55 |
| cp16net | or you can goto the [raw] link which show just the file | 21:56 |
| *** edmund1 has joined #openstack-trove | 21:58 | |
| *** edmund has quit IRC | 22:00 | |
| cp16net | mat-lowery: i've used vagrant with much success in replicating these issues | 22:00 |
| cp16net | although i have not run into an issue today | 22:00 |
| mat-lowery | I can get a clean run (provided I extend the timeouts on a testcase) on a local VM. | 22:01 |
| cp16net | LOL | 22:01 |
| cp16net | sounds like me too :-P | 22:01 |
| *** robertmyers has quit IRC | 22:06 | |
| *** SushilKM has joined #openstack-trove | 22:10 | |
| SushilKM | found that in guestagent\api.py some db calls are going asynchronous and others are synchronous, any specifics about that distinction | 22:10 |
| cp16net | SushilKM: if the api needs feedback from the guest is usually the distinction | 22:11 |
| cp16net | most calls like create user/database its just cast | 22:12 |
| SushilKM | how do we decide on the need of feedback | 22:12 |
| cp16net | usually depends on the api call being made | 22:12 |
| *** achampion has quit IRC | 22:12 | |
| cp16net | creates/deletes shouldnt need it | 22:12 |
| cp16net | but gets might | 22:12 |
| SushilKM | and about grant/revoke | 22:13 |
| cp16net | unless its data stored in the trove database and not the guest instance's db | 22:13 |
| SushilKM | of-course api calls are hitting the guest db | 22:13 |
| cp16net | we've tried to avoid doing too many calls that make sync calls to the db because it could take a while to respond from the api side of things | 22:14 |
| SushilKM | but does not the async call deprives user of seeing an error message from guest db | 22:14 |
| cp16net | depends on what the call is | 22:16 |
| *** edmund1 has quit IRC | 22:16 | |
| *** khyati_ has joined #openstack-trove | 22:16 | |
| SushilKM | "what the call is" means ... | 22:16 |
| cp16net | for create user for example we do validation to make sure its as "valid" as possible for sending it | 22:16 |
| SushilKM | request may be valid, but say if database crashes during the request | 22:17 |
| datsun180b | the thinking is that creates and grants for example take nonzero time and it's not fair to ask for the api user to stay on the line while we build a response | 22:17 |
| SushilKM | so does create and delete users and dbs ... do they take much time .... | 22:17 |
| datsun180b | in practice they don't take very long | 22:18 |
| SushilKM | wat i saw was create and delete are async with cast method but grant & revoke are using call method | 22:18 |
| SushilKM | not able to figure out distinction between y we have difference | 22:18 |
| datsun180b | grant & revoke are calls? that's a surprise, i thought they were casts, too | 22:19 |
| SushilKM | yep they are calls | 22:19 |
| SushilKM | https://github.com/openstack/trove/blob/master/trove/guestagent/api.py#L146 | 22:20 |
| SushilKM | https://github.com/openstack/trove/blob/master/trove/guestagent/api.py#L152 | 22:20 |
| SushilKM | grant & revoke respectively | 22:20 |
| * cp16net shruggs why thats the case | 22:20 | |
| datsun180b | well https://github.com/openstack/trove/blob/master/trove/guestagent/datastore/mysql/service.py#L378 is returning None | 22:20 |
| SushilKM | so thats wat i was questioning too, y do we decide that | 22:21 |
| cp16net | lolz | 22:21 |
| SushilKM | he he he | 22:21 |
| datsun180b | I'm guessing it's because of the validation? | 22:21 |
| cp16net | no clue | 22:21 |
| datsun180b | that is, the guest might just throw an explode at you if you try some shenanigans | 22:21 |
| *** yidclare has quit IRC | 22:22 | |
| SushilKM | so everyone shud be equal .... | 22:22 |
| SushilKM | i mean create/delete/grant/revoke | 22:22 |
| cp16net | datsun180b: the guest likes to throw a wrench in to your party | 22:22 |
| cp16net | lol | 22:22 |
| datsun180b | cp16net: the trick is to catch the wrench and throw it back | 22:22 |
| SushilKM | to throw it back we shud first have it | 22:22 |
| SushilKM | and create/delete does not gives the wrench to u | 22:23 |
| SushilKM | ;) | 22:23 |
| SushilKM | grant & revoke provides u wid the toool | 22:23 |
| datsun180b | well, we do appear to want the guest to do some validation | 22:23 |
| datsun180b | well what's the worst that can happen if you put up a review to change them all to one or the other, except getting shot down by hub_cap's i3 cannons | 22:24 |
| SushilKM | so, call does not supports the validation | 22:24 |
| datsun180b | well if it's a call and it explodes we won't hear about it | 22:25 |
| datsun180b | er wait, other way around | 22:25 |
| SushilKM | yep hub_cap is good at shooting ... have few experiences waiting in my queue ... | 22:25 |
| datsun180b | for casts you just throw it into the ocean and hope it gets there | 22:25 |
| SushilKM | it its a call, and if something bad happened, one wud get an info, for cast nothing happens | 22:26 |
| datsun180b | right | 22:26 |
| SushilKM | we are on with validations only ...... | 22:26 |
| datsun180b | if a cast fails the only way you'll know is you don't see the changes you asked for made | 22:26 |
| SushilKM | yep | 22:26 |
| SushilKM | but if i was on with a script making multiple requests for an application, its all garbage, that i wud get back | 22:27 |
| datsun180b | famous last words but "What's the worst that could happen?" | 22:27 |
| *** demorris has quit IRC | 22:29 | |
| kevinconway | datsun180b: famous last words "Snarf." | 22:29 |
| *** rwsu has quit IRC | 22:34 | |
| *** radez is now known as radez_g0n3 | 22:34 | |
| *** jasonb365 has quit IRC | 22:35 | |
| SushilKM | the worst could i could disturb my database ..... | 22:41 |
| *** jmontemayor has quit IRC | 22:43 | |
| *** jmontemayor has joined #openstack-trove | 22:44 | |
| mat-lowery | Anyone know who has access to trove-gate? Can I look at the config? | 22:49 |
| *** pdmars has quit IRC | 22:50 | |
| datsun180b | you mean https://rdjenkins.dyndns.org/job/Trove-Gate/? ? | 22:52 |
| datsun180b | you need to be in the group reddwarf-drivers to retrigger jobs | 22:52 |
| datsun180b | even then i can't view the configs for that job | 22:53 |
| mat-lowery | datsun180b: thanks | 22:53 |
| datsun180b | do you need anything retriggered, because i can do that | 22:54 |
| *** tanisdl has joined #openstack-trove | 22:54 | |
| mat-lowery | datsun180b: thanks. I wanted to see about archiving some logs as part of the job...and understanding some of the confusing output that I assume comes from the job itself | 22:57 |
| datsun180b | yeah, you'd probably have to configure the jobs to do that, though they do hold onto their output for a while | 22:58 |
| mat-lowery | ok thanks | 22:58 |
| datsun180b | https://rdjenkins.dyndns.org/job/Trove-Gate/1/console their archives go all the way back | 22:59 |
| mat-lowery | datsun180b: Thanks. | 23:00 |
| *** kevinconway has quit IRC | 23:01 | |
| *** michael-yu has joined #openstack-trove | 23:03 | |
| *** rwsu has joined #openstack-trove | 23:04 | |
| *** jcru has quit IRC | 23:07 | |
| *** datsun180b has quit IRC | 23:11 | |
| *** thedodd has quit IRC | 23:12 | |
| *** yogesh has quit IRC | 23:13 | |
| *** yogesh has joined #openstack-trove | 23:14 | |
| *** citizengrey has quit IRC | 23:14 | |
| *** citizengrey has joined #openstack-trove | 23:15 | |
| *** citizengrey has left #openstack-trove | 23:15 | |
| *** yogesh has quit IRC | 23:19 | |
| *** amytron has joined #openstack-trove | 23:19 | |
| *** rwsu has quit IRC | 23:22 | |
| openstackgerrit | Mat Lowery proposed a change to openstack/trove: Test restore full and restore incremental https://review.openstack.org/73736 | 23:31 |
| *** achampion has joined #openstack-trove | 23:39 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!