Closed stephen-soltesz closed 1 month ago
NOTE: we discussed plans in the eng call for monitoring to be from a new prometheus instance in a GKE cluster dedicated to the autojoin platform and data pipeline - for alpha / beta testing this configuration will suffice. TODO to remove byos configs in this PR.
This change adds basic functionality for collecting prometheus metrics from BYOS nodes. The gcp-service-discovery now includes a reference to the Autojoin API List operation, which returns known BYOS nodes and associated ports for monitoring.
See: https://github.com/m-lab/autojoin/pull/32
This change is