Synthetic Monitoring: A Dress Rehearsal For AWS Success
January 14, 2021
By Sungard AS
Imagine having a dress rehearsal for your IT department in which you can preview incidents that could strain or trip up your business’ website before they happen. In a theatrical dress rehearsal, everything runs as a show would normally run – dialogue, acting, costumes, lighting, sound, scenery and props. The only thing missing is the audience; it’s as close to the final performance as you can get, while still leaving room for improvement without the risk of audience exposure.
Synthetic monitoring is the same idea, offering IT professionals valuable feedback before problems happen on a large scale so they can ascertain if a website or application is slow or experiencing downtime before impacting the customer experience. This type of monitoring simulates a typical website user’s behavior without requiring actual web traffic, which is why it is called “synthetic.” This allows companies to test applications in a realistic atmosphere with real-world conditions, but without real-world risk.
Such monitoring services for cloud-scale applications can help DevOps teams prevent downtime, solve performance issues, and make certain that customers will get the best user experience. Sungard AS runs synthetic monitoring for numerous AWS customers who use the service to ensure the best possible security, performance and proper resource allocation.
Synthetic Monitoring with Datadog
Unlike traditional infrastructure and application monitoring, synthetic monitoring specifically focuses on the end user experience, allowing a greater understanding of user actions and predictive behavior. It facilitates better visibility with regards to application uptime, global/geolocation performance issues, and identification of application issues. This helps track your application service level agreement (SLA) or service level objective (SLO) and validates key business transactions or workflows of your web application or application programming interface (API) endpoint. Synthetic monitoring helps reduce mean time to resolution by visualizing uptime data alongside key business metrics that you define.
Sungard AS partners with Datadog, a monitoring service for cloud-scale applications, to provide monitoring of servers, databases, tools, and services through a SaaS-based data analytics platform. This synthetic monitoring managed service supports two categories of tests: API tests, to test a customer's frontend or backend APIs; and browser tests, to create step-by-step recordings of key business flows within a customer's web application.
Tests are configured within the Datadog console, which allows Sungard AS DevOps engineers to create the tests easily on its customers’ behalf and work with them to define the test specifications for applications. At a high level, a test requires the following details:
URL or API endpoint request (including protocol)
Name of the test
Supported devices for the test (laptop, table or mobile)
Locations to run tests
Test assertions (API Tests)
Recording steps (Browser Tests)
API Tests use test scripts that generate traffic simulating HTTP, TCP or SSL requests to a company’s web application or API endpoint. The test scripts are configured to run periodically or continuously and include customisable conditions that dictate the outcome of a company’s tests; for example, HTTP response codes, request headers, and content body.
Browser Tests also use test scripts, however unlike API Tests, these record actions from a company’s web application which reflect a business process or a workflow (for example, logging into your web application). During the recording phase of the test, a customer’s interactions with your company’s web application are documented as series of discrete steps, which can be edited or enhanced with assertions and other validation actions.
What does a Synthetic Monitoring engagement look like?
A Synthetic Monitoring engagement is a collaboration between the Sungard AS Solutions Architect DevOps Engineer resources and a customer’s development, test and operations teams. This involves interactive planning and screen sharing sessions to define a customer’s requirements and implement their tests, assertions and, where applicable, alerts.
Note: In the case of Browser Test recordings, control is given to a customer’s team to perform the workflow within their web application, thus enabling Datadog to capture the input for their test.
How do I get access to the test results?
Sungard AS provides read only access to the Datadog console, which enables customers to view their test configurations and results. This view will also show events of tests that are in alarm status, complementing the alert notifications customers receive via email.
How do I know if Synthetic Monitoring is right for me?
Sungard Availability Services can help you find representatives or engineers who can discuss the Datadog Synthetic Monitoring process for your company. Click here for more information. By measuring uptime, availability and response time of critical pages and transactions, you will be able to measure your performance before the curtain goes up. The show must go on, but with a solid dress rehearsal, you’ll probably get better reviews on opening night.