Migration Guide: Migration guide for Lightbend Telemetry 2.12
New metrics for both Akka’s Persistent Actors and Lagom Persistent Entities. This is in addition to existing instrumentation covering the recovery of a persistent actor. It also includes Akka and Lagom persistence tracing.
Akka Streams Metrics Sampling. When running services containing high throughput Akka streams, every cycle matters for performance. While Telemetry aims to minimize its impact, some performance hit is inevitable, particularly for Akka streams stages that perform relatively little processing (e.g. a quick .map() operation). This can start to add up in a very high throughput application. To minimize this performance impact, Telemetry now offers the ability to only report a sampling of Akka Streams elements moving through the stream.
Akka Actor Mailbox & Stash Sampling. The performance and distribution of actor mailbox and stash metrics for aggregated actors (reported by actor class or group) has been improved with the introduction of sampling. This means that each actor’s mailbox and stash size will only be recorded every 1 second by default.
OpenTracing module for Datadog, enabling Lightbend Platform users to quickly and easily integrate with Datadog’s distributed tracing.
Released: 14 November 2019
- Add support for Lagom 1.6 on Scala 2.13
Released: 21 October 2019
- Fix prometheus akka persistence used/limit permits graph.
- Fix akka persistence stopwatch disabled exception.
- Fix elasticsearch akka persistence commands/events queries.
Released: 09 October 2019
- Add new Akka and Lagom Persistence metrics with the opt-in automatic dimension for the command and event types. See Persistence metrics for the details.
- Add support for Akka and Lagom persistence tracing.
- Stream telemetry improvements by adding sampling.
- Use sampling periods for mailbox size and stash size metrics.
- Add support for sending OpenTracing traces to Datadog.
request_methodas an additional label for endpoint metrics.
- Dashboards improvements including fix filters (first by application and then by server/host), improve JVM graphs, add graph for SBR events, add the http-port filter for the servers and endpoints graphs as well as the request method filter and some other fixes.
- Add per-endpoint MDC filter setting to HTTP configuration.
- Add Stream async boundary metrics.
- Fix support for proper MDC propagation for the Akka persistent actors.
- Add Graphite docker sandbox.
- Fix dispatcher metrics not reported after dispatcher restarted.
- Add new StatsD reporter opt-in TCPChannel that also works with Docker on MacOS as opposed to the default UDPChannel.
- Add support for Akka HTTP 10.1.10.
- Update to Jackson to 2.9.9.