Metrics in the monitoring tool

The Attributes fields in the monitoring tool provide details that can help you to determine the health of the Opportunity Detect runs being executed on your Streams servers.

The monitoring tool can show metrics at the server group, server, deployment, and component level. The following table describes the fields shown in the Attributes node.

The polling interval and retention period mentioned in the following table are configurable on the Settings > Configuration page under HCL Opportunity Detect and Interact Advanced Patterns | System | Monitor.

Table 1. Monitoring metrics
Field Description
Wider Interval An interval of time you can set to see the data over a longer period than the polling interval.

Persisted data from the database is used for this calculation.

The average TPS informs you of throughput at the last polling; but it gives you no perspective on whether that number is normal or slow. The wider interval gives you a broader perspective.

Every time an average TPS is calculated it is written to the database, so the tool has access to historical average TPS values. When you specify a wider interval, the tool averages all the historical TPS values for that time period. In this way, you have two numbers to compare: the current TPS against the historical average.

You should set the wider interval value to less than the retention period.
Average tps The average number of transactions per second processed during a polling interval.

Persisted data from the database is used for this calculation.

Health Status Health is aggregated over all health statuses of all components.

Failure of an engine causes the server to become unhealthy, which in turn causes a deployment to become unhealthy. A deployment involves other jobs that handle input and output and potentially several engines. When polling is done, the tool looks at all these jobs and determines whether they are healthy.

If all are healthy, this value is healthy; if any one is unhealthy, this value is unhealthy.

The possible values are:

  • HEALTHY
  • PARTIALLY HEALTHY,
  • PARIALLY UNHEALTHY
  • UNHEALTHY
Transactions per second (tps) The number of transactions processed per second over the polling interval by a deployment, server, or engine.
Total transactions The total number of transactions processed over the polling interval.
High Engine tps/Low engine tps Transactions per second over the polling interval on the highest performing engine among all engines on the server. Like health status, high and low engine TPS are aggregate values across engine transaction processing rates belonging to that server or deployment.

High engine TPS represents the highest TPS value for any engine recorded in the last poll. Low engine TPS represents the lowest value.

These values give you insight into the overall behavior of the system. For example, the average TPS in the context of the wider interval TPS might be okay; but some engine may be slow performing. By contrasting the high and low TPS, you the gain insight that there is a slow performing engine.