provable/ptokens-eth-syncer repository overview

⁠:arrows_clockwise: pBTC ETH Syncer

The Provable pBTC Token ETH syncing tool, designed to keep up a feed of ETH blocks to the enclave.

⁠Build with docker

./build-image ptokens-eth-syncer:latest-vanilla vanilla ghp_... -f Dockerfile.dev

⁠:point_right: Usage:

First install dependencies:

❍ npm i

Then run the tool via:

❍ node eth-syncer.js

For more information:

❍ node eth-syncer.js --help

❍ ETH Syncer:

  Copyright Provable Things 2020
  Questions: [email protected]

❍ Info:

  A tool to watch for new ETH blocks & feed them to a pToken enclave.

❍ Usage:
  eth-syncer.js
  eth-syncer.js --help
  eth-syncer.js --version
  eth-syncer.js getInitBlock [--file=<path>]
  eth-syncer.js getSubmissionMaterial <blockNum> [--file=<path>]

❍ Commands:

  <no-command>          ❍ Start the ETH syncer.
  getSubmissionMaterial ❍ Get the submission material for a given block number.
  getInitBlock          ❍ Get a block that can be used to initialize the ETH side of a pToken enclave.

❍ Options:
    --help              ❍ Show this message.
    --version           ❍ Show version information.
    <blockNum>          ❍ The number of the bitcoin block in question.
    --file=<path>       ❍ Path to write the output to.

⁠:page_with_curl: Notes:

This tool polls the ETH blockchain for new blocks and upon finding one, assembles that block, all of it's transactions into a JSON, which it then submits to the pBTC enclave.

⁠:wrench: Configurables:

A guide to the various configurable fields in the config.js file in the crate root:

• ENCLAVE_BINARY_NAME Name of the enclave binary.

• MAX_ENCLAVE_RETRIES Max number of times to try and communicate with the enclave binary.

• ENCLAVE_QUERY_MAX_TIME Max amount of time (in milliseconds) to wait for an enclave query to execute.

• DB_LOCK_SLEEP_TIME: The amount of time in milliseconds the tool will sleep when encountering a database lock in the enclave before trying its previous action once more.

• SLEEP_TIME: The amout of time in milliseconds the tool will sleep between successfully finding and submitting a BTC block.

• MONGO_DATABASE_NAME: The name of the Mongo database accessible via the mongod daemon.

• ETH_ENDPOINT: An endpoint for a full ethereum node on the desired ethereum network.

• MONGO_COLLECTION_NAME: The name of the collection in the aforementioned database.

• MONGO_DB_URL: The URL to access the above defined database.

• RELATIVE_PATH_TO_ENCLAVE: Per title: a path to the enclave binary, relative to this configuration file.

• GET_TX_RECEIPTS_CALL_TYPE: Depending on the endpoint in use, you can select 4 different RPC calls methods:

  • parity: perform parity_getBlockReceipts to download tx receipts

  • erigon: perform eth_getBlockReceipts to download tx receipts

  • single: (a.k.a. batch JSON RPC call) perform a single JSON RPC call to download all the block receipts

  • multiple: (a.k.a. non-batch json RPC call) performs a call for each tx receipt inside the block

  • auto: use parity if available otherwise use multiple

⁠:guardsman: Tests

To run the tests simply run:

❍ npx mocha

⁠:black_nib: To Do:

• Make PATH_TO_ENCLAVE configurable!
• More Tests!
• Factor out bit of maybeGetNextBlock... fxn.
• Catch ctrl-c & exit gracefully! (& in this case we can NOT delete the temp file EXCEPT when we catch a sigint)
• Some form of clean up in above case.
• Put main mongo db in state and close the connection if we hit the main catch
• Handle block already in db error by incrementing num in state and continuing!
• Handle receipts root error by decrementing block num and continuing (then, with above, it should solve orphans, bad blocks etc)
• Handle NO_RESPONSE_FROM_NODE_ERR better by counting the retries or something before giving up. Then somehow send an alert that the node is down.
• Implement queueing system for getting blocks so we can submit them faster to the enclave.