DBAs shouldn't be afraid of managed services

DBAs have raised a lot of objections about moving from on-prem (or self-hosted) to using managed services. In this post, we address each in turn.

02 December 2019
John Hammink
John Hammink RSS Feed
Developer Advocate at Aiven

A lot of DBAs these days are starting to warm up to the idea of using managed database and messaging services. And with good reason, which we’ll get into below. But many are still resistant to the idea, as are their colleagues in DevOps, Data Engineering, IT, Data Science, ETL and others. But why?

In this piece, we’ll consider the many objections that DBAs and others have with managed services, look at some cost considerations, and finally look at some of the real benefits of managed database and messaging services.

They took our jobs! Or did they?

Those who follow South Park are probably already very familiar with this particular meme. As silly as it’s presented, this serves as possibly the biggest (often silent) objection we’ve seen to adopting managed services so far.

What about the jobs? What happens to DevOps, Data Engineers and DBAs when managed services are adopted to replace on-prem or self-managed cloud services?

You can be certain there will be less need for on-premise data pipeline engineers moving forward. In the open-source (and software) world, one of the main aims is to build up and share foundations upon which future innovation can take root. As engineers, we don’t want to continually reinvent the wheel and build the same things from scratch.

As a manager, do you actually prefer spending your budget on database infrastructure experts and maintenance engineers? It’s a struggle to find talented DevOps people anyways. And, what about learning costs: do your data scientists and business analysts need a help desk to work with the pipeline?

A lot of these problems are actually solved with a managed solution. Over time, more components, apps and storage schemas will be abstracted away and simplified to offer best-practice complete drag-and-drop pipelines as managed services.

And as we shall see, this means there will be more room in the future for experts who can work with data in more focused, novel, and competitive ways.

Shifting the locus of control

Taking managed cloud solutions into use means there will be less control by the in-house staff over the minutiae of database setup and management: less need for sharding and replication, hardware configuration, set up, updates, reliability, latency and throughput engineering.

For some, this is a cause of concern; sometimes out of a desire to control some of these variables. However, there’s risk associated with that responsibility. With a managed cloud solution, that risk is eliminated. What’s more, there’s a whole lot more freedom to do work in the rest of the data stack.

What are managed services? What is DBaaS?

A managed service is where a complete package of processes and functions are packaged together and sold to customers to improve and simplify operations as well as slash expenses. Managed services involve a relationship between a service provider who builds and maintains the services as well as a subscriber or customer who buys those services. These two parties are bound by a contract known as a service-level agreement (SLA), which outlines the agreed performance, latency and other quality requirements of the service. Typically, managed services are purchased by the hour.

A database as a service (DBaaS) is a refinement of that concept which focuses on databases. DBaaS allows the subscriber to outsource the hosting and management of the database. So how does it compare in terms of costs to consider?

Costs to consider: On-prem vs. DBaaS

As you can imagine, it’s not just about hardware outlay costs - there are other items as well.

On-prem tasks include monitoring, regular upgrades, configuration tuning, updates, extension management, tracking software updates, data migration between versions, test suites, ops on-call duties, certification management, and network encryption.

What about the team you’d need to hire to handle all of that? In fact, a mid-size setup for a company with 100 employees could require around 2-4 people maintaining the databases and servers continuously. If you can find the right talent stack, this may be doable — and scalable — for companies of this size. But when hiring for much smaller teams, It’s less likely you’d find one person with the chops to do all of those things. This is one of those decision points at which a managed cloud service becomes essential.

Nonetheless, if we consider an on-premise solution with a three-year depreciation period, the monthly cost of buying hardware and electricity alone for equivalent servers is about one tenth of Aiven’s total pricing.

Have no fear, DBAs!

But let’s look at some other significant cost drivers that fall into the buckets of infrastructure, network and maintenance costs:

Breakdown of data-management costs by category.

With a managed solution, customers need only be concerned with the italicized features, while enjoying the ability to focus on the ones in bold type.

Or, put differently, your in-house technical staff can focus on tasks that are higher up the value chain, as follows:

Hierarchy of DBA time spent: self-hosted vs. Managed service.

In today’s data-intensive environment, new technologies need to be adopted quickly and purchasing long-term database solutions under enterprise license agreements (ELAs) has introduced risk because such a model is slow-moving by definition.

At the same time, the larger the group of stakeholders maintaining an on-prem, open-source database, the less flexibility users can expect when using the database to its maximum potential.

Other benefits

So what if you could dedicate your data staff resources almost exclusively on the following tasks instead:

  • getting real insights and analytics from ALL the data
  • discovering new things to do with data
  • providing data engineering support (focusing on on data analytics)
  • data science
  • modelling and testing data models
  • data visualization
  • building and training AI and ML
  • building business end-applications?

Imagine being able to up- and downscale your services based on actual usage. While this is not easy to do with a self-hosted or on-prem solution, such flexibility clearly does drive real savings. On managed cloud services like Aiven, usage levels will inevitably be higher due to superior availability: when a node goes down on a hosted service, Aiven’s automation recovers it almost immediately.

In addition to allowing you and your team to focus on your core business, Aiven gives you technical support 24-7 that you’ll never get from a self-hosted service. And Aiven’s DBaaS and hosted Kafka are ISO 27001-2013 and SOC2 compliant, so you know exactly where you stand with industry standard compliance and trust.

Wrapping up

In this piece, we’ve looked at the many objections that DBAs and others have with managed services, examined cost considerations, and zoomed in on a few of the real benefits of managed databases (and Apache Kafka) as-a-service.

The reason cloud computing in general is seen as revolutionary is because it brings elasticity to organizations. With it comes predictability that transforms heavy one-time capital expenditures (CAPEX) into reliable monthly or annual operational expenditures (OPEX).

Most importantly, cloud computing — and managed services — allow businesses to focus on their core strengths: your data talent get to focus on unique value creation, rather than on maintaining infrastructure.

The best way to truly understand what a managed service can do for you is to try it out! There’s a no-commitment, 30-day trial to peruse. Don't forget to follow our blog, changelog RSS feeds, or catch us on Twitter or LinkedIn.

Start your free 30 day trial

Test the whole platform for 30 days with no ifs, ands, or buts.

Aiven logo

Let‘s connect

Apache Kafka, Apache Kafka Connect, Apache Kafka MirrorMaker 2, M3, M3 Aggregator, Apache Cassandra, Elasticsearch, PostgreSQL, MySQL, Redis, InfluxDB, Grafana are trademarks and property of their respective owners. All product and service names used in this website are for identification purposes only and do not imply endorsement.