The Worst Of Web: A Look Back At The Most Significant Web Outages Of 2012 And 2013

There’s a lot on the line

In all of history, no other generation has been as heavily reliant on technology as we are today. Our transactions, assets, and even directions to a destination are all tied in with technology. This means that even the slightest glitch could have effects that are far more reaching than we could have previously anticipated.

Evidence of past events shows that web outages could result in millions of dollars in lost revenue and millions more in repairing the company’s damaged reputation. So we have developed a list of the most significant web outages in 2012 and 2013 that will illustrate these effects clearly.

NASDAQ bug

When one of the largest stock exchanges in the globe grinds to a halt, you know that something significant is happening. On August 22nd2012, a software bug triggered a big halt in trading in the United States. The bug was only able to access the system due to an inadequate redundancy capability of NASDAQ’s system.

This incident, the first of its kind, lasted for about 3hours and affected over 3,200 companies. Thankfully, the halt did not cause a dip in trading, but certainly demonstrated the need to conclusively address vulnerability of exchanges to cyber attacks.

Amazon.com

This multi-billion dollar retail site went offline for about 40 minutes on the last day of January 2012. The outage, whose cause remains unknown, affected Amazon’s main website for US and Canada. As a result, it is estimated that the site lost nearly $5 million in sales. The other pages were restored, but the homepage took a while, a rare occurrence for the Amazon site.

The outage also affected other sites using Amazon’s cloud services. These included Vine, Instagram, Flipboard and Airbnb, among several others. Experts suggest that the popularity of these other sites may have contributed to all of them going down when the Amazon data center was affected.

Gmail

Gmail got a serious dent in its reputation and finances on September 23rd2012 when its users experienced prolonged delays in receiving emails. The 12 hour delay was said to be the result of a failure in a dual network which affected about 29% of Gmail users.

With delays in downloading attachments going as long as 2 hours, tempers flared among users who called on Gmail to get themselves together. Given that businesses are looking to Google to provide an email service that is cloud-based, such outages and delays affect the confidence of these businesses in Google and Gmail, which could result in loss of potential revenue.

Amazon Web Services

2013 saw Amazon Web Services down again, not once, but thrice - in June, October and December. The June outage was caused by severe storms which resulted in power outages in most of Virginia affecting the Amazon facilities in Northern Virginia. This affected other popular sites such as Netflix, Instagram and Pinterest, which use its cloud services.

October’s downtime was the result of a bug in one of AWS’s operational agent for data collection. This caused hardware failure in one of the servers used for data collection, necessitating replacement. During this replacement, the DNS update was not fully completed resulting in massive memory loss. This incident, like the previous one, affected major sites such as Imgur, Reditt and Heroku.

The December outage was the result of a deletion by a developer during routine maintenance that was not realized on time. As with the other occurrences, Netflix and other popular websites using the AWS cloud services were adversely affected.

Google

The globally recognized brand had all of its services down for five minutes on the afternoon of August 16th2013. During that time, the company’s traffic dropped by 40% and it is estimated that it lost $550,000 in revenue.

While the recovery time was clearly impressive, the cause of the glitch remains unknown to the operators. Due to the massive number of Google users globally, this outage is significant more for its scope of its effects than financial repercussions. Millions of users were left unable to access Gmail, Google Search, YouTube and Google Drive resulting in quite some outcry. The outage also indicated just how much Internet users rely on Google, proving its dominance on the web.

Go Daddy gone!

Go Daddy is one of the most important DNS service providers with 5 million websites and 50 million domain names in its care. So when corrupted data caused an outage on September 10th2013, it affected millions of websites all over the world.

There was speculation of a hacker attack which was quickly denied by Go Daddy. Services were restored 9 hours later and it was estimated that over $50,000 in revenue was lost.

The Domino Effect

Friday, 26thOctober, 2013 was a strange day for the web. Following the outage of Amazon Web Services earlier that week, Dropbox, Google App Engine and Tumblr, all came crashing down that Friday.

As Google App Engine started having slow performances and dropping connections, Tumblr began having network problems with one of its uplink providers. Dropbox also had some problems about the same time, and all three outages resulted in a significant drop in web traffic. Major losses were also reported in North America as a result of these outages.

Hurricane Sandy

Mother Nature came into play in 2013 when this super storm occurred. Hurricane Sandy delivered unprecedented challenges to the web. When the storm hit New York and later New Jersey, data centers in these regions faced the impacts of massive flooding and persistent power outages. These data centers host some popular websites such as Huffington Post and BuzzFeed.

The power outages persisted for the week requiring extraordinary measures to be taken in sustaining the backup power systems and keeping generators running, which involved some website staff carrying fuel 17 floors up for 3 days to ensure there was no downtime.

The erratic power situation caused some sites to be down for the entire week, but overall, the effects of the storm began an ongoing conversation about location, disaster prevention and recovery measures that can be taken to protect data centers from similar occurrence in the future.

To know more visit remotedba.com