E-commerce application use case

Consult the following use case for information on how to achieve value with TrueSight Intelligence.

The goal

The e-commerce application manager should be able to know if a problem with the systems affects the key business goals (metrics) and vice versa.

Key business goals examples: low checkout time, low turnaround time for catalog browsing, secure payments, seasonal traffic handling, and so on.
System examples: database server (catalogs, user accounts, etc.), network infrastructure, system memory, and so on.

The workflow

Set up data collection

Use any of the available methods to collect data for relevant business, application, and infrastructure metrics: Amount spent per order, checkout time per order, CPU usage, Memory usage, and Network out.

For example, if you are using a combination of PostgreSQL, Linux , AWS EC2 instances, and NGINX, you can collectively view data from all the sources that are needed to manage the health of your e-commerce application.

For more information, see Collecting-data.

Create and monitor the application

Any metric data brought into TrueSight Intelligence can be made into a key performance indicator (KPI). It is easy to determine and set these KPIs.

In this example the e-commerce application manager has determined that these set of metrics should be set as KPIs: Number of customers logged in (business metric), Amount spent per order (business metric), Checkout time per order (business metric), CPU usage (infrastructure metric), Simultaneous online users (business metric), and Network out (infrastructure metric).

For more information, see Searching-for-data, Setting-metrics-as-KPIs-for-analysis, and Managing-applications.

Add capacity to match request volumes

Checkout time per order (business metric) increases when more than 1000 customers are shopping simultaneously. Multiple abnormality events are generated due to deviation from the baseline.
Upon further investigation with IT and the capacity team, it is determined that additional capacity is required for handling the increase in number of online customers.
For example, if you are using a cloud service, you can decide on the minimum number reserved instances you need though the year.
By looking at last years data it is determined that the number of online customers during certain seasonal periods goes up to 5000 customers, which will require additional capacity.
For example, if you are using a cloud service, consider purchasing additional instances just before the predicted peaks (black Friday, school reopening, and so on) to ensure optimal performance.

For more information, see Viewing-abnormality-and-MVGD-events and Viewing-correlated-metrics.

The benefits

A low turnaround time was maintained.
Continuous monitoring helped perform a timely increase of system capacity.
The increase in capacity helped cater to the increased request volume.