Skip to content
This repository has been archived by the owner on Jul 4, 2018. It is now read-only.

Latest commit

 

History

History
39 lines (28 loc) · 1.62 KB

changes.md

File metadata and controls

39 lines (28 loc) · 1.62 KB

0.9.11.1

  • Fix tracking of pending-size which could become negative under some retry conditions.

0.9.0

  • Added NewRelic metrics sender contributed by LockedOn (http://www.lockedon.com).
  • Deprecated :metrics/workflow-name. Use job metadata via your job submission like so:
  • Bumped recommended dogstatsd-clj version as previous version has a memory leak
{:workflow ...
 :lifecycles ...
 :metadata {:name "YOURJOBNAME"}}

0.8.8.1

  • Added datadog metrics sender, contributed by Cognician!

0.8.2.6

  • Fix bug where throughput would only be recorded for input tasks.

0.8.2.5

  • Undo breaking change in 0.8.0.4. Include peer-id in tags, but not in service name.

0.8.2.4

  • Fix test error causing non-release of 0.8.2.4

0.8.2.3

  • Added measure for counting pending messages "pending_messages_count"
  • BREAKING CHANGE percentile measurements with a period now use an underscore e.g. 99.9 -> 99_9. This is so they are not segmented in grafana and others.
  • BREAKING CHANGE batch latency percentiles now include "th" in their name for consistency with the other services names.

0.8.0.4

  • BREAKING CHANGE Further segment metrics by peer-id, otherwise riemann and others will think measurements from different peers on the same task are just another reading

0.7.10

  • BREAKING CHANGE There is no need for separate latency and throughput lifecycles as these are now part of :onyx.lifecycle.metrics.riemann/calls, and :onyx.lifecycle.metrics.timbre/calls. Please refer to the README for the new lifecycle calls and configuration.
  • New metrics "complete-segment-latency" and "retry-segment-rate"