According to this post the dispatcher now defaults to prevent duplicate queries. This can be turned off with the option "--allow-duplicate-queries=yes", but is not yet configurable through ENV vars.
This PR provides this option.
Second, the default "healthcheck" script fails because it always does the same (thus duplicate) query to node 1.
This PR queries a random node in the healthcheck script if duplicate queries are not allowed.
According to this post the dispatcher now defaults to prevent duplicate queries. This can be turned off with the option "--allow-duplicate-queries=yes", but is not yet configurable through ENV vars.
This PR provides this option.
Second, the default "healthcheck" script fails because it always does the same (thus duplicate) query to node 1. This PR queries a random node in the healthcheck script if duplicate queries are not allowed.