The month of June saw four cloud computing outages where in Google Gmail, Amazon Web Services, Apple iCloud and Twitter saw outages ranging from 2-4 hours and these are not big outages when compared to previous outages that these companies faced in the previous years that lasted for days. But millions of users and companies like Quora, DropBox, Pinterest, Heroku, etc who use the cloud totally or partially for running their day to day operations were severely affected which also led to a lot of negative chatter on social media platforms like blogs, discussion forums, Twitter, Facebook, etc. Despite many precautions taken by cloud service providers there have been outages on a regularly basis caused majorly due to human errors, quality issues, technical glitches and natural disasters which also highlighted that there is no escape from cloud computing outages and companies have to include the outage risk concerns in their data security and disaster management plans. Cloud Computing has become a vital part of the IT infrastructure of many companies and reliance on cloud computing is even more increasing in the near future to power business, government, consumer services, etc and major players in the space include Amazon, Rackspace, Microsoft, Salesforce, AT&T, Google, etc. Investing in Cloud computing is a significant IT decision that the CTO and his IT team along with consultations with CEO, CFO and other stakeholders have to make and also have to frame the necessary policy or upgrade the Organizational IT policy accordingly for the successful transition to cloud computing.
According to a recent report by the International Working Group on Cloud Computing Resiliency, every year minimum of 10 hours are lost because of service disruptions and according to the thirteen biggest cloud computing service providers since 2007 a minimum of five hundred hours has been lost, which also translates in monetary terms to be worth a minimum of $70 million. In that same report, the group claims that a cloud computing service is usually down for an average of 7.5 hours each year, although an electric power service outage is pegged at a low 15 minutes yearly. The group gathered the data from various sources such as Twitter, Amazon, Google, Paypal, Yahoo, Microsoft, Facebook, and others. Cloud service analysis firm Newvem says 40% of Amazon's biggest cloud users are not ready for the next outage which can be totally blamed on those users only as they don't follow rule No. 1 in computing: make backups. This is a major threat for these companies as in case of a major outage they may loose critical and confidential data forever. Cloud users must create a back up of their data either on another cloud or on their premises. Some of the cloud providers like Amazon provide tools like Elastic Load Balancers, which automatically shift traffic around, and Snapshots, which automatically make backups. Cloud service providers should reduce the human errors, technology glitches, improve the testing process and prepare for tackling the natural disasters to avoid frequent outages and consumer confidence on cloud computing will be affected by frequent outages.
Apart from backing up the data cloud users too have to deploy the cloud across multiple geographical regions as highlighted by Amazon and it strongly discourages the practice of deploying in one region only as the two major outages on Amazon Web Services over the past two years were limited to servers in a single region (its
Eastern US servers). So
cloud computing services users must spread their workloads across various
geographical parts of cloud service providers in order to prevent being hugely
affected if an individual region experiences service disruption as for some of
the companies their websites going offline for couple of hours will lead to
significant amount of business loss both in terms of revenues and profitability
added with brand reputation loss and customers moving to their competitors. Cloud computing users should realize the fact
that cloud computing do not work on its own and the cloud service providers
will take total responsibility for the data safety and smooth running of the
business operations rather they should carefully monitor and manage the cloud performance and also have a back up
and disaster plan in place especially for what will happen in the event of a
service disruption. Companies should be very clear on how to integrate the
cloud computing into their IT infrastructure and should also have total
understanding of the cloud computing limitations and should have a plan in
place to tackle the risks. Both the
users and providers of cloud services should have the relevant contracts in
place and should also be very clear about the service level agreements.
List of Cloud Computing Outages in 2012: