Choosing a DBaaS Provider: Why Cost Should Never Be the Sole Criterion

essidsolutions

Implementing a new database platform requires a great deal of research within an enterprise. Purchasing a database management system is an arduous process, considering the enormous investment needed for new hardware and software system components. While DBaaS providers have taken some of the trepidation out of the procurement process, choosing a DBaaS provider isn’t something you should take lightly. Cloud-based services offer agility and elasticity, but cloud contracts aren’t flexible enough. That’s why it is essential to take your time and make sure you are contracting for the right solution. Below are four top considerations when selecting a DBaaS service provider. 

Choosing the Right Data Model

The first thing to consider is what data model you need. We traditionally think of relational databases when it comes to data storage. Options include Microsoft SQL Server, MySQL and Oracle. In recent years we have seen another database technology called Not Only SQL (NoSQL) come to fruition. These include MongoDB, Redis and Cassandra. So which is best for your needs?

If your database transactions need to be ACIDOpens a new window compliant, you can choose a relational model. ACID stands for Atomicity, Consistency, Isolation and Durability. It is a set of principles that help guarantee that transactions are processed correctly. A perfect example is a banking transaction, as both parties must believe that the transaction was processed accurately. An e-commerce transaction would be another example. Any of the mentioned SQL platforms can fulfill these requirements.

If you are looking for speed, you probably want to go with the tried and tested SQL programming language. Relational databases are highly structured, which allows data queries and modifications to be performed quickly. Data is organized into rows and columns. Because SQL technology has been around for so long, it is relatively easy to find people that have worked with it.

In comparison, unstructured data is somewhat of a mishmash of stuff. Unstructured data does not require a defined schema, and it resides in numerous formats. This includes information such as sensory data, social media posts, voice mails or article content. The fact is that a vast majority of generated data today doesn’t reside in a nice organized table. Unstructured data resides across multiple locations, and due to its unorganized nature, processing consumes more time and resources than its organized counterpart. Because of the relative newness of the technology, not many people are familiar with it as well.

Learn More: Running a Database-as-a-Service in Kubernetes

Security and Encryption

An organization has to secure all sensitive data, but it is often difficult to define what isn’t sensitive. Organizations must safeguard financial transactions, but voicemails can also hold sensitive information. Ideally, it would be best to encrypt everything, but encryption takes a toll on performance due to the heavy processor load of encrypting and decrypting data. Those organizations that must meet industry or governmental compliances must ensure that their DBaaS provider can satisfy all mandatory regulations.

At the very least, you need to encrypt all data while in motion using the most recent TLS version, namely TLS 1.2. This will prevent hackers from intercepting any data while being uploaded. Then there is data at rest that requires either full disk encryption at the file system or block-level and Transparent Data Encryption (TDE). 

TDE will involve an encryption key that you must secure and periodically rotate as well.  If you can’t afford the resources necessary to encrypt everything, then you must decide which data deserves protection. You should create policies to apply the right levels of encryption and security to different data types. DBaaS technology is currently advancing to the state that data will no longer have to be decrypted to work with it, saving time and money.

Encryption isn’t the only aspect of security. Your database must have either built-in security or integrated security to authenticate users. Built-in means that usernames and passwords are stored within the database engine itself, while integrated security requires a user directory such as LDAP. Once a user is authenticated, they must be authorized to access different data segments. Authorization can, however, quickly become complicated in unstructured data environments.

Make Sure You Can Exit the Building

It would be nice to say that every match with a DBaaS provider can last forever, but that’s not the case. Needs and technologies change over time, and no one can predict the future. So while you take time for your due diligence when deciding who to best partner with, make sure you spend some time looking into the exit point as well. The last thing you want is to be locked in with a DBaaS provider that isn’t meeting your needs five years from now. 

To avoid vendor lock-in, you need to ask whether their APIs, database engine or automation tools are proprietary or industry standards. Will you have to refactor your code to move it someplace else? Can their systems integrate with other clouds? After all, there isn’t just “one cloud” out there. You also need to find out if there are exit fees involved and how easy it is to migrate data from their platform. After all, data is constantly being updated dynamically, so moving data that is continuously in a transitive state is difficult already.

Learn More: How New Security and Encryption Layers Strengthen Cloud Databases

Read the Fine Print

One of the reasons that enterprises love cloud service providers is that it saves them from attending to rigorous details such as system maintenance, updating and support. Of course, the downside is that you become dependent on someone by passing over the responsibility onto someone else’s shoulders. Third-party reliance is a fact of life when it comes to adopting cloud services, and some providers are better than others. Be sure to read the service level agreements of all proposed providers to compare apples to apples. Also, ensure you understand their availability, who you are supposed to contact and the chain of command if you aren’t getting the attention you think you deserve. 

Calculate the Real Cost

Then, of course, there is the issue of cost. Don’t just count on a provider’s marketing calculator to crank out a rough estimate for you. Do the math yourself if necessary to fully understand how much their cost structure will affect you. Different providers use different equations, but in the end, what matters is the final figure.

Conclusion

Whether you are purchasing for on-premise or in the cloud, database systems are elaborate, complex and expensive. Don’t take any shortcuts in the decision process.  

Do you think organizations are carrying out sufficient checks when choosing DBaaS providers? Comment below or tell us on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We would love to hear from you!