Skip to content

Latest commit

 

History

History
85 lines (63 loc) · 4.27 KB

README.md

File metadata and controls

85 lines (63 loc) · 4.27 KB

Monitoring a live deployment of Hedera Mirror Node

This code runs on an external server outside of the Hedera mirror node deployment, and periodically polls the REST APIs exposed by the Hedera mirror node to ensure that the deployed APIs are working. It also provides a simple dashboard to monitor the status.

Monitoring APIs:

A process runs that periodically polls the APIs exposed by the deployment of a Hedera mirror node. It then checks the responses using a few simple checks for those APIs. The results of these checks are exposed as a set of REST APIs by this monitoring service as follows:

API HTTP return code Description
/api/v1/status 200 (OK) Provides a list of results of all tests run on all servers
/api/v1/status/{id} 200 (OK) If all tests pass for a server, then it returns the results
4xx If any tests fail for a server, or if the server is not running, then it returns a 4xx error code to make it easy to integrate with alerting systems

Monitoring Dashboard:

A dashboard polls the above-mentioned APIs and displays the results.

Quickstart

Requirements

  • List of addresses of Hedera mirror nodes that you want to monitor
  • An existing topic ID of the target Hedera Mirror node environment if you want to run topic message tests
  • An external server where you want to run this code to monitor the mirror node. You will need two TCP ports on the server.
  • npm
  • PM2
git clone [email protected]:hashgraph/hedera-mirror-node.git
cd hedera-mirror-node/hedera-mirror-rest/monitoring

To install the dependencies and configure monitor_apis:

cd monitor_apis
npm install
vi config/serverlist.json // Insert the mirror node deployments you want to monitor

To customize per-resource configuration:

  • set enabled to true or false to enable / disable tests for a resource
  • freshnessThreshold (in seconds) for balance, transaction, topic, and token (balances of a token) can be adjusted independently if needed. Set to 0 to disable freshness check for a resource
  • intervalMultiplier for all resources. The tests for a resource will run every interval * intervalMultiplier seconds. For stateproof, it defaults to 10, so the tests run at a lower frequency to reduce cost
  • limit threshold for account, balance, transaction, topic, and token can be adjusted independently if needed, e.g.,for environments with lower traffic volume. For token, tokenBalancesLimit is the limit for a token's balance query and tokensLimit is the limit for the token discovery query
  • set topic.topicId to an existing topic ID of the target environment. If not set, topic message tests will be skipped
  • set token.tokenId to an existing token ID of the target environment in case it's needed to pass certain token tests, e.g., tokenBalancesLimit for the token. If not set, the tests will use the first token from the token discovery API response

See default.serverlist.json for a full list of configuration properties.

To run the monitor_apis backend:

PORT=3000 pm2 start server.js --node-args="--experimental-specifier-resolution=node"

The server will start polling Hedera mirror nodes specified in the config/serverlist.json file. The default interval to populate the data is 60 seconds. You can verify the output using curl http://<host>:<port>/api/v1/status command.

To run the dashboard (from hedera-mirror-rest/monitoring directory):

cd monitor_dashboard
echo '{"monitorAddress": "localhost:3000"}' > config.json
pm2 serve . 3001 // Serve the dashboard html pages on another port

Using your browser, connect to http://<host>:<port>/index.html