| leifmadsen | oh got splunk :) | 02:27 |
|---|---|---|
| leifmadsen | dmsimard: the real solution to pumping it into ES is probably to use filebeat | 02:27 |
| leifmadsen | s/got/god./ | 02:27 |
| leifmadsen | dmsimard: the other thing around that manifesto is that it can be a bit of a living document that can change over time to adjust to knowledge learned. No reason to not merge it as it's just git :) | 02:28 |
| leifmadsen | My own manifesto... http://blog.leifmadsen.com/blog/2015/11/19/rules-for-the-greater-goodness-a-product-development-guide/ | 02:28 |
| leifmadsen | I already know a lot of people hate me based on number 1 :) | 02:28 |
| dmsimard | leifmadsen: fair but I reached out to a lot of folks to have input on the manifesto, no rush to merge it -- monday is fine for me | 02:29 |
| leifmadsen | for sure | 02:29 |
| leifmadsen | I guess I mean leaving it open for weeks, not days :) | 02:30 |
| dmsimard | sure | 02:30 |
| dmsimard | no java ? | 02:30 |
| dmsimard | okay in my book | 02:30 |
| dmsimard | one of my own is right tool for the right job | 02:30 |
| dmsimard | I've seen people write shell scripts in php :/ | 02:30 |
| dmsimard | it's about the same level of doing shell scripts in nodejs | 02:31 |
| leifmadsen | I write too many shell scripts :) | 02:38 |
| leifmadsen | I should learn a real language some day | 02:38 |
| *** karimb has joined #ara | 09:24 | |
| *** Maximus86 has joined #ara | 10:34 | |
| Maximus86 | dmsimard: unable to reproduce my own issue at the moment, starting from a clean install of centos 7.2 minimal in order to create a bash script for reproducing | 10:35 |
| Maximus86 | dmsimard: I have the script, but it's not doing a lot | 10:36 |
| Maximus86 | dmsimard: my previous installations were from a kickstart file | 10:36 |
| Maximus86 | dmsimard: which I created for my work here, so it already deviated from standard. Afterwards some basic stuff was set on it via Ansible | 10:37 |
| Maximus86 | dmsimard: now I need to track down what could be different | 10:37 |
| *** karimb has quit IRC | 11:04 | |
| *** karimb has joined #ara | 11:06 | |
| *** BlessJah has joined #ara | 11:21 | |
| BlessJah | ara does not report correctly on tasks in loops (do_with), it says "One or more item failed" | 11:23 |
| BlessJah | is this known issue (missing feature)? | 11:23 |
| dmsimard | BlessJah: that is what Ansible sends to ara :( | 11:37 |
| dmsimard | I don't know if there is an issue in Ansible upstream to improve what Ansible sends to callbacks for looped tasks | 11:37 |
| dmsimard | Maximus86: ack, so something coming from your Kickstart perhaps ? | 11:38 |
| BlessJah | does ara have a bugtracker? | 11:39 |
| BlessJah | dmsimard: I've used register and ara_record, the data is available there | 11:50 |
| BlessJah | but I had to "failed_when: false" to allow next task (ara_record) to run, thus loosing information that failure happened | 11:51 |
| dmsimard | BlessJah: it does, it's the new openstack tracker but I need to write down some documentation on how to use it. | 12:02 |
| dmsimard | https://storyboard.openstack.org/#!/search?project_id=843 | 12:02 |
| BlessJah | kthx | 12:02 |
| dmsimard | If that sounds too intimidating, you can use github.com/dmsimard/ara and I'll handle it there or migrate it myself | 12:02 |
| dmsimard | BlessJah: instead of using failed_when perhaps you could use a block/rescue/always :) | 12:03 |
| dmsimard | So that you can properly fail, say run the record task in the rescue and then have a fail task | 12:05 |
| leifmadsen | oh I like the look of storyboard so much more than launchpad | 12:35 |
| dmsimard | leifmadsen: they've done a good job so far. It still needs some work but it's usable. | 12:37 |
| leifmadsen | dmsimard: I'm trying to figure out how to link issues to reviews... | 12:37 |
| leifmadsen | other than just adding a note... | 12:37 |
| leifmadsen | that might be the only way unfortunately | 12:38 |
| dmsimard | leifmadsen: Story: and Task: | 12:39 |
| leifmadsen | worklist needs cleaning up it seems | 12:39 |
| dmsimard | There is gerrit integration (links in review and comments in storyboard) | 12:40 |
| leifmadsen | there is? | 12:40 |
| leifmadsen | hrmmm | 12:40 |
| leifmadsen | looking here https://storyboard.openstack.org/#!/story/2000868 | 12:40 |
| leifmadsen | (mostly playing because I hadn't looked at storyboard yet) | 12:40 |
| dmsimard | leifmadsen: I didn't put the story# in the commit message | 12:41 |
| leifmadsen | gotcha | 12:41 |
| leifmadsen | like I said, I'm mostly just playing, learning how this could operate :) | 12:41 |
| dmsimard | I'll add it in a minute | 12:41 |
| leifmadsen | yea all good, just seems like you should be able to link the other way easy enough as well | 12:42 |
| leifmadsen | but guess project is still pretty new | 12:42 |
| dmsimard | it's been like two years in the making already :p | 12:46 |
| *** karimb has quit IRC | 13:09 | |
| *** karimb has joined #ara | 13:18 | |
| Maximus86 | dmsimard: is ara prefetching the whole page in the UI? | 13:28 |
| dmsimard | Maximus86: define whole page | 13:29 |
| dmsimard | Maximus86: in the reports page, the different result tables are loaded asynchronously (AJAX) and things like host facts, files or task results are iframes so they are not loaded | 13:30 |
| Maximus86 | running some tests atm on anm | 13:30 |
| Maximus86 | *on an underpowered VM and it's locking up my browser, saw a bunch of DB calls in the log | 13:31 |
| Maximus86 | hence my question | 13:31 |
| Maximus86 | (underpowered because of my shitty laptop and too much other stuff running) | 13:32 |
| Maximus86 | still not able to reproduce the issue | 13:32 |
| Maximus86 | (on the minimal install) | 13:34 |
| dmsimard | Maximus86: running a db with a couple hundred thousand records in the embedded server worked fine straight from my laptop (and in fact I used it to optimize performance) so it shouldn't be too bad | 13:35 |
| dmsimard | Maximus86: this one :) https://twitter.com/dmsimard/status/838903230256009218 | 13:36 |
| Maximus86 | that's neat | 13:36 |
| Maximus86 | which db? | 13:36 |
| Maximus86 | *which db server should be the correct question | 13:37 |
| dmsimard | that was sqlite | 13:37 |
| Maximus86 | that's impressive as hell | 13:37 |
| dmsimard | sqlite is really really performant | 13:37 |
| dmsimard | I don't know where it's bad reputation comes from | 13:38 |
| Maximus86 | was it not with data corruption? | 13:38 |
| dmsimard | ¯\_(ツ)_/¯ | 13:38 |
| dmsimard | sqlite is a very fine solution if you don't need stuff like network access or authentication | 13:38 |
| Maximus86 | I'll look into it, it sure is impressive. I'm using ARA with mysql as I already have that db server running for other applications on the same host | 13:39 |
| Maximus86 | what bothers me more than the ajax errors | 13:40 |
| Maximus86 | are KeyError: '001_initial_revision_py' | 13:40 |
| Maximus86 | in the apache log | 13:40 |
| Maximus86 | I don't have them on the complete-from-scratch test machine either | 13:41 |
| dmsimard | Maximus86: so you can't reproduce your issue outside of your kickstarted environment ? | 13:45 |
| Maximus86 | so far, no.. but working hard to break it :P | 13:46 |
| dmsimard | Maximus86: lol | 13:46 |
| BlessJah | dmsimard: block/always works as a charm | 13:46 |
| dmsimard | BlessJah: \o/ | 13:46 |
| dmsimard | Maximus86: perhaps there is a conflict of python libs in your kickstarted environment | 13:46 |
| dmsimard | Maximus86: i.e, a rpm install of python-something and a pip install of something | 13:46 |
| Maximus86 | yeah, ansible is installed from RPM, and later updated via pip | 13:47 |
| Maximus86 | already looking into that | 13:47 |
| dmsimard | BlessJah: your comment came up before, I'll try and look if there's an issue upstream now. | 13:49 |
| * dmsimard searches through 1718 opened issues | 13:49 | |
| BlessJah | dmsimard: do you plan to add tower-like features? triggering ansible, authorisation, dashboard? | 13:53 |
| dmsimard | BlessJah: nope, none of that | 13:53 |
| dmsimard | BlessJah: If you'd like some background/context, there is http://ara.readthedocs.io/en/latest/faq.html#why-don-t-you-use-ansible-tower-rundeck-or-semaphore and also something I'm working on right now: https://review.openstack.org/#/c/446709/4/doc/source/manifesto.rst | 13:54 |
| BlessJah | what's the scope of ara? centralized logging and visualisation? | 13:54 |
| dmsimard | BlessJah: yep. | 13:54 |
| dmsimard | BlessJah: for tower-like features, there is.... tower :) | 13:54 |
| BlessJah | $$$ | 13:55 |
| leifmadsen | gotta pay to play :) | 13:55 |
| Maximus86 | did RH not opensource tower last fall? | 13:55 |
| BlessJah | doesn't seem so | 13:56 |
| dmsimard | Red Hat acquired Ansible but Tower is not open source yet | 13:56 |
| leifmadsen | not open source yet | 13:56 |
| Maximus86 | there's also foreman, which actually plays together very nicely with Ansible | 13:56 |
| BlessJah | yet? | 13:56 |
| dmsimard | They'll open source it, it just takes time | 13:56 |
| leifmadsen | that. | 13:56 |
| leifmadsen | :) | 13:56 |
| leifmadsen | it will be open sourced at some point though | 13:56 |
| dmsimard | It's more work than people realize, you can't just go and do a git push to github and call it a day :) | 13:56 |
| BlessJah | theforeman requires way too much resources | 13:57 |
| BlessJah | It'd be nice to have text editor and crontab build into i.e. ara (yep it's against manifesto) | 13:59 |
| dmsimard | BlessJah: those are features that are available outside of ara | 14:00 |
| *** Maximus86 has quit IRC | 14:00 | |
| dmsimard | BlessJah: you can set a cron on your server or laptop to run ansible :) | 14:00 |
| dmsimard | and you have *gasp* vim or emacs if you're into that | 14:00 |
| BlessJah | yeah, been there done that | 14:00 |
| dmsimard | BlessJah: the other thing with file editing is that ara doesn't store actual files | 14:01 |
| dmsimard | only their content | 14:01 |
| dmsimard | so there's nothing to really edit in the first place | 14:01 |
| BlessJah | I'd prefer VCS hooked into jenkins feeding ara | 14:01 |
| BlessJah | dmsimard: it's good that ara stores file content and not read files on demand (what if user modifies said file after run?) | 14:02 |
| BlessJah | dmsimard: is the schema content-adressable? for deduplication? | 14:02 |
| dmsimard | BlessJah: it stores file content in a git-like way, sha1 hashed and zlib compressed | 14:02 |
| dmsimard | so we only ever store one copy of one file and compressed | 14:02 |
| BlessJah | lovely | 14:03 |
| dmsimard | if you run the same playbook with the same files 10 times, you only have the files once | 14:03 |
| dmsimard | if you do a select in the database you'll see gibberish but that's not exactly an expected way of browsing the file content | 14:03 |
| * BlessJah is inspecting schema right now | 14:03 | |
| *** tbielawa has joined #ara | 15:09 | |
| *** jparrill has joined #ara | 15:32 | |
| *** themurph has quit IRC | 15:44 | |
| *** themurph has joined #ara | 15:45 | |
| *** jparrill has quit IRC | 16:11 | |
| tbielawa | dmsimard: i'm attempting to learn the gerrit + git-review system now. | 16:31 |
| tbielawa | got launchpad hooked up, managed to locate the project on review.openstack too! | 16:31 |
| tbielawa | where does your commit actually live? I had to clone from github, but I guess your commits live on the openstack git repos... | 16:31 |
| dmsimard | tbielawa: gerrit takes some time to get used to but after that you can't work with anything else :) | 16:32 |
| tbielawa | oh snap. i think I can just click a line to add comments. | 16:32 |
| tbielawa | we'll see about that | 16:32 |
| dmsimard | tbielawa: under the hood, gerrit is the one hosting the git repo -- it creates refs with the commits in them | 16:32 |
| dmsimard | So for https://review.openstack.org/#/c/446709/ | 16:33 |
| dmsimard | The commit is actually here https://review.openstack.org/gitweb?p=openstack/ara.git;a=commitdiff;h=a2e0a607de7a78122b94c567ddfac8dd35a205c2 | 16:33 |
| dmsimard | but since the git repo is also mirrored | 16:34 |
| dmsimard | you can also see it here: http://git.openstack.org/cgit/openstack/ara/commit/?id=a2e0a607de7a78122b94c567ddfac8dd35a205c2 | 16:34 |
| tbielawa | many places | 16:34 |
| dmsimard | or here: https://github.com/openstack/ara/commit/a2e0a607de7a78122b94c567ddfac8dd35a205c2 | 16:34 |
| tbielawa | lol | 16:34 |
| tbielawa | it IS on github? | 16:34 |
| tbielawa | $ git show a2e0a607de7a78122b94c567ddfac8dd35a205c2 | 16:35 |
| tbielawa | fatal: bad object a2e0a607de7a78122b94c567ddfac8dd35a205c2 | 16:35 |
| dmsimard | tbielawa: you need to fetch the ref first | 16:35 |
| dmsimard | at the top right in gerrit | 16:35 |
| dmsimard | you have a "download" button | 16:35 |
| tbielawa | so i need to add gerrit as a remote too then? | 16:35 |
| tbielawa | oh | 16:35 |
| dmsimard | nope, just git clone from github | 16:36 |
| dmsimard | and then you can do, like, git fetch https://git.openstack.org/openstack/ara refs/changes/09/446709/4 && git checkout FETCH_HEAD | 16:36 |
| dmsimard | to check out that particular patch | 16:36 |
| tbielawa | my brain is trying to figure out what assumptions there are that make that intuitive in any way | 16:36 |
| tbielawa | heh | 16:36 |
| dmsimard | if you want to submit a new patch set, you check out the patch, edit what you need to edit and then do "git commit --amend" and then another "git review" | 16:36 |
| dmsimard | When you're happy with the review, you can submit it -- but you're not the one merging the code, gerrit is. | 16:37 |
| dmsimard | It's quite different than github pull requests | 16:37 |
| tbielawa | oh snap, git remote add -f gettit https://git.openstack.org/openstack/ara actually worked | 16:37 |
| dmsimard | yeah, but you don't *need* to set the remote | 16:37 |
| tbielawa | different strokes, different people. | 16:38 |
| dmsimard | the "git review" tool has a command for that actually | 16:38 |
| dmsimard | it's git review -s | 16:38 |
| tbielawa | i checked out git review --list, then got the login prompt | 16:38 |
| tbielawa | and then errors. even w/ ssh key uploaded. | 16:38 |
| dmsimard | not familiar with --list | 16:38 |
| tbielawa | -l, --list | 16:39 |
| tbielawa | List the available reviews on the gerrit server for this project. | 16:39 |
| dmsimard | I'm not sure that works | 16:39 |
| dmsimard | well, never used it personally anyway | 16:39 |
| tbielawa | lol. it didn't :) | 16:40 |
| dmsimard | https://review.openstack.org/#/q/project:openstack/ara works well enough for me | 16:40 |
| dmsimard | but you can also make customized dashboards with different criterias | 16:40 |
| tbielawa | interesting | 16:41 |
| tbielawa | ohhhhhhhhhhhhhhhhhhhh snap. I manually set my gerrit name and then cloned from git.openstack. then! then git review --list worked | 16:43 |
| tbielawa | nice nice nice | 16:43 |
| tbielawa | nice! git review -d 446709 | 16:46 |
| tbielawa | perfect | 16:46 |
| tbielawa | $ git review -d 446709 | 16:47 |
| tbielawa | Downloading refs/changes/09/446709/4 from gerrit | 16:47 |
| tbielawa | Switched to branch "review/david_moreau_simard/446709" | 16:47 |
| tbielawa | dmsimard: i clicked a line on here https://review.openstack.org/#/c/446709/4/doc/source/manifesto.rst and entered some review text. it's there now on my screen, says it's a Draft comment. I can edit or discard | 16:59 |
| tbielawa | not sure how to just save the review comment though | 16:59 |
| tbielawa | I am also trying to submit a new patchset for review with some x-refs | 17:00 |
| dmsimard | In the main review page, click reply at the top and then submit | 17:00 |
| dmsimard | tbielawa: a new patchset for the manifesto review or something else ? | 17:01 |
| tbielawa | manifesto review | 17:01 |
| tbielawa | oh my, this interface looks so simple but aparantly it's more web two-point-oh than I though. I figured that navigating away from the manifesto page was going to forget my changes. | 17:01 |
| dmsimard | Don't submit new patches for reviews that aren't yours :p | 17:01 |
| tbielawa | lol | 17:01 |
| dmsimard | Unless you're taking over the patch and it was agreed or something | 17:02 |
| tbielawa | I got ya | 17:02 |
| dmsimard | Proper etiquette :) | 17:02 |
| tbielawa | ah, i did it! first gerrit review reply | 17:02 |
| dmsimard | Achievement unlocked | 17:02 |
| tbielawa | :DD | 17:02 |
| dmsimard | tbielawa: now if you want I can even show you how to add that little message that ARA is loaded before running the playbook if you'd like | 17:04 |
| tbielawa | let me guess first | 17:04 |
| tbielawa | you did write some contributing docs after all. lemme see what you got here | 17:04 |
| dmsimard | tbielawa: okay, sure | 17:04 |
| tbielawa | clone, checkout new branch, commit, git review | 17:04 |
| dmsimard | but I meant like, where in the code :p | 17:04 |
| dmsimard | you don't necessarily need to checkout a new branch | 17:05 |
| dmsimard | you can check out master | 17:05 |
| tbielawa | ah, so not exactly like this https://ara.readthedocs.io/en/latest/contributing.html#sending-a-patch-for-review | 17:05 |
| dmsimard | you can also base patches on top of each other | 17:05 |
| tbielawa | so many options! nice | 17:05 |
| dmsimard | tbielawa: yeah you can use a new branch, just saying it's not necessary | 17:05 |
| tbielawa | You can point me at the right file to add that little loaded message :-) | 17:05 |
| dmsimard | tbielawa: probably somewhere in there: https://github.com/openstack/ara/blob/master/ara/plugins/callbacks/log_ara.py#L255 | 17:06 |
| tbielawa | then I'll do the hard stuff, like following whatever pattern youve' established now for toggling config options, like adding an 'enable "ara loaded" message' switch | 17:06 |
| tbielawa | nice nice nice. ty | 17:06 |
| dmsimard | tbielawa: bah I don't think we need a config option for that | 17:07 |
| tbielawa | nah? i figured some folks would like quiet over loud. make it verbose if you wish kind of approach | 17:07 |
| dmsimard | it's probably fine to just print a little line (and eventually get rid of the obnoxious alembic migration messages) | 17:07 |
| * tbielawa nods | 17:08 | |
| tbielawa | ack | 17:08 |
| dmsimard | tbielawa: so another thing | 17:08 |
| dmsimard | tbielawa: is that there are gate jobs | 17:08 |
| dmsimard | so if you look at https://review.openstack.org/#/c/446709/ | 17:08 |
| tbielawa | i see them there, yeah | 17:08 |
| dmsimard | they're there on the right hand side | 17:08 |
| tbielawa | the jenkins checks table | 17:08 |
| dmsimard | clicking on the job name gets you something like this: http://logs.openstack.org/09/446709/4/check/gate-ara-integration-latest-centos-7/6b27841/ | 17:08 |
| dmsimard | console from the job http://logs.openstack.org/09/446709/4/check/gate-ara-integration-latest-centos-7/6b27841/console.html (timestamps are clickable for direct links) | 17:09 |
| tbielawa | NICE. that'll make tracking down build fails easier | 17:09 |
| dmsimard | and then you have different logs here http://logs.openstack.org/09/446709/4/check/gate-ara-integration-latest-centos-7/6b27841/logs/ | 17:09 |
| dmsimard | for example http://logs.openstack.org/09/446709/4/check/gate-ara-integration-latest-centos-7/6b27841/logs/build/ | 17:09 |
| tbielawa | neat | 17:10 |
| BlessJah | dmsimard: could setup task receive some default taskname? "Setup <host>" or sth would make report look better imo | 17:19 |
| dmsimard | BlessJah: ah, another part of the manifesto :D | 17:20 |
| BlessJah | which one? | 17:20 |
| dmsimard | BlessJah: ARA doesn't change or transform what ansible provides | 17:20 |
| dmsimard | It's pass-through | 17:20 |
| BlessJah | :/ | 17:21 |
| dmsimard | So we could have ansible do a better job and give a proper name to the "meta" setup task | 17:21 |
| * BlessJah starts to really like this project | 17:23 | |
| dmsimard | BlessJah: so I tell you no and you like it ? that's cool too :p | 17:24 |
| dmsimard | BlessJah: so it's probably somewhere in here: https://github.com/ansible/ansible/blob/edcbef27ec0c9ee9bec1b1898c9b3b57ddc2b029/lib/ansible/modules/system/setup.py | 17:26 |
| BlessJah | you have vision (manifesto), and reasonable explanation when you reject such ideas | 17:27 |
| BlessJah | what I'd like to see though is "ongoing" status instead of "aborted" when ansible is running | 17:28 |
| dmsimard | ah, yeah, I think that's written down somewhere | 17:28 |
| BlessJah | and yes, I know why job is "aborted" and why it may not be possible to do it | 17:29 |
| dmsimard | The wording is important though, it says incompleted, not aborted | 17:29 |
| dmsimard | Which, technically, it is incomplete :P | 17:29 |
| dmsimard | BlessJah: https://storyboard.openstack.org/#!/story/2000852 | 17:29 |
| dmsimard | It's not an easy problem from the perspective of ara | 17:30 |
| BlessJah | it reads "Playbook execution was interrupted: data may be incomplete and inconsistent" | 17:30 |
| dmsimard | BlessJah: oh, I screwed up that wording then | 17:30 |
| dmsimard | Let's fix it | 17:30 |
| BlessJah | message is the same for running and actually aborted tasks | 17:31 |
| dmsimard | yeah, I'll removed the wording around interrupted | 17:31 |
| dmsimard | ara doesn't have a way to know or differentiate between the two | 17:31 |
| dmsimard | I explained a bit the context in that storyboard link I sent | 17:31 |
| BlessJah | yep, you'd need ansible/ara-plugin side daemon and ara-server side daemon communicate | 17:32 |
| dmsimard | yes and no | 17:33 |
| dmsimard | ansible just sends events to ara | 17:33 |
| dmsimard | ara doesn't know if there are any more events coming | 17:33 |
| dmsimard | the only thing it knows is if the playbook has finished running because there is a hook for that | 17:33 |
| BlessJah | can you differentiate between different aborts? e.g. network dies between ansible and ara, ansible dies because of ^C, ansible dies because of pause/abort? | 17:35 |
| dmsimard | ansible might, ara wouldn't | 17:35 |
| dmsimard | technically when something is interrupted, it's ansible, not ara | 17:36 |
| dmsimard | ansible just sends data to ara, ara doesn't really "run" | 17:36 |
| dmsimard | on a per-hook basis | 17:36 |
| dmsimard | the hooks look like this: https://github.com/ansible/ansible/blob/devel/lib/ansible/plugins/callback/__init__.py#L218-L385 | 17:37 |
| dmsimard | BlessJah: fyi ansible dev told me that gather_facts will have a task name in Ansible 2.3 :) | 17:44 |
| dmsimard | 'Gathering facts' | 17:46 |
| BlessJah | any SQL injections lurking in task names? :D | 17:46 |
| dmsimard | I would hope not but I haven't tried or had a formal security audit done | 17:48 |
| dmsimard | ARA isn't coded in PHP with it's own SQL unescaped queries :p | 17:50 |
| leifmadsen | :D | 17:50 |
| BlessJah | dmsimard: indeed the task has a name | 17:54 |
| * dmsimard needs to test ansible 2.1.5, 2.2.3 and 2.3 | 17:56 | |
| dmsimard | oh, 2.2.2 | 17:56 |
| * BlessJah tested pip install https://github.com/ansible/ansible/archive/v2.3.0.0-0.1.rc1.tar.gz | 18:04 | |
| dmsimard | ohhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh | 18:09 |
| dmsimard | https://twitter.com/dmsimard/status/842800964948189185 | 18:13 |
| dmsimard | jrist: ^ | 18:14 |
| dmsimard | tbielawa: oh no you're an emacs guy | 18:38 |
| * dmsimard runs away | 18:38 | |
| tbielawa | somebody just saw my twitter | 18:41 |
| tbielawa | :-) | 18:41 |
| tbielawa | who wouldn't like M-x ansible-doc | 18:42 |
| tbielawa | to look up module documentation | 18:42 |
| *** tbielawa is now known as tbielawa|lunch | 18:45 | |
| dmsimard | :) | 18:46 |
| jrist | they look good dmsimard ! | 19:17 |
| BlessJah | dmsimard: have you thought about exposing API for plugin, instead of connecting directly to database? | 19:31 |
| BlessJah | for local setup there is no difference (local sqlite may even be preffered) but for remote ara server it doesn't feel right to directly expose db to probably semi-trusted nodes (i.e. jenkins executors) | 19:34 |
| *** tbielawa|lunch is now known as tbielawa | 19:45 | |
| dmsimard | BlessJah: yes | 20:24 |
| dmsimard | BlessJah: eventually want to separate callback/client/server/api/webapp | 20:24 |
| BlessJah | great | 20:25 |
| BlessJah | that's what I wanted to hear :D | 20:25 |
| *** tbielawa is now known as tblablabrb | 20:28 | |
| *** tblablabrb is now known as tbielawa | 20:45 | |
| *** tbielawa has quit IRC | 20:55 | |
| *** karimb has quit IRC | 21:51 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!