2

To monitor nodes' healthiness, there are tools for Bitcoin and Ethereum (e.g. https://github.com/cubedro/eth-net-intelligence-api). Some of them are tools that sends heartbeats representing an uprunning node.

Currently, for Stellar-Core, I set PUBLIC_HTTP_PORT=true, and my firewall limits that port from external access.

Other than tuning firewall policies, is there any other way / tools for external parties to check my node healthiness?

3

You can get the latest ledger header from horizon, for instance the following command would work from the command line:

curl 'https://horizon.stellar.org/ledgers?limit=1&order=desc'

Obviously substitute the node you actually care about for horizon.stellar.org.

The sequence and closed_at fields give you the ledger number and close time, which will give you an indication of how recently the node reached consensus. If you are worried about actual forks from misconfigured quorum slices, then you can look at the hash and prev_hash fields and make sure those match the same ledger number at other validators.

More documentation is available in the ledger endpoint description in the REST API documentation.

  • Thanks mate, good solution! but any way to check it more intensively? From this method, any malfunctioning cannot be detected in 5 seconds... It may matter in the way that my Core/Horizon is still trying to take in transactions, although it is down. – cesarm Jul 30 '18 at 2:25
  • 1
    I mean technically your node has not malfunctioned until it fails to process ledger close at the next 5-second interval. I guess you could try to connect to stellar-core over TCP to make sure the unix process hasn't died, but that's not secure. – user3188445 Jul 30 '18 at 22:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.