This documentation is a work in progress. It needs to be adapted for use by persons outside of the CSS Lab

Preflight Checklist

Setup

[ ] create a GH repo to receive data, invite deliberation-machine-user to be a collaborator on that repo with push permissions
[ ] set up or identify an appropriate S3 bucket and region

[ ] visit https://app.terraform.io/app/css-lab-deliberation/workspaces/ and enter the workspace for the server you want to lanch
[ ] in the "variables" tab, update the deliberation_empirica_tag to the latest version from https://github.com/Watts-Lab/deliberation-empirica/pkgs/container/deliberation-empirica. This is the value of the docker image tag, everything following the colon.
[ ] check that the cpu and memory are set for much larger than we will need. A 2048 cpu and 8192 memory were overkill for a group of ~80. More computer is always cheaper than a failed run.
[ ] start a new run, and apply the changes, checking that there are no run errors

[ ] visit CloudWatch on the AWS console and inspect the server logs. You should see:
[ ] the correct Container Image Version Tag
[ ] the correct Subdomain
[ ] redacted values for the Daily APIKEY, Qualtrics Token,
[ ] github authentication succeeded
[ ] no error messages
[ ] a message that includes "serve: server started"

[ ] develop and test treatment files locally
[ ] push treatment files to github, and wait 5-10 mins to make sure they are up to date on the CDN
[ ] check the treatment file is correct by visiting it on the CDN, for example: https://s3.amazonaws.com/assets.deliberation-lab.org/projects/example/treatments.demo.yaml (update with the relevant path)

[ ] Visit the AWS S3 Bucket list, and get the bucket name and region you want video files to save to, and enter these in the config
[ ] set the data repo and prereg repo to a test folder in the repo you want the data to get pushed to
[ ] set the batch name to indicate that this is a test or demo run
[ ] set cdn to prod
[ ] set centralPrereg to false

On the production study admin page

[ ] write a config json file for the test, e.g. test.config.json, using deliberation lab docs for specific options.
Use batchname "test" or similar
Use the same videoStorage bucket you intend to use in the study. Check that the bucket name and region are all correct
Use the same repo you intend to use in the study, but use a different folder e.g."test"
set "centralPrereg" to false
[ ] visit the study subdomain admin page e.g. study.deliberation-lab.org/admin
[ ] Check that the AWS region specified in the batch config matches the region of the intended bucket.
[ ] set the data repo and prereg repo to a test folder in the repo you want the data to get pushed to
[ ] enter the batch config in the Empirica Admin Console > New Batch > Assignment Method: Custom text box, and hit "Create"
[ ] Wait 5 seconds for the validation to run
[ ] Check that the batch shows the green Start button
[ ] Visit the AWS server logs and look for any errors
[ ] Check for Initialized Batch ... at ...
[ ] Click Start batch button

visit the server directly through the participants' link, eg. study.deliberation-lab.org
[ ] check that you see (or don't see):
[ ] the webcam/mic check (for videocall studies)
[ ] the appropriate login instructions
[ ] the expected intro steps
[ ] walk through the experiment to the end steps
[ ] look at the AWS console logs for any errors

After everyone has finished

On the production study admin page

[ ] enter the batch config in the Empirica Admin Console > New Batch > Assignment Method: Custom text box, and hit "Create"
[ ] Wait 5 seconds for the validation to run
[ ] Check that the batch shows the green Start button
[ ] Visit the AWS server logs and look for any errors
[ ] Check for Initialized Batch ... at ...
[ ] Click Start batch button