Metrics and the access log

Metrics

TESSARIDB_S3_METRICS_LISTEN makes the node serve GET /metrics in Prometheus text:

• requests by operation, status and error code, and their duration by size class;
• requests shed under load;
• authentication refusals by reason, and policy denials;
• the heal and replication backlogs;
• the cluster's members, and those answering a probe.

Backlogs and members are read at the scrape; s3_collect_up says whether that read succeeded.

The access log

With TESSARIDB_S3_AUDIT=on (the default) every S3 request is recorded in the region's s3_access_log topic in TessariDB, kept 30 days, under its request id: operation, bucket, key, signing access key, status, error code, bytes, duration and client address. It never holds a secret or a signature.

A full queue drops the record and counts it in the metrics; the request is never held up. TESSARIDB_S3_AUDIT=off records nothing.

Read it with a topic consumer, as any TessariDB topic.