-
These sets of playbooks were deployed with Ansible 2.6
-
You will need the following Ansible Galaxy playbooks installed:
ansible-galaxy install geerlingguy.java
-
Ansible uses the boto3 library for AWS access. That will automatically configure itself from several mechanisms — either use a tool such as
aws-vault
to manage the environment (recommended) or follow the boto3 configuration guide to ensure that your API keys are available. It is recommended that you use an AWS named profile (e.g.ndnp
) and letaws-vault
manage short term credentials with rotation. -
Create an EC2 SSH key-pair and make sure the
AWS_KEY_PAIR
environmental variable specifies it by name. -
Add
Chronam-deploy-aws-ansible-vault
to your system keychain:$ security add-generic-password -a ${USER} -s "Chronam-deploy-aws-ansible-vault"
-
Data has already been uploaded to S3
Before we can deploy to AWS we need to setup several environments so that we conform to A&A security requirements. See setup.sh
run deploy_<ENVIRONMENT>.sh
depending on where you want to deploy
- Load the very old titles that only exist in our archive ** https://s3.console.aws.amazon.com/s3/buckets/chronam-original-titles/?region=us-east-1&tab=overview
- pull and load the other titles
- load the essays
- load the holdings ** https://s3.console.aws.amazon.com/s3/buckets/chronam-holdings/?region=us-east-1&tab=overview
- then create a script to split up the batches based on release date. This was then used to load the batches on the various django servers we created early. This ensures that we don't try and load 2 batches from the same awardee at the same time which causes problems
- then we ran
django-admin release --reset
to make the release dates the same between systems - then we ran
django-admin dump_ocr
to generate all the OCR data - look at /status page and check against known good instance and tweak as necessary
- Ensure that
pre-commit
is installed (e.g.brew install pre-commit
orapt-get install pre-commit
) and that you have runpre-commit install
at least once