MinerConnectionError when ECC is bad

[posterzh]

Follow up to:

• MinerConnectionError: Unable to connect to miner http://helium-miner:4467 #350

Thoughts:

• For any miner, if helium-miner container is down, then one of the following actions makes MinerConnectionError.
  • open the diagnostics page.
  • call diagnostics/json endpoint
  • upload the diagnostics to the Google big query

The bad ECC can be identified as follows:

• Balena device: https://dashboard.balena-cloud.com/devices/c4cb830c21b581fb483d197b571158f9
• SSH into diagnostics container
• Run /opt/python-dependencies/hm_pyhelper/gateway_mfr --device=ecc://i2c-7:96?slot=0 test
  Result of test is fail.

{
  "result": "fail",
  "tests": {
    "ecdh(0)": {
      "error": "elliptic_curve error",
      "result": "fail"
    },
    "key_config(0)": {
      "checks": {
        "auth_key": "0",
        "intrusion_disable": "false",
        "key_type": "ecc",
        "lockable": "true",
        "private": "true",
        "pub_info": "true",
        "req_auth": "false",
        "req_random": "false",
        "x509_index": "0"
      },
      "result": "pass"
    },
    "miner_key(0)": {
      "checks": "112GoL4dxhdTkAZ1Ny7Kqe1mTgH6scA3Sh3fMv77uY87JvH5YdwV",
      "result": "pass"
    },
    "sign(0)": {
      "checks": "ok",
      "result": "pass"
    },
    "slot_config(0)": {
      "checks": {
        "ecdh_operation": "true",
        "encrypt_read": "false",
        "external_signatures": "true",
        "internal_signatures": "true",
        "limited_use": "false",
        "secret": "true"
      },
      "result": "pass"
    },
    "zone_locked(config)": {
      "checks": "ok",
      "result": "pass"
    },
    "zone_locked(data)": {
      "checks": {
        "expected": "locked",
        "found": "unlocked"
      },
      "result": "fail"
    }
  }
}

• helium-miner container is still accessible through SSH, but all miner operation is timed out.
  For example, if run miner peer book -s command: