Continuous learning is essential to keeping up with new and changing technologies that will impact all businesses and data professionals’ jobs in the coming years. During our free webinar, The Top 4 Data Technology Trends Every DBA Should Know About in 2019, I was part of a panel of data technology experts, including dbatools.io Creator Chrissy LeMaire (@cl), Sabin.io DevOps Data Engineer Mark Allison (@dataguzzle), and Heraflux Technologies Founder David Klee (@kleegeek), who discussed the following four data technology trends:
- DevOps for databases
- Cloud computing, administrated by PowerShell
- Containers and orchestration
- Machine learning, the natural next step for data
Let’s take a closer at each of these trends and how they will affect you and your business.
Trend #1: DevOps is for the Database Too
DevOps has traditionally been used on the application side, allowing developers to continually push out code in small increments and, if any problems are discovered, roll the code back. The promise of reliable releases and reduced risk is expected to drive the adoption of data DevOps.
Planning is key to successful DevOps implementation; changes need to be designed so that they’re small and have minimal complexity. Once developed, the code should be checked into a source control tool such as GitHub Enterprise, Jenkins, Azure DevOps (formerly Visual Studio Team Services, or VSTS), or any other tool selected by your organization.
A golden schema, where all databases within a multitenant environment or across multiple clients are the same, allows changes to deploy to one or more databases reliably. Some companies deploy a small update across 500 or 1,000 databases at once. Once deployed, SolarWinds SQL Sentry allows these changes to be monitored so that performance problems with the new code can quickly be identified and fixed.
DevOps for Databases
“The main message is to make small, frequent changes and validate and test them as soon as you can. We use SentryOne for that.” –Mark Allison
Trend #2: PowerShell Is a Key Tool in the Expanding Cloud Computing World
Cloud has moved beyond the early adopters. Today, businesses of all types and sizes see the benefits of the cloud, whether for all workloads or just for disaster recovery (DR) and high availability (HA) in a hybrid cloud deployment.
PowerShell is an ideal tool in the highly automated cloud computing environment, streamlining most processes, including management, provisioning, orchestration, DR, and HA. In 2019, PowerShell is a must-have skill for DBAs, especially as businesses continue to move workloads into the cloud and in businesses with large SQL Server estates. The tool is ideal for automating repeated tasks, such as a database migration checklists and integration testing.
“PowerShell has made my life as a DBA so much more relaxed because I can restore an entire instance, no problem. DR is so much easier with PowerShell.” –Chrissy LeMaire
PowerShell is included as part of Windows and can be installed on Linux and iOS.
Trend #3: Containers and Orchestration Streamline Deployments
Virtualization allows multiple systems to run on the same physical system, each instance with its OS and applications. Containers and orchestration streamline deployments to virtual machines (VM), allowing VMs to spin up quickly and without the significant overhead associated with deploying each VM individually.
Without containers, a VM is deployed with a full copy of the OS and all required applications, which can be time-consuming and expensive, especially when multiple VMs need to deploy at 20-25GB per deployment. A container strips out the OS and packages up just the necessary applications and their dependencies, decreasing both the time to spin the system up and the hardware overhead.
“Containers really are the future. You don’t need to transport 50GB or 100GB of data just to move a 2GB database.” –David Klee
Orchestration uses workflows to automate deployments as well as spinning up new VMs or new Azure Functions or identifying the need to scale a system up or down. It uses a workflow process to determine when it needs to launch a process, such as when it sees critical errors on a server; it sends the workload to another server and shuts down the one that’s failing.
Trend #4: Machine Learning Turns Managed Data into Actionable Information
Machine learning allows businesses to turn the data they manage into actionable information that’s useful for decision making. It can be used to answer five question types:
- Is it A or B? For example, what type of writing instrument is this: a pen, a pencil, or a crayon?
- Is something an oddity or an outlier?
- How much or how many are there?
- How is this organized?
- What should I do next?
“The promised land for machine learning is telling you what you should do next. This is very hard to do with machine learning, but when done right, it is explosively valuable.” –Kevin Kline
One of the dangers with data sciences such as machine learning is that the system will answer the question asked, even if it’s the wrong question. Decisions based on an answer to a wrong question can lead to problems and perhaps even disastrous results.
Most successful data scientists or DBAs who begin using data science start small and slow, using a process such as the Microsoft Team Data Science Lifecycle. These data scientists become business subject matter experts, understanding not just what the business does, but how well the data represents the business.
The Microsoft Team Data Science Lifecycle
Eager to learn more about SQL Server–related data technology trends? Be sure to check out our upcoming webinars and in-person events, our blog, and our Resource Library for more content about the database performance monitoring and data DevOps topics that matter most to data professionals.
Is there a data topic or trend you’d like to learn more about? We would love to hear from you!