This repository has been archived by the owner on Jul 14, 2020. It is now read-only.
More info in log files: source address, machine, tag, timestamp #17
Assignees
Labels
Comments
|
AnotherKamila
added a commit
that referenced
this issue
Oct 10, 2019
AnotherKamila
added a commit
that referenced
this issue
Oct 10, 2019
* Add teams folder back * Add generation of configs directly from webserver * Add test files * Minor * Rename destination address file * Force directories to show up and ignore the content * Remove secrets from code :D This is a public repo. Yes, nobody cares, but still :D Since you appear to be using pipenv here, you can just add a .env file with the variables and it will be used automatically. I will add an example .env file and document it in the README. * add example .env file and document it * Minor refactor * Start implementing new webserver * Secret for management not needed * Implement webserver according to new specification Still some things left to be fixed * Update webserver dependencies * Add wrapper for web requests * Refactor paths of resources * Minor refactor and improvements Add the multithreaded processing * Ignore pycache * I have no idea if this works. TODO rebase this commit :D * README and bash wrapper * Load the secrets directly from the env * Refactor the bash wrapper * Fix bug from new script * Add test of one round * Remove useless sleep and test file * Remove unused * Default submission file to make buildbot happy :) * all the buildbot stuff in one huge commit, because I'm horrible * add ansible-inv-to-infrastructure.py script * Hack to make buildbot happy All the sources have a runnable python script, even the ones without player * Idempotent /prepare * Prepare for prod deployment, act 1: Managing workers * Sanitize team names (with regex) * update manage_workers script to create remote workers * downgrade to Python 3.5 in master Pipfile because old ubuntu is old * Config generation to support for minimum of src and dst transfer size * Add influxdb anb credentials to environment * Add influx db name and address to .env * Move scoring under webserver and fix scoring bug * Add scoring to /finish and test-run * Push to influx * change constants :D * webserver: downgrade to python 3.5 because old ubuntu is old * Fix to api call * Revert "webserver: downgrade to python 3.5 because old ubuntu is old" This reverts commit b00b0ce. * master: add a way to quickly toggle running rounds * Hack in round configs for the clients * forgot to add a file * Check that the webserver handles empty scores * master: if not RUN wip * master: maybe hopefully wire things together? we'll see what breaks * bla * Changes to force users to export token * blah * Fix bug in timestamp * || true :D * add shittiness scaling * stuff * commit changes made in prod :D * commit changes made during test * Implement Kamilal's code review * write some documentation :D * Create ./clearstate.sh scripts. Also bump python version for master to 3.6. * Remove `cleanup_dir()` calls from webserver startup. Now the webserver keeps state on restart by default. * fix example.env in webserver * random fixes * Add dummy worker and enable toggling rounds. Closes #7, #23. * also disable rounds when clearing state * Add machine info to log: src addr and hostname pokes at #17 * small fixes
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
"Tag" would be a thingy identifying a submit that the webserver would create and echo on submit.
The rest is just in infrastructure.
The text was updated successfully, but these errors were encountered: