This release adds additional prometheus metrics to the S3 backend for better understanding of errors and latencies.
What's Changed
- New metrics added:
storage_s3_call_error_by_type_total- Which adds an error type label; this is intended to replace the existing storage_s3_call_error_total metric over time, which only counts errors but not what type of error. Alsostorage_s3_call_duration_seconds- Tracks the duration of each s3 method in a histogram. In #110
Full changelog in: v0.3.1...v0.3.2