Black Friday Web Traffic Crashes HP, Best Buy and Other Ecommerce Sites

This was the day crashes were supposed to be literal rather than figurative.

But in a sign of the changing nature of consumer demand, it was websites, more than stores' glass doors, where the breakage occurred.

A number of prominent retailer and etailer sites crashed, including HP and Best Buy. Others experienced delays usually expected more at snowy airports than at ecommerce platforms.

The heavy traffic appears to be the result of  converging trends: less impulse buying and more specific 'search and deploy' tactics from consumers on the hunt for certain items; heavier-than-expected volume prior to what was expected to be the Cynber Monday rush; and a sense that, despite lower gas prices, shopping may be losing its exalted status as a favored pastime for the entire family.

This simply confirms what analysts have been predicting based on accumulated trend data. But it also confirms that simply having the data is not sufficient if organizations are unwilling or unable to act on it. JL

Sean Gallagher reports in ars technica:

While content delivery networks (CDNs) have made it possible to push static content out closer to Web and mobile shoppers and reduce overall traffic hitting e-commerce sites, the load on the sites is still causing some to buckle and break
They don’t call it Black Friday for nothing. And as shoppers increasingly use websites and mobile apps to do their shopping—even as they brave the crowds in brick-and-mortar stores—some retailers are finding it hard to handle demands.
While content delivery networks (CDNs) have made it possible to push static content out closer to Web and mobile shoppers and reduce overall traffic hitting e-commerce sites, the load on the sites is still causing some to buckle and break, albeit briefly.
Even Hewlett-Packard has gotten a bit of a bloody nose as HP Shopping went down around midday Eastern Time. And Best Buy’s website went offline briefly—though by noon those who had left their browsers pointed at the website were switched back to the electronics store. The trials and tribulations of major retailers’ websites are being live-blogged by David Jones of Dynatrace, an application performance management subsidiary of Compuware.

Cloud bursts

The hiccups in HP’s retail site are especially stinging because though the company is in the midst of corporate mitosis, the consumer-facing website’s outage was apparently because of HP’s internal enterprise cloud systems—an embarrassment for a company trying to position itself as a major cloud provider. The outage was not apparently the result of problems with its content delivery network provider, Akamai, as the site kept serving pages—albeit not the kind visitors wanted to see. Instead they saw an error page that indicated the back-end applications for the site couldn’t keep up with traffic or failed under load—not exactly good news considering HP’s recent announcement of another decline in revenue.
Best Buy also appeared to suffer from an internal error. The electronics retailers’ site had been performing at peak efficiency for most of Wednesday night and this morning, until just before 10am Eastern Time—when the site suddenly dropped offline.
As Ars attempted to reach a Best Buy spokesperson about the outage at 11:27am, the site refreshed itself through the magic of JavaScript, and everything appeared to return to normal. “It wasn’t an external issue —it was not an issue with a third-party provider or a peering issue,” said Jones in a phone conversation with Ars about Best Buy’s outage. “It basically looked like an internal error.”

Unsportsmanlike conduct

Best Buy and HP have had plenty of company in the site outage department. The sporting goods retailer Cabela’s has suffered a series of outages, apparently due to internal factors.
Cabela’s site was showing signs of instability on Thanksgiving night. But by this morning, the site was in full retreat, and visitors to the Web store were greeted with a message saying that the site “is temporarily down for updates.” Customers were encouraged to place orders over the phone and were told that the issue was “routine system maintenance. Rest assured our team at Cabela’s is working diligently to make the site available soon.”
Part of the problem at Cabela’s may simply be how complex the retailer’s commerce site infrastructure is, Jones noted in his blog—the site pulls content from servers all over the US to serve up its e-commerce pages and relies on “a lot of Third Party and CDN services.”
Shoe retailer Foot Locker also had issues early today—apparently due to problems with their content delivery network provider’s configuration for their site. “Foot Locker tweeted out to people to refresh their page and try again,” Jones told Ars. “When we looked at the details, we can see that they’re using a CDN. And so typically when they say to a customer that they want you to refresh a page, the issue was a configuration problem with a CDN.”
Foot Locker’s site is front-ended in most of the US by Akamai, though other CDN providers may be involved in serving up elements of the site.