0.8.0

Improvements

• Enabled the following instrumentation by default
  • Java Executor Service (#295)
    • Caveat: you still need to manually toggle it in your custom instrumentation.
      ((TraceScope) scope).setAsyncPropagation(true);
    • Some frameworks (servlets and jms message listeners) have this flag toggled automatically (#316)
• Dynamically rename services via configuration (#314)
  • Useful for making databases have distinct names across different services.
  • Configure mapping as follows:
    • System property: dd.service.mapping=mysql:my-service-db,cassandra:my-service-cassandra
    • Environment variable: DD_SERVICE_MAPPING=mysql:my-service-db,cassandra:my-service-cassandra
• Automatically set service name based on servlet context (war file name) if not already configured or configured, but empty. (#320)
• Improved instrumentation for AWS client: (#299)
  • This includes a change to the resource name.
• Automatically detect non-Datadog @Trace annotations (#311)
  • List of annotations can be replaced (or disabled with an empty list) using the following setting:
    • System property: dd.trace.annotations=com.some.Trace;io.other.Trace
    • Environment variable: DD_TRACE_ANNOTATIONS=com.some.Trace;io.other.Trace
• Configure additional methods to be traced without code change: (#311)
  • System property: dd.trace.methods=package.ClassName[method1,method2,...];AnonymousClass$1[call]
  • Environment variable: DD_TRACE_METHODS=package.ClassName[method1,method2];AnonymousClass$1[call]
• Automatically set a configured header value as a span tag (#317)
  • System property: dd.trace.header.tags=CASE-insensitive-Header:my-tag-name,User-ID:userId
  • Environment variable: DD_TRACE_HEADER_TAGS=CASE-insensitive-Header:my-tag-name,User-ID:userId
• Improve naming for @Trace generated spans on methods inside anonymous classes: now includes more class name. (#303)
• Upgraded to Byte Buddy 1.8.8 (#293)
• Remove guava dependency for dd-trace-ot (#307)

New Instrumentation (disabled by default)

• Hystrix (#306)
  • System property: dd.integration.hystrix.enabled=true
  • Environment variable: DD_INTEGRATION_HYSTRIX_ENABLED=true

Bugfixes

• Fixed Ratpack instrumentation -- still need to manually enable it (#301) (Thanks @jonmort!)
• Minor Kafka instrumentation change to reduce chance of breaking (#297)
• Prevent nested JDBC spans (#302)

0.7.0

Improvements

• Add startTime and duration to MutableSpan interface #283
• Improve trace count accuracy #289
• Priority Sampling config for Javaagent users #288

New Instrumentation (disabled by default)

• Play 2.4 - 2.6 #277 #282
  • System property: -Ddd.integration.java_concurrent.enabled=true -Ddd.integration.play.enabled=true
  • Environment variable: DD_INTEGRATION_JAVA_CONCURRENT_ENABLED=true DD_INTEGRATION_PLAY_ENABLED=true
• Support for Ratpack #276
  • Shoutout to @jonmort for the third-party contribution!
  • System property: -Ddd.integration.ratpack.enabled=true
  • Environment variable: DD_INTEGRATION_RATPACK_ENABLED=true
• Support for Spark Java #254
  • Shoutout to @JorgenG for the third-party contribution!
  • System property: -Ddd.integration.sparkjava.enabled=true -Ddd.integration.jetty.enabled=true
  • Environment variable: DD_INTEGRATION_SPARKJAVA_ENABLED=true DD_INTEGRATION_JETTY_ENABLED=true

Bugfixes

• Fix NoClassDefFoundError in jaxrs instrumentation #294
• Decrease log level for hot paths #291
• Correct units for start timestamp parameter name #286
• Instrument IBM-DB2 #281
• Fixed missing httpclient traces #296

For more details, check the full changeset and the release milestone.

0.6.0

Improvements

• Enabled the following instrumentation by default (#274)
  • JAX-RS annotation based resource naming
  • Jedis
  • Kafka client
• Upgraded to Byte Buddy 1.8.1 (#271)

New Instrumentation (disabled by default)

• JAX-RS Clients (#265)
  • System property: dd.integration.jax-rs.enabled=true
  • Environment variable: DD_INTEGRATION_JAX_RS_ENABLED=true
• Akka async actor propagation (#264)
  • System Property: dd.integration.java_concurrent.enabled=true
  • Environment variable: DD_INTEGRATION_JAVA_CONCURRENT_ENABLED=true

Bugfixes

• Fix instrumentation not loading for servlet 2 in some environments (#270)
• Instrumentation for Kafka Consumers weren't loading correctly in some environments (#275)
• Instrumenting old (pre Java 1.5) class files no longer fails (#266)
• Dectect Tibco temporary queue names properly (#268)

0.5.0

Improvements

• Configure tags to add to every span #245
• New TraceInterceptor API to modify and examine traces before sending to trace-agent #253

New Instrumentation (disabled by default)
(Will be default enabled in a future release. For this release, enable with the below config)

• scala async propagation for Promises and Futures #255
  • System Property: dd.integration.java_concurrent.enabled=true
  • Environment variable: DD_INTEGRATION_JAVA_CONCURRENT_ENABLED=true

Bugfixes

• Set max stack size correctly for error trapping #256
• Fix NPE edge-case in httpclient #258
• Mark all spans with an http status code of 5xx as an error #261
• Check up the classloader hierarchy to locate resources #259

For more details, check the full changeset and the release milestone.

0.4.1

Bugfixes

• [aws] Fix ClassNotFoundException when parsing aws headers (#244, #247)

For more details, check the full changeset and the release milestone.

0.4.0

Improvements

• Update javaagent bootstrap process to support enterprise application servers (#228)
  • WildFly/JBoss (#231)
  • OSGi (#232)

New Instrumentation (disabled by default)
(Will be default enabled in a future release. For this release, enable with the below config)

• Jedis redis client (#212)
  • Big shoutout to from @gihad from @uken for the first community contribution!
  • System Property: dd.integration.jedis.enabled=true
  • Environment variable: DD_INTEGRATION_JEDIS_ENABLED=true

Bugfixes

• Update DDApi connection with Datadog Agent v4 endpoint (#240)
• Use nanotime for additional clock resolution (#239)

For more details, check the full changeset and the release milestone.

0.3.3

Improvements

• Integrations can now be disabled individually (#213)
  • Set the system property dd.integrations.enabled=false or environment variable DD_INTEGRATIONS_ENABLED=false to disable all instrumentation by default.
  • Individual integrations can be enabled by name even if all are disabled by default:
    • System property: dd.integration.{integration-name}.enabled=true
    • Environment variable: DD_INTEGRATION_{INTEGRATION_NAME}_ENABLED=true
  • To disable a specific integration:
    • System property: dd.integration.{integration-name}.enabled=false
    • Environment variable: DD_INTEGRATION_{INTEGRATION_NAME}_ENABLED=false
• Spring Web exceptions that are handled via an error handler are now captured as errors on the span. (#227)
• Better URL based resource naming to reduce high cardinality issues (#221, #224)
• Improved AWS integration
  • Additional details captured for AWS Requests (#220)
  • Instrumentation now applies on the client constructor, not just the builder (#214)

New Instrumentation (disabled by default)
(Will be default enabled in a future release. For this release, enable with the below config.)

• JAX-RS annotations for resource naming (#230)
  • System property: dd.integration.jax-rs.enabled=true
  • Environment variable: DD_INTEGRATION_JAX_RS_ENABLED=true
• Kafka client and Kafka streams (#209)
  • System property: dd.integration.kafka.enabled=true
  • Environment variable: DD_INTEGRATION_KAFKA_ENABLED=true

Bugfixes

• Fix instrumentation for servlet 2 (#219)

For more details, see the release milestone.

0.3.2

Bugfixes

• Java Client no longer initializes the global java.util.logging.Logger: #210

0.3.0

General Availability
As of this release, Datadog Java APM is now generally available to the public.

Improvements

• Display trace service type in UI: #207
  • This feature requires the 5.21.1 trace-agent

Bugfixes

• Prevent breaking modification of signed aws client calls: #206
• Use thread safe map for jdbc connection storage: #203

0.3.0.RC2

Priority Sampling
Priority sampling is a new feature for sampling traces. #192

• The sampling rate is set per service name by the core agent.
• Additionally, a priority flag is propagated for external calls between java hosts.
  • This allows sampled traces to be saved across external calls.
• Requires the Datadog Agent 5.19 or greater.
• Further reading: https://github.com/DataDog/datadog-trace-agent/wiki/Distributed-Priority-Sampling

Bugfixes

• Log exceptions to datadog debug log instead of stderr #196
• Defend against NPEs in Spring Instrumentation #195

Internal Changes

• Test Java 7 and 9 in CI #201