Skip to main content

Verifies that Kinetica is running at the cluster in use, failing over to another where it cannot be reached.

The head node in use is sent a GET, waiting as long as the request timeout, and must answer exactly “Kinetica is running!”. Any other answer fails with “Server returned invalid response: <answer>”, without failing over. Where the head node cannot be reached, or answers an error status with nothing to read, the connection fails over and the next cluster is tried. Once every cluster has been tried, the failure is a connection error with the last attempt’s message.