The popularity of cloud computing services keeps rising, with providers like Amazon and others offering public cloud hosting services at reducing prices and growing bandwidth capacities. Still, even if this computing technology is virtual, it’s far from infallible.
Why Should You Care?
Cloud computing outages like a failed component or total service meltdown can’t be avoided at all. When they happen businesses lose their access to important data or resources and inevitably experience downtime which impacts the revenue. How quickly you recover from these largely depends on the quality of your cloud contingency plan. Your business continuity plan is really something that cannot be overestimated, especially in business world. Your physical network already has a plan covering all eventualities and you need to build another for your virtual one.
Tips On Planning For Outages In Your Cloud Environment
Your plan for cloud outages should include such contingencies as for continuity, data recovery, and infrastructure failures (in extreme cases). While the parameters of the plan will vary across different business sizes and patterns of data usage, there are some basic tips every company should refer to in this matter.
Your contingency plan should include an updated list of data access permissions, dependencies, and data controls. This is important because if you have no idea what your cloud’s running, it’ll be very difficult to put everything back.
Again, if your business mission-critical data and key services are heavily cloud-dependent, you need to make a comprehensive map with all the core components included and the recovery order described. Normally, your company won’t need to recover the entire data to keep functioning during such an outage, but knowing what and in which order you should recover will save you much time and effort when the worst happens.
- Consider user experience issues
Since the end user usually experiences the full effect a cloud outage can cause, it’s critical to factor your UX issues into the contingency or backup plan. Your aim during the design and building phases should be to minimize the disruption from the point of view of a user. If the strategy is properly formulated, the user access to some alternative resources will be allowed in some specified time, like a service-level agreement usually provides for.
- Get control over the cloud traffic
If you are in control of wide-area, or WAN, network traffic in your cloud, your network performance is optimized and you are able to direct your users when a network outage occurs. A cloud environment manages data so that it is spread across different locations and times of access, and when these are added up, the process can be slow down considerably. That’s why understanding your WAN traffic as well as optimizing your network is a great practice usually achieved by creating as efficient network paths connecting your data sites as possible. A few software solutions easily handle this traffic and are deployed at virtual routing layers. These include, for example, Riverbed and Silver Peak.
Finally, it’s crucial for every environment to be extensively tested along with subsequent documentation. However foolproof your contingency plan may look, it’s just as important to test it in real-world operation conditions to make sure it’s perfectly dependable, especially in an organization which is based on cloud infrastructures. Providing every test with proper documentation will prove invaluable in similar, actual events later.
To sum it up, the exact contingency measures in cloud outages differ among businesses, but planning for disasters won’t be much hassle for you if you know for sure:
- what your cloud is running
- what data needs your users have
- what data is critical for your company to keep running smoothly