Build 132

05 December 2017

Features

Flexible Alerting

It’s now possible to configure multiple alerting rules with an unlimited number of integrations to be alerted through. To filter for the specific events to be notified on, enter a Dynamic Focus query by clicking on “Advanced Filter”. For example, to notify on events affecting the prod zone of the monitored system, specify the query entity.zone:prod. The configuration has moved from “Management Portal → Integrations” to “Settings → Alerting”.

alerting configurations dialog

AWS Support

To provide deep visibility and insight into Amazon Web Services-based cloud environments, Instana now automatically discovers and monitors AWS services. Read more about our AWS support.

Improvements

  • New Node.js version 8 support. The Node.js sensor now will instrument apps running on Node.js versions 8.2.1 and higher. See the Node.js documentation for specifics.
  • The trace and event view will now present only the list of item in a fullscreen mode. When selecting an item, users will find the view as it was before, split into two columns: The item list on the left and details about the selected one on the right.
  • Interacting with the comparison table or the event view will no longer manipulate the dynamic focus input field.
  • Infinite scrolling for trace and event list is now replaced by an explicit interaction button.
  • Tables are now unified in their look and behaviour.
  • Changes to the entity type, selected entities and metrics in the comparison table are now encoded to the URL
  • Improved support for Azure Instance metadata
  • JVM crashes are now detected and reported automatically
  • Java Tracing now supports Amazon DynamoDB
  • Instana Java SDK will automatically mark the first INTERMEDIATE span as ENTRY
  • The Instana agent will now consume less peak memory and thus support tracing higher call volumes. It is recommended to “re-install” agents from a fresh download which are older than 6 months to take advantage of this.
  • Improved parsing of MySQL configuration for auto-discovery
  • Improved support for Docker, Kubernetes and Weave Networking which does not appear in docker inspect
  • Initial support for running the agent in Amazon ECS and EKS
  • Reduced overhead introduced by .NET-Tracing (memory-consumption and CPU-cycles)
  • Reduced cpu usage of InstanaPCP
  • InstanaPCP now checks for prerequisites being installed upon starting
  • Added EUM-support for .NET applications (ASP.NET WebForms, WebAPI and MVC)
  • Added instrumentaton for HttpWebRequest
  • Services created based on ASP.NET Spans will not be case-sensitive anymore (thus no duplicate services)
  • Marathon application ID (appId) and labels can be used for service definition and naming in the Service Extraction Rules

Fixes

  • OR connected terms in dynamic focus will now lead to correct results in the agents overview.
  • Node.js Sensor: Node.js <=4.5.0 can have Buffer.from, but it does not accept a string. Thanks @libozh!
  • Node.js Sensor: Support announce to agent even when the Node.js process is renamed.
  • Python Sensor: Don’t report submodules in snapshot.
  • Python Sensor: Lower log levels for non-critical issues
  • JBoss Data Grid Sensor: Improved hit rate calculation
  • NGinx Sensor: Fixes parsing problems with Naxis Rules
  • Correctly detect Docker socket when the “-H” argument is empty (like in OpenShift)
  • .NET Tracing: Correlation of traces between WCF-client and -service on basicHttpBinding
  • .NET Tracing: Stack-walking for .NET Classic 4.7.1
  • .NET Tracing: Duplicate addition of headers on nested Http-calls
  • .NET Tracing: Fixed unknown MsSQL server
  • Fixed a problem with reporting metrics for CLR-processes when Performance-Counters are not installed