voting-alpha

Repo for storing all the resources for the Stack.

Contents:

• /ethereum - smart contracts, etc
• /stack - AWS CloudFormation stack resources

Website and Livestream

https://flux.vote

deps

Note: only linux supported atm

system

• Ubuntu deps: sudo apt install python3 python3-pip zip git build-essential
• Recommended: Install pyenv and pyenv-virtualenv

The recommended things are for tests, etc, used automatically via ./manage

local repo

• ./setup/pyenv.sh -- ensures python and venv are set up
  • DEPRECATED: pyenv install -- installs python matching .python-version if it isn't installed. (If there's any trouble with the .python-version in the repo, run pyenv install 3.6.9)
  • DEPRECATED: pyenv virtualenv venv-voting-alpha
  • DEPRECATED: If you don't have pyenv-virtualenv shell integration then you'll need to activate the venv yourself / manually. Shell integration does this automagically.
• Python deps (for ./manage): pip install -r _manager_lib/requirements.txt
• To install python deps for lambdas: ./manage pip all

deploying and ./manage

WARNING: The ./manage util will use your default AWS .credentials or AWS_PROFILE if present. You'll also need export AWS_DEFAULT_REGION=ap-southeast-2 if region isn't specified in your .credentials.

run ./manage --help for cli commands, or ./manage CMD --help

Env Vars

For each env edit the vars in .env.(name) -- e.g. .env.dev

Additionally copy .in.sample to .in and update the vars; source .in if you need. (zsh-autoenv will automatically source .in and .out files for you)

• S3_DEV_BUCKET=
• TEST_SUBDOMAIN=
• VOTING_DOMAIN=<the domain hosted zone you have in route53; subdomains are okay>

Deployment & Pre-reqs

• make sure you've installed python deps for lambdas as per above (works on linux; on mac you need to use docker which is currently disabled - though code is still in ./manage)
• ./manage pip -- installs deps for custom resources and things
• ./manage deploy-macros -- run this after pip but before stack deployment
• ./manage sync-env dev -- deploys the main cfn stack -- will fail if deploy-macros has not been run

To deploy two stacks along side eachother the following vars must be unique:

• STACK_NAME
• NAME_PREFIX
• SUBDOMAIN

You can edit .env.dev, etc to keep them neat

Update nested stacks directly (useful during dev)

• example ./manage deploy-ns ./stack/nested/sv-members-app.yaml flux-dev-chain-rMembersApp-1EPU6A6JKYGSX

Tests

• ./manage test chaincode
• ./manage test api-members
  • note: ensure a containerized openethereum node and RPC endpoint is running before hand: ./manage tools node-up

Curl based, can be run from CLI

• #: curl -v -X POST -d '{"method":"#","params":{"ethereumAddress":"0xe9b0722856FA4EfF597Bd2123C00a4'$(date +%s)'", "unsafeChecksum": true}}' https://api.blockchain.suzuka.flux.party/members/api
• ballot_publish: curl -v -X POST -d '{"method":"ballot_publish","params":{"specHash":"0x0f74cba1f103c8636cb87f241cd27a6ef1776185fc956acdb999cd'$(date +%s)'"}}' https://api.blockchain.suzuka.flux.party/members/api
  • note: this handler might need more data in future, like a ballotSpec param.

DEPRECATED

• export OFFSET='1' -- this is just used to allow us to create more than one stack in parallel for dev purposes (by changing offset)
• ./manage --offset $OFFSET deploy --watch --step 0 voting-dev -- the --step N param allows us to incrementally deploy, which is useful for testing and getting the stack to a state that makes subsequent deployments faster.
• ./manage --offset $OFFSET deploy --watch --use-existing --step 1 voting-dev
• ./manage --offset $OFFSET deploy --watch --use-existing --step 2 voting-dev
• ./manage --offset $OFFSET deploy --watch --use-existing voting-dev -- --step N can be omitted for a full deploy; at the time of writing there are only 3 steps available.

Developing

• cnf-lint stack/**/*.yaml -- cfn-lint is installed as a python dep (and github action)
• ./manage deploy-ns [path/to/template.yaml] [stack_name] will deploy a nested stack only (useful when testing stacks that are late in the CFN deployment process)

Adding pypi packages to lambda functions

The target you specify must be one of:

• cr -- installs in stack/cr/common
• members -- stack/app/members
• app -- stack/app/common

Run ./manage pip TARGET pkg1 pkg2 pkg3 ... - this will install dependencies and update requirements.txt

license

Deciding on the default license for the repo is a WIP. Some files are licensed under apache (inherited from AWS templates I've altered). Suggestions welcome. See issue #1