DOT METRICS
A robust application metrics tool offering a broad range of benefits both to developers and the run-the-bank operations team
FOR DEVELOPMENT...
DOT Metrics combines the benefits of a robust metrics tool, allowing developers to report continuously on the availability and performance of their applications, with an easy-to-use means of publishing metrics to a range of system monitoring platforms.
Compared to common home grown instrumentation solutions, DOT Metrics offers a simple API to capture different types of metrics, which can then be published to multiple channels, including log files,Windows performance counters or JMX.
DOT Metrics also includes a useful viewing and alerting system to use during debug and testing.
AND FOR OPERATIONS
However the biggest benefits come when applications built with DOT Metrics are released into production. Its daily use by developers results in a system with a more clearly-focused set of metrics, any subset of which can be published to the monitoring system such ITRS's Geneos or HP's OpenView simply by changing a configuration file.
This richer set of metrics strengthens Operations’ ability to spot problems before they occur - queues lengthening or latency increasing - and to diagnose production incidents promptly.
For smaller organisations without a Systems Monitoring Platform in place, DOT Metrics provides a simple monitoring and alerting system to improve support of production systems.
INSIDE DOT METRICS
-
Simple programming model: basic system metrics such as cpu and memory usage require only an initialisation call
-
Additional metrics easily add through an intuitive API.
-
Metrics can be published to a range of channels including TIB RV, TCP, Windows performance counters and system monitoring platforms.
-
DOT Metrics viewer runs on desktop providing continuous view of system health
-
Shows “What’s running where? “
-
Visible alerts on loss of heartbeat
-
Data can be logged to a csv file for easy analysis in Excel
The DOT Metrics system supports Java and Streambase with a set of .jar files and .Net with a set of .Net assemblies. In both cases, simply linking and initialising the DOT Metrics system is enough to generate a heartbeat and a basic set of metrics, including processor and memory usage. A simple configuration file controls where and how often the data is published.
The DOT Metrics viewer listens for the messages and organises the data into a variety of displays including a tabular display and a network representation.
Developers can add application-specific information such as status, queue lengths, updates-per-second, server addresses and other key configuration data or whatever else is germane to the purpose of the application. These quantities can be grouped and built into a hierarchy, if required. This structure is defined programmatically in terms of simple AddChild, AddItem methods and can be as simple or as complex as is necessary for the application. Developers can also set up an overall status which is reflected in the metrics viewers and any other alerting system.
The DOT Metrics system automatically renders this structure into a schema which is broadcast along with the items themselves. There's no need for any complex external definition or configuration to structure the broadcast data.
The viewer can make use of the schema to drill down into the data structure and display as much or as little detail as required. Processes that have stopped sending heartbeats or have raised a “non-Green” status will be highlighted.
