Service Monitoring with Business Space

Service monitoring measures the response time and request throughput for services invoked by and exposed by an SCA module. You choose what operations to monitor on the services exposed to requestors (SCA exports) and consumed (SCA imports), and can optionally define thresholds for response time and throughput.

Service monitoring is available for IBM® Process Server and Enterprise Service Bus from the Service Monitor widget in Business Space. Use service monitoring to gather and analyze response time and throughput metrics so you can answer questions like the following:

How much time do specific services require?
Does service duration degrade over time?
How often are specific services called?
Does throughput adhere to expectations you have defined, or does it degrade over time?
Do any of the calls exceed a defined threshold?

The information it provides helps you monitor ongoing problems and pinpoint which part of your solution is not responding as expected.

The service monitor plots response time and throughput data on graphs, visually distinguishing those calls that exceed any threshold you have defined. The graphs always show the latest monitoring statistics; however, you can see historical data by increasing the length of time shown on the graphs.

Response time

The Response Time graph indicates the time elapsed between a service request and response. (For service operations with two-way asynchronous implementations, the graph indicates only the time the operation needed to handle the request, not the time that elapsed between request and response.) Response time on the graph is plotted over seconds or minutes; in addition, the Statistic Measurements Table shows you response times for the last second or minute and for the entire monitoring session.

Throughput

The Throughput graph shows how many calls have been completed over a unit of time (seconds or minutes). In addition, the Statistic Measurements Table shows you throughput for the last second or minute and for the entire monitoring session.

Service monitoring architecture

The service monitor performs all service monitoring tasks. It has a client/server architecture:

Service monitor server: The service monitor server gathers and aggregates response time and throughput measurements from all running service monitor agents, and then calculates and stores the statistics. The Service Monitor widget queries the server for these measurements.
Service monitor agent: The agent measures the throughput and response time for operations and sends the measurement data to the service monitor server.

In a deployment environment, the server runs on a support cluster, while the agent runs in the application cluster on the server where you deployed your module. In a stand-alone server environment, the server and agent both run on the stand-alone server.

Service monitoring data is stored in memory. When the buffer is full, the oldest data is discarded and replaced by the newest data. The stored data for an operation is automatically removed when all users turn off monitoring for that operation.