Rich real time execution analytics and profiling combined with source-level root cause diagnosis

The ReplayLIGHTNING rapid execution profiler is the newest member of the ReplayDIRECTOR family of application recording and replay products. ReplayLIGHTNING is a fully integrated, interactive analytics front end to the application record/replay functionality of ReplayDIRECTOR. The combination is designed for the agile development and DevOps environment where time is at a premium and collaboaration is essential for finding, communicating and fixing code defects, performance and security issues.

Real Time Dashboard, Event Analytics and Performance Profiling

ReplayLIGHTNING streams highly actionable application execution data directly into the Eclipse IDE.

 

ReplayLIGHTNING Eclipse IDE Integration
Click Image to Enlarge

With ReplayLIGHTNING activated, development, test, operations and support have instant access to comprehensive diagnostic information WITHOUT PRE-DEFINING THE TIMEFRAME, EVENT TYPES OR BEHAVIORS THEY ARE LOOKING FOR:

• A time-based, dynamically configurable “radar screen” that summarizes any critical execution event type such as CPU and memory utilization, log generation, exceptions, transactions, database calls, etc. The axes of the real time display are selectable and can display the time based behavior of any event type that ReplayLIGHTNING collects.
  
  
• An integrated list of all execution events within a specific time window
  
  
• A windowing function that can be used to narrow down the timeframe of interest in examining execution events and activity. With the click of a mouse the few seconds of keys events describing a specific defect occurrence can be selected from a multi-hour execution session.
  
  
• Within these windows, a high-precision event view can be generated by drilling-down into an amplified (expanded) set of events. A general-purpose filtering function will return specific event types (such as database calls, exceptions, etc.) to reduce the noise in defect analysis.
  
  
• Profiles of the executed code coverage along with a metric highlighting the complexity of the executed code. This will provide feedback on where additional test cases should be generated.
  
  
• Visual indication of hot spots and slow-downs in the code via the Performance Profiler.

Using Real Time Event Data to Navigate Execution Recordings

Replay LIGHTNING is also the “window” into the source-level diagnostics provided by the ReplayDIRECTOR application record and replay functionality. The ReplayLIGHTNING data can be used to generate “diagnostic probes” (much like data mining probes) to identify and request additional execution information within a specific time window from the ReplayDIRECTOR recording.

Examples of the deep diagnostic actions that ReplayLIGHTNING can initiate are:

Turn Back the Clock-- Event Amplification. Many applications at best produce minimal event output such as logs because of the overhead of generating and capturing truly debug-level data. With the ReplayLIGHTNING execution amplifier working in conjunction with a Replay recording, debug-level logs, exceptions, database calls, etc. can be captured on recording playback as if they had been originally generated. This “turn back the clock” functionality is unavailable in any other application diagnostic or monitoring solution.

Source-level Defect Marking and Debugging. ReplayLIGHTNING event “cues” such as exceptions and log messages can be used to put bookmarks in Replay recordings so that a developer can directly jump to the exact spot in the source code where the problem occurred. The markers can be dropped by a tester during a manual test when a issue is detected on the screen, or programmatically during automated testing when a specific type of exception or other error condition occurs.

Source-level Defect Marking and Debugging
Click Image to Enlarge

Memory Leak Locater. Unlike traditional memory leak detectors that can indicate where a leak impacted a program and at best what class it occurred in, ReplayLIGHTNING can bring a developer to the exact point in source code where the object was allocated and provide the context (db call, transaction, etc.) in which the object was used. Through the use of the replay functionality to recreate the exact program execution, all the guess work in memory leak detection and resolution is eliminated.