Let’s dispel 3 myths on cloud computing

Make no mistake, the cloud is here and it’s here to stay. The cost savings, flexibility and agility benefits cloud computing offers are too great to assume otherwise.

However, many organizations are struggling to figure out exactly how to most intelligently leverage the cloud; specifically, what infrastructure elements to migrate off-premises and which to leave on-premises. In fact, according to a recent SolarWinds survey, three out of five organizations have migrated less than 25 percent of their infrastructure to the cloud at this point in time and one in ten have not yet migrated anything, though plenty plan to do so once they figure out what and how.

On that note, one of the most controversial infrastructure elements when it comes to cloud migration is the database, with a whopping 72 percent of IT decision makers hesitant to move their databases to the cloud, according to another survey. But why?

Well, the database is the heart of most applications, and often the most complex and critical piece of infrastructure when it comes to performance. This leads to a lot of fear that making any changes to the way a database operates, or where it operates in this case, could have negative effects that ripple far and wide.

While from a high level this fear is justifiable, I believe there is some specific fear out there preventing more from considering cloud databases a viable alternative that is based more on myth, or at least misconceptions, rather than modern reality.  Three cloud database myths in particular come to mind: lack of performance, comprised security and loss of control.

Let’s do some myth busting.

Myth 1: The cloud is not a fit for databases that require high performance and availability

This myth, as many do, has roots in reality, but has become outdated. Yes, several years ago in the quite early days of cloud computing, consistent performance and predictable availability were issues for databases. But with the natural maturation of the technology and the introduction of vital cloud performance monitoring, management and tuning tools, we’re at a point where databases in the cloud can meet the performance and availability requirements of even some of the most demanding applications.

In fact, the architecture of today’s cloud storage systems, often based on SSD drives, storage-optimized instances and guaranteed performance options, offer up to 48,000 IOPS and 800MBps throughput per instance, more than enough performance to meet the requirements of most organizations. And yes, while two to three years ago, outages in the cloud were a commonplace occurrence, today’s cloud provider SLAs combined with the ease of setting replicas, standby systems and the durability of data stored in the cloud often exceed what most IT departments are able to deliver with the average on-premises systems.

This isn’t to say, however, that the onus is completely removed from you, the database administrator (DBA). While giving the cloud provider control of your database infrastructure may remove some of the heavy lifting from you that is involved with administrative and configuration tasks, you are still ultimately responsible for overall database performance. Thus, you still need to pay attention to resource contention, bottlenecks, query tuning, indexes, execution plans, etc. This may mean new performance analysis tools are needed, as many do not work properly on the cloud, especially those requiring agents, but some work great.

Myth 2:  The cloud compromises database security

Now, since security should always be a concern and you should always assume your data is not secure enough, there is some legitimacy here. However, just because you can walk into a server room and stare at the server racks doesn’t mean the physical hardware is more secure than the cloud. In fact, there have been very few high profile security breaches in the public cloud, while such breaches involving on-premises data centers continue to mount.

The reality is that the cloud can be secure, you just need a plan. This should start with not expecting the cloud provider to completely take care of security for you. You need to think of security as a shared model where the cloud provider provides reasonably secure systems and tools like encryption, but you are just responsible for secure architecture and processes.

This begins by being very clear about the risks you are trying to mitigate, the corporate security regulations that need to be followed and the compliance certifications that must be achieved, all on a workload-by-workload basis. Then you must work with your cloud service provider to jointly build a plan to meet workload requirements.

You also need to develop a thorough understanding of your cloud provider’s security model, capabilities, tools and processes. For example, you need to understand their encryption and be involved in considerations such as which keys will be used, who has access to them and what algorithm will be used to do the encryption. You also need to understand how your database will be monitored for malicious or unauthorized access. With this depth of understanding of and involvement in the security your cloud vendor provides, you can ensure your data is as safe, if not safer, than if it were down the hall.

Myth 3: The cloud means I no longer control my database

Conclusion