Skip to content

Latest commit

 

History

History
93 lines (75 loc) · 8.29 KB

README.md

File metadata and controls

93 lines (75 loc) · 8.29 KB

lido-dot-ksm-oracle

Oracle service for LiDo liquid staking. Collects staking parameters from relay chain and reports to the parachain liquid staking module.

How it works

  • Upon the start daemon determines the reportable era and retrieves the list of stash accounts to report for.
  • If no stash accounts were found, waits for the beginning of the next era. Otherwise, daemon starts collections staking parameters for each stash account separately, signs and sends transaction to the oracle contract.
  • After a report has been sent for all stash accounts, it moves on to waiting for the next era.

Requirements

  • Running relay chain node, from which the staking parameters are read.
  • Running parachain node with contracts deployed.
  • ABI for the OracleMaster contract (see above).
  • Python 3.7+
  • The application functions as a linux service under systemd or as a docker container.

Setup

python -m pip install --upgrade pip
pip install -r requirements.txt

Run

By default, the assets directory contains oracle ABI and the service tries to get it from the assets/oracle.json file, but you can change it as follows:

  • Clone lido-dot-ksm repository.
  • Run the brownie compile command.
  • Copy the contents of the build/contracts/OracleMaster.json with the 'abi' key to assets/oracle.json or change the ABI_PATH environment variable (see below).

The service receives its configuration parameters from environment variables. Export required parameters from this list and start the service:

./oracleservice/start.py

To stop the service, send a SIGINT or SIGTERM signal to the process.

Full list of configuration options

  • WS_URL_RELAY - WS URL of relay chain node. Required.
  • WS_URL_PARA - WS URL of parachain node. Required.
  • CONTRACT_ADDRESS - OracleMaster contract address. Required. Example: 0xae7ab96520DE3A18E5e111B5EaAb095312D7fE84.
  • ORACLE_PRIVATE_KEY_PATH - The path to the oracle private key file. Required if the ORACLE_PRIVATE_KEY parameter is not specified.
  • ORACLE_PRIVATE_KEY - Oracle private key, 0x prefixed. Used if there is no file with the key. Required if the ORACLE_PRIVATE_KEY_PATH is not specified.
  • ABI_PATH - Path to ABI file. The default value is assets/oracle.json.
  • GAS_LIMIT - The predefined gas limit for composed transaction. The default value is 10000000.
  • MAX_PRIORITY_FEE_PER_GAS - The maxPriorityFeePerGas transaction parameter. The default value is 0.
  • FREQUENCY_OF_REQUESTS - The frequency of sending requests to receive the active era in seconds. The default value is 180.
  • MAX_NUMBER_OF_FAILURE_REQUESTS - If the number of failure requests exceeds this value, the node (relay chain or parachain) is blacklisted for TIMEOUT seconds during recovery mode. The default value is 10.
  • TIMEOUT - The time in seconds the failure node stays in the black list in recovery mode. The default value is 60.
  • ERA_DURATION_IN_SECONDS - The duration of era in seconds. Needed for setting the SIGALRM timer. The default value is 180. Required.
  • ERA_DURATION_IN_BLOCKS - The duration of era in blocks. The default value is 30. Required.
  • SS58_FORMAT - The default value is 42. Required.
  • TYPE_REGISTRY_PRESET - The default value is kusama. Required.
  • REST_API_SERVER_IP_ADDRESS - REST API server IP address. The default value is 0.0.0.0.
  • REST_API_SERVER_PORT - REST API server port. The default value is 800.
  • LOG_LEVEL_STDOUT - Logging level of the logging module: DEBUG, INFO, WARNING, ERROR or CRITICAL. The default level is INFO.
  • ORACLE_MODE - If the value is DEBUG, the oracle will not send transactions, but only prepare a report.
  • ERA_UPDATE_DELAY - The maximum delay in seconds with which an era can be updated before the service stops working. The default value is 360.
  • ERA_DELAY_TIME - The maximum delay in seconds with which an era can be updated comparing to the OracleMaster before the service stops working. The default value is 600.
  • WAITING_TIME_BEFORE_SHUTDOWN - Waiting time in seconds before shutdown the service. The default value is 600.

Prometheus metrics

The Prometheus exporter provides the following metrics.

name description frequency
agent
Info
The address of the connected relay chain node Each reconnection
is_recovery_mode_active
Gauge
Is oracle-service in recovery mode or not: 1, if the recovery mode, otherwise - the default mode Starting and ending recovery mode
active_era_id
Gauge
Active era index Every change of era
last_era_reported
Gauge
The last era that the Oracle has reported After completing a sending of reports for an era
last_failed_era
Gauge
The last era for which sending the report ended with a revert During sending a report for an era
previous_era_change_block_number
Gauge
Block number of the previous era change Every change of era, if at least one stash account was found
time_elapsed_until_last_era_report
Gauge
The time elapsed until the last report from the UNIX epoch in seconds After each successful sending of a report
total_stashes_free_balance
Gauge
Total free balance of all stash accounts for the era Every time a new report is generated
oracle_balance
Gauge
Oracle balance Every change of era
tx_revert
Histogram
Number of failed transactions Every unsuccessful sending of a report
tx_success
Histogram
Number of successful transactions Every successful sending of a report
para_exceptions_count
Counter
Parachain exceptions count Every exception in the parachain
relay_exceptions_count
Counter
Relay chain exceptions count Every exception in relay chain

REST API

Prometheus metrics are provided by URL '/metrics'.

The status of the oracle is provided by URL '/healthcheck'. The following states are possible:

  • not working - the service is in parameter preparation mode;
  • starting - the service is starting but has not yet started monitoring the event;
  • monitoring - the service is monitoring the event;
  • processing - the service is preparing data for a report or sending a transaction;
  • recovering - the service is in recovery mode.