Build 130

  1. October 2017

Features

  • Python: New TextMap OT propagator (thanks @elgris!) #16

Newly Supported Technologies

  • Mule ESB: Initial discovery and dashboards
  • Java 9 support

Improvements

  • Go: Delete pre-existing HTTP Header keys to avoid append (thanks @vetinari for pointing this out!) #40
  • Go: Better locking to avoid data races #41
  • Go: Only report custom service name from tracer (if configured) #43
  • Python: Handle errors and make best effort to report what we have for snapshot collection (thanks @elgris!) #24
  • Python: Fix exception logging: convert to string (thanks @elgris) #22
  • Python: Add handle_fork function for eventual pid change detection #21
  • Python: Implement advanced announce for containers #20
  • Python: Improved detection of the default gateway (thanks @elgris!) #18
  • Python: Logging: Make sure the arg we are passing is a string #19
  • Python: Better process/entity naming #27
  • Python: Custom service name support #25
  • Ruby Performance: Don’t create a new Integer Range on each ID generation #92
  • Ruby: Loosen up the dependency on timers to >= 4.0 #95
  • Ruby: Require entire CGI due to Ruby 2.4.1 bug #13539 #94
  • Ruby Performance: Use clock_gettime instead of converting Time object #93
  • Ruby Performance: Disable backtrace collection by default; add config option #98
  • Ruby: Report spans in batches #96
  • Ruby: Improved debug checks and logging #97
  • Ruby Performance: Switch to Oj for faster json processing. #101
  • Java: Reduced memory usage for tracing
  • Java: Trace remote vertx communication
  • Java: automatic attachment improvements for early Java 6 builds
  • JBoss: AS 7.0 compatibility, fixed incorrect servlet invocation counts
  • PHP: Performance Improvements
  • Kafka: Corrected partition count
  • MongoDB: Support strings in bindIP config
  • Etcd: added more client url discoveries
  • RabbitMQ: Corrected reported port list. Performance improvements.
  • Docker: fixed missing network metrics under some configurations
  • .Net: Performance improvements
  • Varnish: Support Varnish v3 metrics
  • PostgreSQL: Fix communication with postgres when running on default interface

Agent Health Monitoring

Based on feedback from the last release we’ve introduced a chart to visualize change in reporting agents over time and improved the design of the overview and agent dashboard. Enjoy!

Improved Agent Health