Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

How Much Can We Trust The Cloud?


Disaster Recovery Testing should be a top priority for all businesses

The recent failure of AWS is a wake up call for business owners

Cloud services are increasing at a rapid pace, with no signs of slowing

Should businesses trust cloud services with their backups? How good is the DR plan, really?

The answer is to the first question is mixed. More and more businesses are relying on cloud services to back up their business systems and data, but should they? With the way technology is advancing, the answer is basically a yes, but as we have recently seen in the news, there is increasing proof that you should not entirely be reliant on cloud services. A few weeks back, Amazon Web Services (AWS) went down due to a human error. The closest number I could find as to how many services AWS actually utilizes seems to be around 1.3 million servers. Every day, AWS installs enough server infrastructure to maintain their entire e-commerce site. While it was difficult to find exact numbers, those facts are very telling, and it only continues to grow.

When AWS went down, it caused issues ranging across the entire world. That alone should scare the hell out of CEOs and BODs. But this event brought us back to reality a bit. It was a good reminder that while cloud hosting is very useful and helpful is most circumstances, things can always go wrong. As previously stated it was caused be a human typographical error, which had massive implications. The article published in the Data Center Journal, discussing this issue and also the British Airways fiasco that occurred a couple months back. The journal stated that the AWS event caused so much error with their system, that they could not even open their dashboard to notify users of the issue. Moving over to The British Airways incident that shut all of their systems down completely for a period of time, is another great lesson that just because a company has a DR plan does not mean it is effective, or in their case will work at all.

When BA’s systems went offline completely, their 2 backup data centers (LESS THAN A MILE AWAY), did not kick on. Testing a DR plan and various scenarios is just as important, if not more, than actually having one. Reports of employees and passengers on BA stated that both parties had no knowledge of what happened, or how to proceed. Now, while this wasn’t a bomb or a natural disaster that passed through, a situation like this one, where no was informed on procedures or damage control with irate passengers, it is the same situation really. Critical functions were no longer possible, where do we go from here? That question wasn’t able to be answered by employees on the service level.

The moral here is that companies and businesses are becoming more reliant on cloud services, so much so that estimates say many businesses will have 0 data centers and everything will be hosted in the cloud in the next decade, but there needs to be a backup to the backup it seems. As for Disaster Recovery plans, it is evident that testing the DR plan is a crucial piece to whether it will be effective when/if the time comes where they need to be put into effect.