mirror of
https://github.com/neondatabase/neon.git
synced 2026-01-07 13:32:57 +00:00
13feda0669
## Problem We don't know how much time PS is losing during ingest when waiting for remote storage uploads in the flush frozen layer loop. Also we don't know how many remote storage requests get an permit without waiting (not throttled by remote_storage concurrency_limit). ## Summary of changes - Add a metric that accumulates the time waited per shard/PS - in [remote storage semaphore wait seconds](https://neonprod.grafana.net/d/febd9732-9bcf-4992-a821-49b1f6b02724/remote-storage?orgId=1&var-datasource=HUNg6jvVk&var-instance=pageserver-26.us-east-2.aws.neon.build&var-instance=pageserver-27.us-east-2.aws.neon.build&var-instance=pageserver-28.us-east-2.aws.neon.build&var-instance=pageserver-29.us-east-2.aws.neon.build&var-instance=pageserver-30.us-east-2.aws.neon.build&var-instance=pageserver-31.us-east-2.aws.neon.build&var-instance=pageserver-36.us-east-2.aws.neon.build&var-instance=pageserver-37.us-east-2.aws.neon.build&var-instance=pageserver-38.us-east-2.aws.neon.build&var-instance=pageserver-39.us-east-2.aws.neon.build&var-instance=pageserver-40.us-east-2.aws.neon.build&var-instance=pageserver-41.us-east-2.aws.neon.build&var-request_type=put_object&from=1731961336340&to=1731964762933&viewPanel=3) add a first bucket with 100 microseconds to count requests that do not need to wait on semaphore Update: created a new version that uses a Gauge (one increasing value per PS/shard) instead of histogram as suggested by review