TABLE OF CONTENTS
Minimum version to update directly: Build 197
In release #203 we will release a bigger migration that will be executed in the background as we are retiring ElasticSearch 2 and migrating to ElasticSearch 7. We anticipate no deviations from the standard update procedure.
Added instrumentation for AWS Kinesis
AutoProfile support for Ruby is available.
The name for infrastructure entities for runtimes (Golang, JVM, Haskell, .NET, Node.js, PHP, Python, Ruby) can now be customized with the environment variable
- Unbounded Analytics: Improved ungrouped view for calls and traces by supporting the same chart metric and aggregation selection as in grouped view.
- Java Trace Sensor: Add tracing support for Java 16 and 17, add support for Java HTTP Client.
- IBM Cloud Platform Services Sensor: Add support for IBM Cloud Databases for Redis and IBM Cloud Databases for PostgreSQL.
- AWS icons update. Both application and documentation are using the latest AWS Architecture icons now.
- Docker: Normalized CPU usage values in the range [0, 100]% are displayed for total, kernel, and user metrics.
- Go: versions
v1.28.1were released. Support for custom tags was added.
- Process Sensor: Improvements of snapshot handling
- IBM Cloud Platform Services Sensor: Added support for Cloud Foundry and Event Streams.
- Top List widget can now display up to 10 groups
- Application charts (pie and time series) can display up to 20 groups
- IBM MQ: Correct dependency of topic and listener.
- Kubernetes: Fix some situations where cluster metrics were not correctly reported or were reported as zero.
- Application events: The link to Unbound Analytics applies the correct boundary scope of the application name filter used.
- ClickHouse disk total usage is now correctly calculated taking into account all tables
- IBM MQ Channel metric gathering improvements
- PostgreSQL Sensor: process discovery improvements
- Migrate audit logs from ElasticSearch to Cockroach
- Migrations now use an improved logic for start/shutdown of datastores