August 6, 2018
This release introduces application perspectives for our SaaS customers! We’ve put together a dedicated transition page to provide guidance on successfully adopting the changes and new capabilities.
- Kubernetes 1.10 support
- Azure Kubernetes Service (AKS): 1.10.x
- Node.js Sensor: Support recording of user-configurable HTTP headers in HTTP client and server instrumentation.
- Node.js Sensor: PostgreSQL instrumentation (docs)
- Node.js Sensor:
httpsclient calls are now instrumented
- Node.js Sensor: express.js and fastify path templates, e.g.
/articles/:articleId, are now captured for better insights into the executed server side logic.
- Python Sensor version 1.0!
- Python MySQLdb Instrumentation (docs)
- Failed Dropwizard healthchecks are now displayed as an issue in Instana
- Metric capture support for Spring Boot 2
- Tracing of Async OkHttp3 requests
- Support Tracing Java -> Amazon SNS -> Amazon SQS -> JMS -> Spring call flows
- Widened JBoss AS Monitoring support to 4 - 13
- Extended custom metric collection to standalone usage of Dropwizard metrics library
- Display voluntary and non-voluntary context switches on process dashboard
- Added consul cluster view with list of live nodes and list of critical services
- Integrations now include container labels whenever applicable.
- Improved Slack integration which makes use of Slack’s formatting capabilities for message that are a lot easier to read.
- Node.js Sensor: Allow retrieval of package.json files.
- Python Sensor: Updates to support OpenTracing v2.0
- Python Sensor: Django & WSGI Improvements; More Safeties, Precision & Tests
- Python Sensor: Add setuptools check on install with clear message
- MongoDb Sensor: Supports connecting agent over TLS/SSL to MongoDb servers.
- GWT RPC Calls are now displayed as such
- Status Code Canges performed by ServletFilters are now captured correctly
- Spring Web Path templates are captured and used for endpoint mapping
- Better hostname detection on Windows Servers
- Improved tracing through Mule ESB
- Etcd sensor now more reliably detecs different etc configurations
- Httpd Sensor will now prefer non-ssl status urls to reduce CPU usage
- Improve Java 9 + 10 compatibility. Experimental Java 11 support.
- Improved compatibility to Docker Container User Namespaces
- Agent CPU usage reductions
- WebLogic Sensor: Collect datasources metrics
- Issue severity levels for issues reported for infrastructure entities were redefined. See full list of tuned issues here. Critical severity now means that the entity is not able to fulfill its purpose anymore or will not be able to fulfill its purpose in the nearest future. An example would be a split-brain state in the Elasticsearch cluster or a disk empty on a host. Warning severity means that the performance of an entity is negatively affected, e.g. increased response time or GC time below 80%.
- Node.js Sensor: Only try to instrument bluebird if it is actually available.
- Node.js Sensor: Avoid sending batch size 0 for Redis multi commands.
- Python Sensor: Format & clean up imports via isort
- Dropwizard Sensor: Failing health checks are now correctly discovered.
- RabbitMQ “queue filling up” health issue is now less sensitive.
Aug 23, 2018
Application Perspectives mode is not yet available on-prem.