While cloud services have promised advantages of redundancy and resilience from the start, there is still the spectre of failure. Even the largest operators can be affected. Amazon’s EC2 (Elastic Compute Cloud) suffered breakdowns in 2011 because of a wrongly applied change of configuration, and again in 2012 owing to ‘historic’ thunderstorms in the neighbourhood of one of its data centres. Facebook recently experienced significant slowdowns in displaying user pages. But what if these operators could now put self-healing technology in place to avoid such occurrences?
That’s the promise of a start-up cloud solutions provider called HybridCluster. The underlying observation is that cloud hosting companies are still exposed to unpredicted breakdowns or rapid fluctuations in traffic, as well as users’ mistakes. The HybridCluster solution targets these problems to help providers improve performance and availability. What will you see as an end-customer? With a bit of luck, almost nothing. The cloud service you use for example as part of your data storage and backup strategy should run without problems of the kind described above for Amazon and Facebook.
What you may see is simply added functionality made available to you in what HybridCluster terms a ‘time machine’ for rapid restoring of data. This is designed to let you move back to the last good version of your files in case you accidentally provoke data loss or suffer hack attacks on your databases. Otherwise, the self-healing (allowing an entire region to recover within one minute) and auto-scaling functionality of HybridCluster should guarantee that for your cloud resources usage you can ‘set it and forget it’. The company has recently received funding to produce the next version of its software for cloud providers, so you’ll have to stay tuned to see who decides to run it and what the effect is.