Significant Performance and Scalability Improvements |
Super Charged Probes Runtime and Extensions
With 100+ complete SPECjvm2008 benchmark runs executed with various configurations and optimizations, each run firing 2-3 billion probes, JXInsight 5.6 is our most performance focused release of our open and extensible resource metering runtime. JXInsight is 6-50 times faster than alternative solutions making it the only Java application performance management solution on the market that could provide fine grain event multi-meter metering (counts, max, mins, avgs) in a real-world enterprise application production environment.
Note: Comparing the raw overhead (exclusive of benchmark costs) JXInsight is nearly 20 to 100 times faster than its nearest rival.
Improved Hotspot Metering Strategy
JXInsight 5.6 introduces an improved and default metering strategy,
hotspot, that determines whether to continue to meter the firing of a particular named probe after a specified number of firings based on a threshold defined on a meter statistic. The strategy is useful in eliminating any overhead in measuring the execution of probes with consistently small metering consumption.
Additional Metering Strategies
JXInsight 5.6 also introduces additional metering strategies offering unparallel flexibility in fine tuning the collection of measurements. The
exclude strategy prevents the metering of particular probe names fired. The
include strategy restricts the metering of probes fired to a particular name list. The new
random strategy performs the metering of a randomly chosen firing of a named probe.
Billing Histories
JXInsight 5.6 now provides access to individual
Billing (metering) records posted to a specific probe
Group by a thread. For each active thread a limited history of billing is maintained and added to the snapshot model when a probe snapshot is requested by the application management and monitoring console or terminal.
AspectJ Hybrid Load-Time Weaving
JXInsight 5.6 extends it load-time weaving cache to now store the transformed bytecode image across application executions turning load-time weaving into binary weaving on consecutive executions. Our inspired approach makes the switch between load-time weaving and binary weaving transparent - the first time an application is executed JXInsight will use load-time weaving the second time it uses binary weaving via the persisted transformed bytecode from the initial load-time weaving transformation. This reduces the (re)start-up time as well as the memory footprint - both which is extremely important in production environments.
| New and Improved Grid Computing Extensions |
JXInsight 5.6 includes more than 50+ new and enhanced extension libraries for monitoring of popular commercial and open source grid computing platforms including
IBM eXtreme Scale,
Oracle Coherence,
JBoss Cache, and
GridGain. JXInsight offers a complete resource metering, call trace analysis, system metrics, and runtime state and request diagnostics solution that works consistently with multiple platforms and up and down technology stacks and across every phase in the application life cycle.
With JXInsight there is no need to learn a different monitoring & diagnostic tool and model for each technology or to pay multiple subscriptions for a basic colorized technology "themed" web dashboard - JXInsight delivers an uniform software performance engineering process combined with sophisticated software and system execution models and supported by an award winning Java application.
Production-Ready Adaptive Metering of the Grid
JXInsight 5.6 probe extensions for IBM's eXtreme Scale solution accurately determine the service times of various grid requests and at a very fine and configurable level of granularity across hosts, processes, and active threads.
Call Trace Analysis for the Grid
JXInsight 5.6 trace extensions for IBM's eXtreme Scale solution profile, trace and determine the call origin of requests to the IBM's eXtreme Scale grid.
Effective Problem Resolution
Problem diagnostics is greatly simplified with JXInsight's 5.6 extensions that allow operations staff to inspect on-demand the current in-flight requests along with object field state for arguments and target objects.
Diagnostic snapshots can be exported and emailed to support teams including third-party vendors for offline analysis within the JXInsight console which is available at no cost to all companies.
The JXInsight application monitoring and analysis console has been enhanced with new and improved table views related to resource metering.
Customization of Probe Icons
New in JXInsight 5.6 is the ability to customize the icon displayed alongside a probe group within the console providing a visual classification of particular probe groups sharing a common responsibility within an application that is not necessarily reflected in the package name space.
Cluster Wide Marking and Tracking of Metering
JXInsight 5.6 provides the ability to mark the current metering data across multiple hosts, process, and threads and then to track and compare changed in metering count, totals, maximums and minimums per group per meter.
Greater Flexibility in Analyzing Multi-Host-Process-Thread Metering Information
A new flat metering table view has been added that allows greater flexibility in filtering and sorting.
A number of bug fixes have been applied to the console, terminal, server and extension libraries.
