Suffering a cloud outage? Look closer to home for the potential cause
If your systems are down, and you’re just about to get on the phone to tear a strip from your cloud vendor, remember this: technical errors in failed cloud implementations are more likely to come from the user organisation itself than the supplier.
That was the surprising finding from a report recently released by The Economist Intelligence Unit, with 36% of respondents saying errors were more likely to come from within than outside (29%). The report noted commercial errors were the most common type of supplier failure, with a third of respondents said they were unaware of any failures in the cloud infrastructure they used.
Public cloud services were more likely to throw up technical failures than those using the private cloud, according to the report. Downtime is most likely to come from what the report called ‘significant outages’ (23%), as well as failure to integrate with existing systems (20%) and data breaches (17%). Inevitably, a skills shortage was also blamed for exacerbating, if not directly causing, technical issues, as well as a lack of business continuity and disaster recovery planning.
The report notes: “It would be misleading to state that public cloud is always riskier. In the early days of the cloud, users may have experienced greater security issues since the technology was not yet mature and because of their own inexperience. Conversely, the bespoke nature of private clouds allows for a greater level of security, though with possibly higher initial costs.”
In a more positive vein, the survey showed that when cloud failures did occur, they were rarely catastrophic, with only 9% of those polled saying their incidents were “high” risk overall. 34% opted for “medium”, while 55% said damage was “limited.” Significantly, the loss of customer data was the biggest fear executives had over a failed cloud implementation.
The report concludes by assessing the maturation of cloud computing, calling it a “core component of the IT landscape”, yet arguing there is still more work that needs to be done, in terms of skills implementation and improved disaster recovery strategies. Businesses, wherever possible, need to be proactive, not reactive – and given Databarracks recently argued disaster recovery as a service (DRaaS) was going to be the most important cloud service in 2015, many seem to be on their way.
- » Hyperscale operators invest hard in data centres amid modest overall capex, says Synergy
- » Cloud computing accelerating climate change is a misnomer, scientists find
- » How AI is bringing a new dimension to software testing
- » Waste not, want not: How enterprises can avoid an idle cloud estate
- » What is cyber insurance truly worth? Analysing the risks and responses