February 22, 2018
DBaaS – Database as a Service Pros and Cons
Almost every business these days is data-centered. Whether the data is for internal applications and systems or for other services that are offered, let’s face it,
managing data is a key to success.
Before listing the pros and cons of DBaaS, we need to explore a few decisions business have to make.
Businesses are faced with numerous quick decisions about data handling that can set them on a path that, if incorrect, is difficult and costly to correct. Those decisions around data are:
- What database type to use, SQL or No-SQL? This is an article all by itself.
- What are the data storage and query needs? Transactional? Big Data?
- What database system to use? A few SQL choices might be Oracle, MySQL, MSSQL, and Sybase. A few No-SQL choices might be MongoDB or Cassandra.
- Do we have DBA (database administrator) talent or do we have to hire?
- What kind of server or resources are needed? What are my power, server, disk, processing, network, and IO requirements?
- How do I maintain, backup, administer and otherwise own the database framework?
- What is my cost of ownership?
Let’s explore the first decision. What database type to use, SQL or No-SQL?
Traditional database types that are classified as SQL have a significant place in businesses and are a mainstay for business choices, however, as companies start to create applications that drive decisions based on significant database analysis of large, almost unfathomable amounts of data, they migrate to No-SQL solutions like MongoDB or Cassandra. The architecture of No-SQL makes it a good choice for big data solutions while the built in protections of a transactional based system like Oracle make it a better choice for banking or similar solutions.
When it comes to picking a specific system, businesses tend to stick with what they know. In other words, if they already have Oracle, and Oracle talent, then when management asks those individuals which database system they should use on Project X, it should be no surprise that they pick…Oracle.
Matching a specific database system to a set of business requirements is an arduous task that should always be looked at with a fresh perspective and not just based on what talent is already employed or what systems a business is comfortable with.
Let’s face it, if a business picks correctly, all is good. If they pick incorrectly, they have wasted a lot of resources which equates to dollars. Enter DBaaS.
Where DBaaS excels is that it gives businesses the ability to test the waters a bit, to try before they invest heavily.
Even if DBaaS is a stepping stone to total ownership, it is a cost effective solution to help figure out your needs prior to investing heavily.
Let’s explore the pros and cons.
Note, we need to distinguish between “hosting database systems” and DBaaS. There are many cloud based solutions that “host” a database system but provide no significant help in configuration, tuning, consulting, and providing the talent needed to actually use those systems. True DBaaS provides both the system and the talent to help you utilize the database and figure out how to store, query, and analyze your data.
The value of DBaaS goes way beyond the hosting.
Pros of DBaaS:
- No equipment or software licenses
- Flexibility. Multiple choices to test drive your applications and pick the right platform for your business requirements.
- Significantly less staffing requirements. The DBaaS provider handles installation, configuration, and in many cases development.
- Offsite provides protection from local power failures or disasters. Many businesses design with power redundancy in mind but rarely meet those goals in reality.
- SLA agreements that have redundancy, up-time, and backup protections. A DBaaS provider has intent focus on protecting your data.
Cons of DBaaS:
- Limited access to underlying servers. This can present itself as a feeling of no control.
- Very little knowledge of how your data is protected from cyber security threats. This can be dangerous for sensitive data.
So how to you decide?
Is there a transition from one to the other. Yes there is almost always, but by following a few guidelines to start with, you can properly use DBaaS.
Forensic IT’s Guidelines for DBaaS:
- Do all development using DBaaS. This is your chance to test drive different architectures and features.
- Unless you have full disclosure of how your data is protected, managed, and secured by DBaaS providers, we suggest consulting with database architects to host sensitive data internally. Note, this is typically not big data. When we use the terms sensitive data, we mean just that. Data like SSNs, account details, financials, personal data, etc. Does this mean that you cannot use DBaaS for this? No, it means that you first have to find a DBaaS provider that will show you everything from how your encrypted data gets in their system to storage, access, etc.
- When you are not sure of what your database needs really are, use DBaaS first. This lets you try SQL or No-SQL. This lets you explore the encryption capabilities of Oracle versus MySQL. Think of DBaaS like buying a car. You test drive sedans, trucks, and SUVs, and try different manufacturers and features. You may decide to lease or buy.
- Always monitor and evaluate the cost of ownership. As your system grows, the operating costs might make sense to drop DBaaS and build an in-house system. By then, however, you have already decided on what you really need. You have test driven multiple solutions and can now buy only the licenses and hardware you know you need to be successful. You can hire the correct talent to manage your system.