Kevin/load test cleanup

[kevsul]

Overview

Clean up and refactor load test code, making the scenarios more modular and composable.

Changes

Describe your changes and implementation choices. More details make PRs easier to review.

• Add FundAccount scenario
• Add SpendEthUtxo scenario
• Refactor duplicate submit_transaction code
• Allow a delay before sending transactions
• Allow gas_price be configurable

[coveralls]

Coverage remained the same at 77.983% when pulling 1ac3b15 on kevin/load-test-cleanup into 97e2aa5 on master.

[boolafish]

Some high level questions and comment:

1. What problem has you bumped so we start to configure gas_price? (Just curious, not saying it really needs justification)
2. Same, why for the delay? Seems like you must encountered some instability?
3. Might consider to add the essential config values in each scenario to their documents.

[boolafish]

n00bie comment, but would it be more consistent for public api to return {:ok, data} instead?

[kevsul]

Hmm, good question.

The way I see it is that this function never returns an {:error condition, so returning {ok, is unnecessary. That is, the function will either successfully return the outputs of a tx or it will fail, but it won't return an :error that can be handled.

I'm happy to be taught otherwise though...

[kevsul]

1. What problem has you bumped so we start to configure gas_price? (Just curious, not saying it really needs justification)

Mainnet and Rinkeby gas prices are higher than Ropsten.

2. Same, why for the delay? Seems like you must encountered some instability?

No, this is so that you can control how fast the txs are sent. Setting this to 0 means send txs as fast as possible, but for example with a long running test we want to send txs at a slower rate

3. Might consider to add the essential config values in each scenario to their documents.

Yes, good idea