Multi-Cloud is a term used to describe IT infrastructure that uses multiple Public Cloud providers. Given that Cloud offers native redundancy, you may wonder why go that route? In the article below we explore several reasons why a Multi-Cloud provides an advantage to your cloud architecture.
Organizations opt to leverage multiple data centers for a variety of reasons. One strategic reason would be for disaster recovery. The old age idea is to spread risk across multiple platforms so you can minimize the possibility of downtime. Being able to failover to another data center, or multiple data centers, is common practice these days.
But what if you decided to put all your workloads into one Public Cloud with the intent to no longer use physical data centers? Do you still need the ability to failover somewhere else? It may seem counter-intuitive, but you do.
While Public Cloud providers have done as much as possible to convince you that they have redundant everything, that still doesn’t change the fact that they DO have outages from time to time. Most recently, AWS had an outage in their Simple Storage Service (S3) that affected numerous customers for approximately 6 hours. Given that S3 promises 99.999999999% durability, you can imagine everyone’s surprise when the service started malfunctioning.
The above outage is a prime example of why you might consider spreading your workloads across multiple Public Cloud providers such as AWS, Microsoft Azure, Cisco, Rackspace, etc.
Not all cloud providers are equal when it comes to compliance. You might choose one cloud provider to handle a set of application(s), but that provider may not be suitable to host other applications that have strict regulatory rules bound to them because of the type of data that is stored or handled. Examples of common compliance that is needed for your business might include HIPAA, HITECH, HITRUST, or PCI DSS.
Below is a published list of compliance offerings by Microsoft Azure. Microsoft says they have the most comprehensive set of compliance offerings of any cloud service provider.
The pricing for High-Performance Computing (HPC) vs. standard computing will vary by provider. And chances are, not all of your apps will need HPC. A Multi-Cloud approach gives you the option of hosting your HPC apps with one Cloud provider, and your standard apps with another based on the provider’s services and cost. By splitting your workloads across clouds based on performance needs, you can avoid the cost of compromise.
In the case of High-Performance Computing, most providers have a whitepaper detailing their HPC service, so pay close attention to which cloud meets your HPC requirements. At the end of the day, not only does your workload need the right server cluster, but it also needs these resources at the right price.
The location of your Cloud deployments can and will impact the speed of your applications. In the current age of immediacy, poor network latency can frustrate users and impact businesses. Within Cloud computing, latency is becoming an increasingly critical concern as the demand on Internet speed and connectivity continue to rise.
Cloud providers are racing to build data centers at a record pace, but not every provider has a data center where you need it. Your Cloud provider should have multiple geographic locations and offer dedicated network connections. If you have some applications that require dedicated bandwidth and others that do not, it may make sense to run these workloads in separate Public Cloud environments.
Not Enabling Your Competition
Because Amazon and Microsoft do more than just provide Cloud services, they may be your direct or indirect competitor. If that’s the case, you may feel that it would strengthen their position to give them any or all of your cloud business. Both companies offer a leading Public Cloud service, but you have other choices too.
As we mentioned when discussing redundancy, a single point of failure is never ideal in business. To put it bluntly, if you ever find yourself in a situation where you want to fire your Cloud provider, it will be easier if you have a Multi-Cloud strategy from the beginning. By going the Multi-Cloud route early on, you will have the tools and know-how to move your applications from one provider to the next, without much disruption or headache.
Added Flexibility and Agility
After considering the above reasons to adopt a Multi-Cloud strategy, you might feel concerned about adding complexity, time, and cost to your infrastructure. For Multi-Cloud to be practical, you would need the ability to configure and deploy your Cloud workloads from a single management console using the same procedures each time, allowing you to provision to any Cloud with ease. Luckily there are comprehensive Cloud Management platforms that will help you accomplish this. For example, Red Hat’s CloudForm and Cisco Cloud Center will enable you to mix and match the best technologies and services from the different cloud providers and manage your Multi-Cloud from a single tool.
Ultimately, this ability to continually mix and match as you see fit is one of the strongest reasons I am for a Multi-Cloud.
The idea of Multi-Cloud is not yet mainstream, but it is up-and-coming because of the many advantages it offers. Once companies understand that there are viable Multi-Cloud management tools available that provide ease-of-use, flexibility, and agility while buying down their IT risk, I believe Multi-Cloud will quickly become the new norm.