| *** etienne has quit IRC | 00:14 | |
| *** harlowja has joined #ara | 00:23 | |
| dmsimard | oh, right, some people put *spaces* in their file or directory names | 02:07 |
|---|---|---|
| openstackgerrit | David Moreau Simard proposed openstack/ara-server master: Fixed config code example if the path has spaces. https://review.openstack.org/631030 | 02:09 |
| openstackgerrit | Merged openstack/ara-server master: Fixed default configuration loading and do not write default config if a config is specified. https://review.openstack.org/631029 | 02:13 |
| dmsimard | oh, sweet. | 02:24 |
| dmsimard | The new integration tests save the database as an artifact of the jobs | 02:25 |
| dmsimard | so I just used one to update the demo server \o/ | 02:25 |
| dmsimard | apollo13: I guess we're a bit redundant: https://i.imgur.com/HEAt8pb.png I'll fix that | 02:34 |
| openstackgerrit | Merged openstack/ara-web master: Move playbook details page in a dedicated page https://review.openstack.org/628451 | 02:53 |
| openstackgerrit | David Moreau Simard proposed openstack/ara-web master: Do not merge: test playbook info https://review.openstack.org/631130 | 04:00 |
| *** sshnaidm has quit IRC | 04:52 | |
| *** sshnaidm has joined #ara | 04:55 | |
| ara-slack | vandanakec: @vandanakec has joined the channel | 06:56 |
| ara-slack | kelian.saintbonnet: @dmsimard I've finally managed to get ARA working with AWX, thank you for your help | 08:06 |
| ara-slack | kelian.saintbonnet: I will redo my steps with a fresh install of AWX to have a full view of what must be done | 08:08 |
| ara-slack | kelian.saintbonnet: And it may help you to write your documentation | 08:10 |
| *** themroc has joined #ara | 08:23 | |
| *** herald85 has joined #ara | 09:36 | |
| *** sshnaidm is now known as sshnaidm|afk | 10:28 | |
| ara-slack | kelian.saintbonnet: @dmsimard In fact you don't need to create a custom virtualenv | 11:01 |
| ara-slack | kelian.saintbonnet: I've just installed ARA in the awx_task container but not in a virtualenv | 11:02 |
| ara-slack | kelian.saintbonnet: Changed the ansible.cfg file according to the plugins path | 11:02 |
| ara-slack | kelian.saintbonnet: Added the path to callback plugin in AWX UI | 11:02 |
| ara-slack | kelian.saintbonnet: And, this is important because it doesn't work without that, I need to remove the HOME extra environment variable from AWX UI | 11:03 |
| apollo13 | awx seems to do something fishy | 12:35 |
| *** sshnaidm|afk is now known as sshnaidm | 12:41 | |
| dmsimard | apollo13: it mixes rpm installs with virtualenvs :( | 12:42 |
| dmsimard | https://github.com/ansible/awx/issues/1737#issuecomment-442454519 | 12:42 |
| openstackgerrit | Florian Apolloner proposed openstack/ara-server master: Renamed most of config to settings and dropped the default_ prefix. https://review.openstack.org/631034 | 12:44 |
| apollo13 | dmsimard: I think I fixed the merge issues, lets see what zuul says | 12:44 |
| ara-slack | dmsimard: @kelian.saintbonnet it's not recommended to install things outside the virtualenv since it can conflict with python dependencies otherwise installed through distribution packages | 12:46 |
| ara-slack | dmsimard: That's why getting ara to work in the default virtualenv would be nice | 12:46 |
| openstackgerrit | Merged openstack/ara-server master: Fixed config code example if the path has spaces. https://review.openstack.org/631030 | 12:47 |
| ara-slack | dmsimard: But the default virtualenv inherits from system site packages and Ansible comes from rpm, not from pip | 12:47 |
| ara-slack | dmsimard: Anyway, this is all documented in that AWX issue I linked | 12:48 |
| openstackgerrit | Florian Apolloner proposed openstack/ara-server master: Renamed most of config to settings and dropped the default_ prefix. https://review.openstack.org/631034 | 12:56 |
| apollo13 | zuul hates me | 12:57 |
| apollo13 | dmsimard: feel free to review the last one and merge :) | 12:58 |
| dmsimard | apollo13: I'm not fully woken up either | 12:59 |
| apollo13 | no rush, that is it for me for the next few hours anyways | 12:59 |
| apollo13 | unless I get bored at work | 12:59 |
| dmsimard | apollo13: any minute of your time is highly appreciated :) | 13:00 |
| dmsimard | I'm happy we ironed out the manage.py and secret key things yesterday, that was productive | 13:01 |
| dmsimard | apollo13: web is slowly coming together too: http://logs.openstack.org/30/631130/1/check/ara-web-build-dashboard/43582ec/npm/html/playbooks | 13:02 |
| dmsimard | Hmmm, that link doesn't actually work | 13:02 |
| dmsimard | I guess it's http://logs.openstack.org/30/631130/1/check/ara-web-build-dashboard/43582ec/npm/html | 13:03 |
| apollo13 | yeah, weirdly enough zuul didn't recheck my last rebase | 13:03 |
| dmsimard | Link ? | 13:03 |
| apollo13 | https://review.openstack.org/#/c/631034/ the last patchset here | 13:03 |
| dmsimard | It's still running | 13:04 |
| dmsimard | See http://zuul.openstack.org/status | 13:04 |
| dmsimard | And filter for "ara" | 13:04 |
| apollo13 | how do you find those dashboard links after it's finished? | 13:06 |
| apollo13 | ah toggle ci | 13:09 |
| apollo13 | there it is | 13:09 |
| dmsimard | You can also search in builds, projects, etc | 13:09 |
| apollo13 | jupp | 13:09 |
| apollo13 | anyways, did pass now | 13:09 |
| apollo13 | at least the server part of it | 13:10 |
| dmsimard | Ok, will take a look after dropping kids at school | 13:10 |
| *** spiette has quit IRC | 13:34 | |
| *** spiette has joined #ara | 13:41 | |
| ara-slack | amurphy: @amurphy has joined the channel | 13:47 |
| *** Jmainguy has joined #ara | 14:17 | |
| Jmainguy | can ansible being run from a centos6 box, connect to a ara server? tried installing ara on the cent6 box and it says I need python 2.7 | 14:17 |
| Jmainguy | we have 2 main cent6 boxes alot of ansible is run from, so i wanted to capture input from them | 14:17 |
| Jmainguy | by input I mean ansible run output | 14:18 |
| dmsimard | Jmainguy: ohai | 14:29 |
| dmsimard | I believe 2.7 is as low as it will go but I forget exactly why | 14:29 |
| Jmainguy | wuzza buddy, sounds about right | 14:30 |
| dmsimard | What version of Ansible are you using on c6 ? | 14:30 |
| dmsimard | Ansible dropped 2.6 support a while ago already iirc | 14:30 |
| Jmainguy | ansible-2.6.7-1.el6.noarch | 14:30 |
| Jmainguy | https://docs.ansible.com/ansible/latest/reference_appendices/release_and_maintenance.html | 14:31 |
| Jmainguy | gives me an excuse to update our ansible central host to centos7 I guess =) | 14:32 |
| dmsimard | Jmainguy: if ansible works on 2.6 maybe ara does ? I would need to test | 14:32 |
| dmsimard | I know that we don't run integration tests with 2.6 anymore for sure | 14:33 |
| Jmainguy | it might, but I am nub and just following the documentation on how to install ara | 14:33 |
| Jmainguy | pip doesnt want to install it | 14:33 |
| Jmainguy | pip complains it needs 2.7 for ara basically | 14:33 |
| dmsimard | Jmainguy: can you pastebin that ? | 14:33 |
| Jmainguy | sure standby | 14:33 |
| Jmainguy | https://paste.fedoraproject.org/paste/i9e7acQpHHy2ijimWF~nYA | 14:34 |
| Jmainguy | flask sqlalchemy requires 2.7 | 14:34 |
| dmsimard | Hmm, perhaps you can find out what is the latest release that supports 2.6 and see if it breaks anything | 14:35 |
| dmsimard | No guarantee though :( | 14:35 |
| Jmainguy | ack, one more nub question | 14:36 |
| Jmainguy | to setup a remote ara server, and send stuff to it | 14:36 |
| Jmainguy | is that the host,port settings in ansible.cfg | 14:36 |
| Jmainguy | or is there more settings then that to connect to a remote ara module (with callbacks configured ofc) | 14:36 |
| dmsimard | So in ara 0.x, everything is bundled together and the callback talks directly to the database -- sqlite by default but this can be mysql/postgre/etc | 14:37 |
| Jmainguy | gotcha, so for seperating ti out, I should use mysql, and config database in ansible.cfg | 14:37 |
| dmsimard | In the upcoming 1.0, we're introducing an API in between | 14:38 |
| Jmainguy | nice | 14:38 |
| dmsimard | What this means is that you can have callbacks send their data to a remote API if need be | 14:38 |
| Jmainguy | yeah thats ideal | 14:38 |
| Jmainguy | I can do the database trick until then | 14:38 |
| dmsimard | But we still default to everything local without requiring an API server to be up | 14:39 |
| dmsimard | This is what the API looks like: https://api.demo.recordsansible.org/api/v1/ | 14:39 |
| Jmainguy | nice | 14:39 |
| dmsimard | And the new WIP web UI reads from that API: https://web.demo.recordsansible.org/ | 14:40 |
| *** ssbarnea|rover has quit IRC | 14:40 | |
| dmsimard | The architecture is very different, almost everything has been rewritten from scratch | 14:40 |
| dmsimard | I mention this because the new back end will be py3 only (no official centos support until 8) | 14:41 |
| *** ssbarnea|rover has joined #ara | 14:41 | |
| Jmainguy | nice, that should be fine | 14:41 |
| Jmainguy | I plan on running it in a container on kube, just need to be able to connect to it with ansible from laptops and central server | 14:42 |
| dmsimard | However, I'd like to have a simple callback that is py2 compatible | 14:42 |
| Jmainguy | I like running ansible from laptop, others on team use a server to do it, I wanna record all our runs regardless of where we run it from | 14:42 |
| Jmainguy | so if the callback plugin supports python2 that would be ideal | 14:43 |
| Jmainguy | or the ability to support python2 as an option | 14:43 |
| dmsimard | yes | 14:43 |
| dmsimard | whether we like it or not, py2 is going to be around for a while | 14:44 |
| Jmainguy | yup =) | 14:44 |
| Jmainguy | we were so excited when centos5 finally actually died, so we didnt have to do python 2.4 anymore | 14:44 |
| dmsimard | Jmainguy: setting up 0.x should hopefully be straightforward -- the web ui, the callback and the CLI all rely on having access to the database | 14:46 |
| dmsimard | so as long as everything is set up to point to the right place it should work fine | 14:47 |
| Jmainguy | sweet, ty ty, I will give that a shot | 14:47 |
| dmsimard | apollo13: I'm looking at the integration tests and I was wondering if an ara_api ansible module would be helpful even if just to test our things, pseudocode: https://gist.github.com/dmsimard/318682878b90041b4b0e47a11c67aa00 | 15:15 |
| apollo13 | good idea imo | 15:15 |
| *** etienne has joined #ara | 15:16 | |
| dmsimard | also, another thing that I need to look into | 15:16 |
| dmsimard | is that there are callback errors during the integration tests but they aren't fatal to the actual playbook so they don't fail the task that runs the playbook | 15:17 |
| openstackgerrit | Merged openstack/ara-server master: Renamed most of config to settings and dropped the default_ prefix. https://review.openstack.org/631034 | 15:17 |
| dmsimard | I feel like it would be hacky to register the output of the task and then grep for like http 400 or whatever | 15:17 |
| openstackgerrit | Guillaume Vincent proposed openstack/ara-web master: Add tasks plays and records number in Playbook summary https://review.openstack.org/631248 | 15:22 |
| openstackgerrit | Merged openstack/ara-web master: Add tasks plays and records number in Playbook summary https://review.openstack.org/631248 | 15:56 |
| Jmainguy | got mysqldb running on kube, got my ansible output to go to it | 15:59 |
| Jmainguy | ran webserver on lappy, first click on reports shows my run, if I reload, or try and click on that run, webserver crashes | 15:59 |
| Jmainguy | just running the devleopment flask one | 15:59 |
| Jmainguy | https://paste.fedoraproject.org/paste/rq0Fplupzj4M3uoQ9rdIgw | 15:59 |
| Jmainguy | all subsequent loads of http://127.0.0.1:9191/ show error 500 | 16:01 |
| dmsimard | Jmainguy: ouch, not sure what is going on there | 16:03 |
| dmsimard | is the playbook large by any means ? | 16:03 |
| dmsimard | Two things to look at is that there's been reports of the UI taking more memory than it should and when using mysql, need to ensure a reasonable max_connections | 16:05 |
| Jmainguy | nah small play, the data seems fine in mysql db, prolly somethign wrong with my mysql setup | 16:05 |
| dmsimard | export ARA_LOG_LEVEL=DEBUG might have a bit more info | 16:06 |
| dmsimard | (from where the webapp is running) | 16:06 |
| dmsimard | there might be a delta between what is logged in the foreground and in ~/.ara/ara.log | 16:07 |
| Jmainguy | yeah its something with my mysql, if I run mysql in a container locally it works great, ty sir | 16:19 |
| Jmainguy | first time using it, the display is amazing, really nice work. Love all the detail, I can even dive into the tasks and how they were written | 16:24 |
| Jmainguy | really top notch work | 16:24 |
| Jmainguy | each of my tasks when run by ansible, show this on the cli, any way to get rid of this? | 16:26 |
| Jmainguy | https://paste.fedoraproject.org/paste/GlsdpjVTSQUckReLNAimxA | 16:26 |
| dmsimard | Jmainguy: haven't seen that before, would need to look into it | 16:31 |
| apollo13 | dmsimard: if I were to write a simple permission framework for 1.0 that only decides between noaccess/readonly/rw and supports basic auth (from a static config) would that something you'd be okay with? | 16:34 |
| apollo13 | and no fine grained permissions yet, just api wide | 16:34 |
| dmsimard | apollo13: I think the first step is to support basic authentication in the client so people can use a web server for security | 16:42 |
| apollo13 | dmsimard: basic auth comes for free with DRF | 16:42 |
| apollo13 | so we can support it at that level | 16:43 |
| dmsimard | Not the python client though | 16:43 |
| apollo13 | I know | 16:43 |
| apollo13 | well rather free | 16:43 |
| apollo13 | requests makes it easy | 16:43 |
| dmsimard | Yeah | 16:43 |
| apollo13 | but I could do you a PR with the server side stuff so you can configure and test the client accordingly without having to setup a webserver | 16:43 |
| dmsimard | For the permission model, we probably need to have a wider discussion | 16:43 |
| apollo13 | for 1.0 read/write on the whole api should suffice | 16:44 |
| apollo13 | I think | 16:44 |
| apollo13 | everything more finegrained will be interesting | 16:44 |
| dmsimard | yes, agreed on either read/write or read only | 16:48 |
| dmsimard | I'm not interested in the business of managing multitenancy, hierarchical and endpoint-granular permissions | 16:50 |
| dmsimard | either you have access or you don't, IMO | 16:50 |
| apollo13 | I currently have API_READ_LOGIN_REQUIRED and API_WRITE_LOGIN_REQUIRED as bool | 16:51 |
| dmsimard | doing otherwise would become complicated very quickly | 16:51 |
| dmsimard | apollo13: oh, that's the DRF built in you mentioned ? | 16:51 |
| apollo13 | no that is the setting and permission class I just added | 16:51 |
| apollo13 | for lack of better imagination | 16:51 |
| dmsimard | API is perhaps a bit redundant | 16:52 |
| apollo13 | I mean we can drop read for now and say we can always read | 16:52 |
| apollo13 | which would leave us with write_login_required | 16:52 |
| dmsimard | both are probably okay if people don't want to manage it through a webserver I guess | 16:52 |
| dmsimard | they would both default to false, yeah ? | 16:52 |
| apollo13 | yeah | 16:53 |
| apollo13 | that said, api with basic auth for read might be fun for the web interface to implement | 16:53 |
| apollo13 | cause after all the client makes those requests directly | 16:53 |
| apollo13 | I think in the longrun we'd probably add support for remote_user | 16:53 |
| dmsimard | the javascript client uses axios which supports basic auth | 16:53 |
| dmsimard | https://stackoverflow.com/questions/44072750/how-to-send-basic-auth-with-axios | 16:53 |
| apollo13 | yeah sure, but where to get the auth data from? | 16:54 |
| apollo13 | actually the user could enter it and the web iface would store it locally, so far so good | 16:54 |
| dmsimard | apollo13: config file https://github.com/openstack/ara-web/blob/master/public/config.json | 16:54 |
| apollo13 | ok, I'll drop api_ but leave the two | 16:54 |
| apollo13 | sure but those credentials are then __public__ | 16:54 |
| apollo13 | you might as well have none :D | 16:54 |
| dmsimard | oh, wait | 16:54 |
| dmsimard | are you talking about the auth from the web to the api | 16:55 |
| dmsimard | to just to the web | 16:55 |
| dmsimard | or just to the web* | 16:55 |
| apollo13 | well axious runs on the client in the webbrowser | 16:55 |
| apollo13 | so yeah I am talking about how the web dashboard authenticates itself to the api if one sets read_login_required to true | 16:55 |
| dmsimard | not sure | 16:56 |
| apollo13 | well the javascript could ask the user for credentials and then put them in localstorage | 16:56 |
| apollo13 | either way that is something we can adjust along the way | 16:57 |
| dmsimard | I guess how it would work is that the web UI would have a login screen | 16:57 |
| dmsimard | the auth is sent to the API server | 16:57 |
| dmsimard | API returns a token or something like that | 16:57 |
| apollo13 | right, and now you also have to manage tokens and renewals and whatnot | 16:57 |
| apollo13 | something for post 1.0 :D | 16:57 |
| dmsimard | yeah... | 16:58 |
| dmsimard | I mean, for the time being | 16:58 |
| dmsimard | we can leave read_login_required to false and have users implement basic auth on the web ui instead | 16:58 |
| apollo13 | yeah | 16:58 |
| dmsimard | leaving the API private or something | 16:58 |
| apollo13 | no you cannot leave it private | 16:58 |
| apollo13 | because the webbrowser needs to reach it | 16:58 |
| dmsimard | oh, right | 16:58 |
| apollo13 | but, if one puts the whole web & api behind a webserver with basic auth and we enable the remoteusermiddleware in django we can handle that… | 16:59 |
| apollo13 | will be two PRs | 16:59 |
| apollo13 | one for auth as such and the other for external auth | 16:59 |
| dmsimard | sure | 17:00 |
| dmsimard | feel free to send something, we'll rope in gvincent for input | 17:01 |
| apollo13 | as long as he doesn't want to use jwt :D | 17:01 |
| dmsimard | eh, I know zuul has a spec for implementing jwt | 17:01 |
| dmsimard | I know nothing about jwt though | 17:01 |
| dmsimard | https://github.com/GetBlimp/django-rest-framework-jwt | 17:02 |
| dmsimard | ¯\_(ツ)_/¯ | 17:02 |
| apollo13 | I know, but still | 17:03 |
| gvincent | tldr? need authentication? | 17:06 |
| dmsimard | gvincent: we were wondering how the web interface might eventually authenticate against the API | 17:12 |
| dmsimard | gvincent: we're not going full blown enterprise RBAC, just basic auth for now | 17:12 |
| dmsimard | apollo13 will submit a proposal we can talk about | 17:14 |
| dmsimard | gvincent: to be clear, this isn't a blocker for 1.0 | 17:14 |
| dmsimard | but it's something we need to think about to make things easier later | 17:15 |
| openstackgerrit | Florian Apolloner proposed openstack/ara-server master: Added very basic read/write authentication. https://review.openstack.org/631285 | 17:36 |
| openstackgerrit | Florian Apolloner proposed openstack/ara-server master: Added support for external auth via REMOTE_USER provided by webservers. https://review.openstack.org/631287 | 17:41 |
| *** themroc has quit IRC | 17:44 | |
| gvincent | JWT or basic auth will work | 17:44 |
| *** herald85 has quit IRC | 18:14 | |
| *** gvincent has quit IRC | 18:46 | |
| *** gvincent has joined #ara | 19:23 | |
| *** pabelanger has joined #ara | 19:57 | |
| pabelanger | dmsimard: is there a way to run ara generate html, but just for the last run only or a specific run | 19:57 |
| dmsimard | pabelanger: --playbook | 19:58 |
| dmsimard | pabelanger: ara generate html --playbook <uuid> | 19:59 |
| dmsimard | you can find uuids with ara playbook list | 19:59 |
| pabelanger | k, testing | 20:00 |
| dmsimard | if you need the currently running playbook uuid during the execution, it can also be done | 20:00 |
| Jmainguy | dmsimard: looks like if it loses connection to mysql, it never reconnects | 20:00 |
| dmsimard | pabelanger: note that there's a bug when specifying a playbook id https://storyboard.openstack.org/#!/story/2001013 | 20:00 |
| Jmainguy | dmsimard: does it open one connection when flask starts, and then never again? | 20:01 |
| dmsimard | this works: ara generate html /tmp/test --playbook 7ccdd306-1414-4d11-aa53-108035cb712b | 20:01 |
| dmsimard | this doesn't: ara generate html --playbook 7ccdd306-1414-4d11-aa53-108035cb712b /tmp/test | 20:01 |
| dmsimard | Jmainguy: this is abstracted heavily by flask and sqlalchemy but ara does expose settings to tweak some of the configuration | 20:02 |
| dmsimard | Jmainguy: SQLALCHEMY_POOL_SIZE, SQLALCHEMY_POOL_TIMEOUT and SQLALCHEMY_POOL_RECYCLE | 20:02 |
| dmsimard | https://ara.readthedocs.io/en/stable/configuration.html#parameters-and-their-defaults && http://flask-sqlalchemy.pocoo.org/2.3/config/#configuration-keys | 20:03 |
| Jmainguy | ty | 20:03 |
| Jmainguy | pool-size for a nub like me, is number of connections it maintains in a pool? | 20:03 |
| pabelanger | dmsimard: thanks, I think ara generate html --playbook=08c65c0e-9334-4e53-a718-2190c1474503 /tmp/x123 worked | 20:03 |
| dmsimard | Jmainguy: According to https://docs.sqlalchemy.org/en/latest/core/pooling.html#connection-pool-configuration, I think so | 20:06 |
| Jmainguy | u so good to me | 20:06 |
| pabelanger | dmsimard: how do you get the UUID from current execution? | 20:12 |
| dmsimard | pabelanger: something like this: https://gist.github.com/dmsimard/0335667e69539c1dd22a169cbd43e616 | 20:13 |
| dmsimard | It's not pretty but it's what we've got for now | 20:14 |
| pabelanger | k | 20:14 |
| dmsimard | tl;dr, registering an ara_record task provides the playbook_id | 20:14 |
| pabelanger | yah | 20:15 |
| dmsimard | if you use ara_record, you'll need to supply an action_plugins path | 20:15 |
| dmsimard | $ python -m ara.setup.action_plugins | 20:16 |
| dmsimard | /usr/lib/python2.7/site-packages/ara/plugins/actions | 20:16 |
| dmsimard | pabelanger: note that it's an action plugin and it should ideally only ever run once (i.e, not a task that runs on all the hosts) | 20:16 |
| pabelanger | ara playbook list | grep playbooks/site.yaml | tail -1 | cut -d' ' -f2 | 20:18 |
| pabelanger | writing up job now | 20:18 |
| dmsimard | pabelanger: yeah.. that's about what the integration tests are running today.. | 20:18 |
| dmsimard | https://github.com/openstack/ara/blob/74557137f8755efe71067a4650597011e40f3c53/run_tests.sh#L142 | 20:19 |
| pabelanger | cool thanks | 20:19 |
| pabelanger | this is to finish up nested zuul deployments | 20:19 |
| pabelanger | eg | 20:19 |
| pabelanger | sf.io zuul, deploying zuul control plane in vexxhost | 20:20 |
| pabelanger | https://object-storage-ca-ymq-1.vexxhost.net/v1/a0b4156a37f9453eb4ec7db5422272df/logs/69/69b77a3434320b75bf5d6c711d0c5837023ebe97/post/windmill-config-deploy/c8f0bf2/ | 20:20 |
| dmsimard | rock on | 20:20 |
| *** gvincent has quit IRC | 20:26 | |
| pabelanger | okay, testing now | 20:34 |
| Jmainguy | dmsimard: to ease your troubled mind, once I removed quota from openshift its running like a dream, so who knows | 21:10 |
| Jmainguy | gonna write a bloggy type post tomorrow on how to use laptops and servers to connect to a central ara on openshift. its slick =) | 21:11 |
| Jmainguy | womp, I spoke too soon, it died with No handlers could be found for logger "sqlalchemy.pool.QueuePool" now | 21:13 |
| ara-slack | nj.chandu: @dmsimard I have found the solution to the problem. The tool is just amazing. | 21:26 |
| ara-slack | Do you think it would make sense to add a note that Hosting statically generated reports is not very efficient at a large scale. in this page https://ara.readthedocs.io/en/stable/usage.html?highlight=ara%20generate%20html? If so, please let me know I will raise a PR with the changes. | 21:26 |
| Jmainguy | added --processes 1 --threaded to runserver and its stableish atm | 21:26 |
| dmsimard | Jmainguy: 0.x is still supported from a bugfix standpoint so if we find anything we can fix and release | 22:27 |
| dmsimard | Jmainguy: happy to share a bloggy type post, just let me know when it's up :) | 22:28 |
| dmsimard | @nj.chandu Yeah I think we can do that -- I can walk you through the process of sending the patch if you'd like, we're not using pull requests right now | 22:29 |
| dmsimard | there's doc on contributing here: https://ara.readthedocs.io/en/stable/contributing.html or a tl;dr version I gave a while back https://github.com/openstack/ara/pull/12#issuecomment-433119539 | 22:31 |
| pabelanger | dmsimard: yay, it worked: https://object-storage-ca-ymq-1.vexxhost.net/v1/a0b4156a37f9453eb4ec7db5422272df/logs/20/2035ce05811e66986b9f3e1f94990652fafe8ff9/post/windmill-config-deploy/d5faf2f/logs/ara-report/ | 22:59 |
| pabelanger | that is production control plane | 22:59 |
| dmsimard | pabelanger: cool, but that's also 1200+ html files :( | 23:00 |
| *** apollo13 has quit IRC | 23:00 | |
| pabelanger | dmsimard: yah, swift | 23:00 |
| pabelanger | dmsimard: still using sqlite for it, need to get up a dedicated db server, then likely launch ara wsgi app and expose ara.poctrong.xyz | 23:01 |
| *** apollo13 has joined #ara | 23:02 | |
| pabelanger | ara.poctron.xyz* | 23:02 |
| dmsimard | pabelanger: I hope that 1.0 will be in a somewhat usable state soon (tm) | 23:03 |
| dmsimard | We still don't have an "ara-report" implementation but it's on my plate | 23:03 |
| Jmainguy | can confirm, neither domain works for me | 23:06 |
| Jmainguy | but now I know that ATT is intercepting my dns requests despite me specifying dns servers, so theres that | 23:06 |
| *** etienne has quit IRC | 23:07 | |
| *** sshnaidm is now known as sshnaidm|afk | 23:43 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
