monitoring

STOM

Useful links

• Starknet on-chain monitor generated docs.
• On-chain monitoring (OM) framework architecture docs in blueprints.
• OM framework generated docs.

Local development

Note: Previously, this monitor also wrote to Kafka, but the dependency on Kafka has been removed in order to simplify deployment. The kafka topics were unused anyway.

• Start an http server that mimics weiwatchers locally. It needs to export a json configuration file for feeds:

[
  {
    "name": "LINK / USD",
    "path": "link-usd",
    "symbol": "$",
    "heartbeat": 0,
    "contract_type": "numerical_median_feed",
    "status": "testing",
    "contract_address": "<CONTRACT_ADDRESS>",
    "multiply": "100000000",
    "proxy_address": "<PROXY_ADDRESS>"
  }
]

It also needs to export a json configuration for for node operators:

[
  {
    "id": "noop",
    "nodeAddress": [<NODE_OPERATOR_ADDRESS>]
  }
]

One option is to create a folder /tmp/configs and add two files feeds.json and nodes.json with the configs from above, then:

python3 -m http.server 4000

• Start STOM locally. You will need and RPC endpoint and the address of the LINK token. Make sure you cd ./monitoring.

STARKNET_RPC_ENDPOINT="<RPC_ENDPOINT>" \
STARKNET_NETWORK_NAME="devnet" \
STARKNET_NETWORK_ID="1" \
STARKNET_CHAIN_ID="1" \
STARKNET_READ_TIMEOUT="5s" \
STARKNET_POLL_INTERVAL="10s" #test \
STARKNET_LINK_TOKEN_ADDRESS="<LINK_TOKEN_ADDRESS>" \
HTTP_ADDRESS="localhost:3000" \
FEEDS_URL="http://localhost:4000/feeds.json" \
NODES_URL="http://localhost:4000/nodes.json" \
go run ./cmd/monitoring/main.go

• Check the output for the Prometheus scraper

curl http://localhost:3000/metrics