What I find really weird is that companies have optimised their software, data, and cloud infrastructure to be utilising a single cloud provider. ????
Whereas, what should be happening is they should be optimising the usage of different clouds to benefit from the best technology and prices.
Companies feel they have to stay with a single provider. As if they owe something to the cloud providers.
It doesn’t have to be like this. For the purpose of data analytics and heavy data engineering, you can use Google BigQuery and Bigtable. And for other needs, you don’t actually have to stick with GCP. You can use AWS if you want top security out of the box. If you want a lot of phone support, Azure is the best. They will happily help you give you all the phone call support.
However, that’s not necessarily true for AWS (Amazon Web Services) or GCP (Google Cloud Platform).
I believe that’s how it should be. We should be utilising the best cloud solutions from different providers so we can stay cost effective and release software quicker. We shouldn’t be putting all our eggs in one basket.
You can also use multiple clouds for additional failover solutions. For example, if Azure services are down in a region near you, you could use AWS services. If AWS is down, you can use GCP.
At the end of the day, different cloud providers are just data centres and they do get spikes and services outages. This is more common now due to so many companies doing their cloud transformation.
You can setup Kubernetes and load balance between multiple clouds. This is just one way of mitigating the risk of relaying on one cloud and having the chance of failure more likely.
If you have a cloud solution sitting in a single cloud and you would like to diversify and use multiple clouds, drop me a message. I can help you plan out which cloud technologies would be best suited for your needs.
Zain Daniyal is a Data Strategy Consultant.