Do you know Jennifer? (3)

Jennifer Dynamic Monitoring
In Application Performance Management, the concept of enterprise performance monitoring is classified into two part: Service Monitoing and Resource Monitoirng

Service Monitoring
When monitoring the status of the application services, it is import to retrieve the performance data in real-time and grasp the mutual relationship between each service transactions. A delay in transaction response time may have originated from not just that transaction but from a combination of all transaction’s resource bottleneck. JENNIFER provides following features in service monitoring.

• Concurrent User
• Active Service
• TPS, Arrival Service Rate
• Real-Time Transaction Response Time
• Application Naming Dynamic Substitution
• SQL Tracing (Including BIND Variable)
• Transaction Profiling(Class/Method Level)
• HTTP POST Request KEY/VALUE Tracking

Resource Monitoring
Resource monitoring includes managing the physical and logical resources used in running an application service. JENNIFER provides following features in resource monitoring.

• JDBC Connectio Pool Status
• JVM CPU Usage (KERNEL/USER/IO)
• JVM Heap Memory Usage
• FILE Read/Write Status
• TCP/IP Socket Input/Output Tracking
• Collection/Live Object Count
• Status of Java Application Server Resource Used by JMX

Performance Data Analysis
The performance problem in a web system can be categorized into two types: 1) conditional performance problems caused by anomalous circumstances or 2) relative performance problems
caused by an increase in load and throughput that is more than the amount system can handle. JENNIFER possesses various types of data collection and analysis modules for detecting all
types of performance problems.

• Memory Leak Detection
• Java Collection / Live Object, Application Tracing
• JDBC Connection / Statement / ResultSet Tracing
• JDBC Transaction (Commit/Rollback) Tracing
• Application Exception Tracing
• SQL, SQL Exception Tracing
• Dump on Running Service
• HTTP Session Dump Capability
• Reporting Loading Class Hierarchy Analysis
• Loaded Class Bytecode Deassembling

Real Time Active Service Monitoring
JENNIFER illustrates active service runtime in different color and displays it via the equalizer graph in real time.

JENNIFER transmits the performance data through reverse-direction protocol that connects from JENNIFER Server to JENNIFER Agent whenever user requests for the data. Using this method, JENNIFER can extract the snapshot information of the running active services. Active service data shown in JENNIFER is not a summarized data of events that occurred in the past but the
real-time data concerning active services.