Saving Cyber Monday: A Retailer Averts Disaster with ExtraHop

While Black Friday diehards continue to fight over television sets and parking spots, droves of other shoppers prefer the comfort and safety of their couch. Indeed, last year was the biggest year ever for Cyber Monday, with sales growing 16.8 percent over 2016.

This massive influx of website traffic is a significant boon for retailers, and also rewards those retailers who've invested in efficient, productive IT departments. Considering Cyber Monday is now by far the biggest shopping day of the year, the ability to quickly identify and fix website glitches makes all the difference when it comes to holiday retail supremacy.

Performance Monitoring Tools Behaving Badly

Early on the morning of one Cyber Monday past, a major home goods retailer and ExtraHop customer started seeing slowdowns on its website. The time required to process web requests was averaging 20-30 seconds. For web pages with multiple requests, this resulted in loading delays of several minutes. Thanks to wire data from ExtraHop, however, the retailer spotted this issue in real time and ExtraHop support engineers were able to immediately dive into the problem.

Working with the retailer's IT team, the ExtraHop engineers spotted unusual activity on the SQL Server database supporting the website. A third-party SQL Server monitoring tool was pushing massive volumes of traffic to the production database. Overwhelmed, the database server was throttling the transaction rate for all requests. Ironically, the monitoring tool was originally put in place to warn against such events. This incident highlights the danger of relying on invasive agent-based application performance monitoring (APM) tools, which can actually do more harm than good.

The retailer's IT department quickly shut down the misbehaving SQL Server monitoring tool. Almost immediately, the database transaction processing time improved seven times over, dropping from 24 seconds per transaction on average to 350 milliseconds. The website that had previously taken minutes to load now did so almost instantly.

With real-time analytics based on wire data, ExtraHop enabled the retailer's IT team to see a broad picture of all the infrastructure supporting the retailer's website and to quickly identify the root cause of the slowdown. This end-to-end visibility helped troubleshoot a problem before it tanked Cyber Monday sales.

IT Operations Can Save Millions of Dollars for Retail

As e-commerce continues to grow, the ability to ensure the performance, availability, and security of websites is what will separate the real contenders from the soon-to-be defunct during the holiday season. Retailers who use real-time IT analytics to secure and maintain their holiday sales are going to win this season.

That's not hyperbole, and here's why:

1. Unprecedented visibility into network traffic (even traffic encrypted in accordance with TLS 1.3) means IT Ops can see every unusual spike, with full context of what it's affecting and why, in real time
2. Definitive insights based on machine learning trained on over 4,600 wire data metrics, which include every transaction from your brick and mortar stores to the cloud, for the highest fidelity analytics
3. Immediate answers about at-risk systems, vulnerabilities, end-user experience issues, and more, plus automated investigation workflows to help IT Ops remediate the problem before customers even notice

ExtraHop customers report a 75% reduction in lost user time due to app degradation, as well as an 86% reduction in unplanned downtime. Over Black Friday weekend, those percentages can translate to millions of dollars saved.