Cloud outages: what they are and how to mitigate their impact
When organisations today lose their link to the cloud, most hit problems. These range from mild inconveniences to total shutdowns of key workflows and income streams. Using the cloud is highly convenient and has transformed the way in which many businesses can work, but cloud outages happen, and have knock-on effects on organisations’ access to data and key tools.
It is essential to consider the likely impact of outages on your organisation. Outages might be something you decide you can live with. Or they might require detailed back-up and contingency plans.
Read on to find out what you need to know about:
- how your organisation might be affected
- steps you might want to take to mitigate the impact
What is a cloud outage?
The cloud is a vast set of servers that we access via the internet. Organisations store data on these servers in the cloud, such as those provided by Google, Amazon Web Services or Microsoft. Organisations may also depend on apps, SaaS platforms and security services that run from the cloud.
A cloud outage happens when these servers are, for whatever reason, unavailable.
Do cloud outages happen?
Yes – they happen to all the major cloud providers despite the redundancy and resources built into their platforms. Plus, an outage can have a domino effect, taking apparently unrelated services down with them.
In January 2023, a networking outage took down Microsoft’s cloud platform Azure for more than four hours, along with services such as Teams and Outlook, which hundreds of millions of people use. This single event knocked 3.2% off Microsoft’s share price.
In 2021 Amazon Web Services had at least 27 outages. One caused delays and shutdowns for seven hours, affecting a vast number of websites and apps. These ranged from Amazon’s own sites, to Netflix, to tools used by millions of smaller businesses. It even affected automatic cat feeders.
In the same year, Google Cloud went down for two hours in mid-November, taking services like Snapchat, Etsy, Discord, and Spotify with it.
Cloud outages happen, and it is essential to consider the likely effect of one on your organisation in advance so that you can mitigate any effect to your business of downtime.
What problems might an outage cause for my organisation?
There are four critical knock-ons from outages, and you should consider whether your business is exposed to these.
- An outage might mean your people can’t do their core work. Do they need access to cloud-based phone systems? To cloud-hosted customer databases or records? To cloud-based development tools or workflow-management apps? Look at the list above and consider how reliant you are on these cloud services and what could be used instead. If there are options, how quickly should staff swap to these alternatives? Otherwise, check the provider you have chosen has the necessary redundancy in place to make any downtime as minimal as possible.
- An outage might mean you can’t make products. With the advent of Industry 4.0, many production and manufacturing systems depend on cloud-based apps and data storage. Would a cloud outage force you to stop production?
- An outage might mean you can’t sell. Do you sell directly through your website? Are you an e-commerce organisation? What impact would customers not having access to your site have on your business? Would you lose sales? What message should you give customers during and following any problems and is it worth looking at additional contingency?
- An outage might put you at risk of breaching data and security regulations. Some sectors – from legal work to financial services – are highly regulated and your systems will need to maintain data integrity regardless of downtime. Many have to work to industry or government regulations about handling data and information. Have you built your systems to ensure a cloud outage would not leave you exposed to breaching data regulations?
- Would a cloud outage lead to any of these four outcomes for your organisation?
Which types of businesses are especially vulnerable to outages?
It’s hard to think of a business today that would experience zero effects from a major outage. The issues vary from mild inconvenience to serious loss of revenue or possible compliance breaches.
- How disruptive or damaging would a major outage be to your organisation? This will help you understand what an appropriate safeguard will look like.
Can we lose data through an outage?
It is unlikely as cloud data should be replicated at a second data centre, so even a catastrophic fire at a data centre should not lead to data loss.
However, organisations that store mission-critical data on the cloud still need to consider other possible ways that their data could be lost. The cloud is not necessarily as safe as it may seem. Problems – some of them catastrophic – hit organisations as a result of:
- Human error. Your SaaS provider is not responsible for data recovery that stems from mistakes made by your team.
- External malicious deletion. External hackers, ransomware and viruses target cloud data, mirroring organisations’ increasing reliance on it. Cloud data is susceptible to ransomware attacks, as we look at below.
- Internal malicious deletion. If someone with valid credentials deletes or corrupts material, cloud platforms will process it and there may be no way to recover the data lost.
Hold on, doesn’t the cloud provider do back-ups for me?
No. Your cloud platform is responsible for uptime and availability. But you are responsible for data protection and access management.
While cloud platforms do replicate your data, they don’t back it up. Replication copies your data to a second data centre. That minimises the effects and length of downtime – if a data centre goes down, you failover to another with a real-time copy.
But in a catastrophic failure, such as a ransomware attack or internal malicious data destruction, you would be wise to have to hand a backup held in a cold or offline location, as it would remain unaffected by any compromise in your live environment.
The point is that cloud platforms’ replication function is not back-up. Genuine back-up is vital for many organisations. And it is your responsibility – not that of the cloud platform – to put it in place.
What should we do about the possibility of cloud outages?
Any organisation should consider the possible effects of cloud outages on their work. But only 23% of UK companies have so far calculated how much outages might cost them.
In your outage contingency thinking, you should consider:
- The regulatory frameworks you operate within. What sets of rules must you follow in handling information?
- How dependent are you on cloud data and apps to conduct business?
- How damaging or costly would it be for your key workflows to be interrupted by an outage?
- What is the value of your data to your business?
- Are there critical data assets like customer databases that you would feel happier hosting or backing up in-house?
- How comfortable are you with storing crucial data assets remotely without a back-up?
- Is replication (rather than true back-up) enough for your purposes?
What can Beaming do to mitigate the impact of outages?
As one analyst in the field writes, ‘Instead of saying “My data is in the cloud”, try saying, “My data is on a rack of servers in an industrial park in California” – or it may be Warsaw, Washington, or Warwickshire. “So, what are the implications of that?”’
Beaming work with customers to think through implications of cloud outages and put in place contingencies that are bespoke to the scale and nature of your business. We guarantee that we will only advise you to put in place what you need.
Further recommended reading
- Not For Profit
- Data Security
- Remote Working
Advocacy for All
Data security is a hot topic that concerns everyone. For a charity that supports the vulnerable, a reliable and secure data handling system is not just critical; it’s mandatory.