Virtualization Technology News and Information
Percona 2022 Predictions: Cloud, Kubernetes and Skills Will All Affect Data

vmblog predictions 2022 

Industry executives and experts share their predictions for 2022.  Read them in this 14th annual series exclusive.

Cloud, Kubernetes and Skills Will All Affect Data

By Donnie Berkholz, Senior Vice President Product, Percona

According to a quote attributed to Bill Gates, people overestimate what will happen in a year and underestimate what will happen in a decade. With this in mind, what will happen around the cloud and data in 2022.

1. More and more databases will move to cloud

It's a simple observation that after the past few years, more and more data will be in the cloud. AWS at Re:Invent 2021 said it had migrated more than half a million databases using its cloud migration service, and there are a plethora of options to run your database instances across public, private or hybrid cloud environments. From using your own databases on cloud images, to getting someone else to manage your images, through to Database as a Service and serverless database options, there are more options and choices than ever before.

Behind this variety, there is a pattern emerging - developers care less about how they run any database, and more about the data it will provide. In practice, this means that they will use whatever option provides the right mix of immediacy, service and cost to fit their needs when they start a project. This will continue in 2022 as more workloads shift to cloud.

2. More problems will come up over time

Following on from this shift, more developers will start to find problems with their initial assumptions. From poor performance due to data being distanced from the application components or the wrong deployment approach being used, through to issues with cost and scale, developers will start to encounter problems after that honeymoon period is completed.

Two good examples of this in action are schema design and query performance. Developers are not experts at putting together schemas in databases, so they can easily overlook some of the best practices involved here. This can then affect performance for the application. If you don't look at this in more detail, you can then spend more upgrading to the next level of cloud instance - upgrading by credit card, in effect - rather than tuning or improving your queries.

The challenge here is not that things don't work. They all do what they describe. The problem is that developers expect cloud databases to be just the same as their internal deployments, but with all the headaches and management overheads magically removed.

The aim for many is that they don't ever want to think about databases ever again. Instead, they will be able to achieve this right up to the point that something goes wrong, or assumptions prove to be inaccurate. In 2022, we'll see more migration between different options for the same data start to occur as developers confront some of these issues.

3. Kubernetes will win for data, too

For companies deploying applications in the cloud, Kubernetes has become a de facto standard. According to a survey by the Data on Kubernetes community, developers in that group are running 50 percent or more of their production workloads on Kubernetes. For advanced users, they saw two times or greater gains on their performance.

Kubernetes works, and it makes it possible to deploy and manage applications at scale. It's open source, so it should help developers avoid lock-in to a particular cloud service. And it is being used more and more by developers as part of their workflows and processes.

However, the same cannot be said for data yet. This will change in 2022 - around 90 percent of respondents to the DoK survey thought Kubernetes was ready to support their stateful workloads like databases.  That early work on Kubernetes operators and management tools will see running data on Kubernetes change from being an option for those at the cutting edge of cloud native to being suitable for mainstream companies to use. The result is that developers should be able to manage their applications and their data in one place, making their lives easier.

4. Lock in will be an issue for many around data, but open source can help

Alongside all the possible permutations around running data, many are still concerned about lock in. Using a cloud service or Database as a Service option can open up new opportunities and speed up how quickly developers can get new applications up and running, but there is still hesitancy around committing to those options without a suitable route to migrate.

For those developers who are concerned about this, new open source options that support running your own Database as a Service platform will help with this. The biggest challenge will be for developers to understand how fully open source options stack up against those that advertise themselves as ‘open source compatible' services. In reality, these ‘compatible' options are not fully open source. Getting through all the marketing will be essential.

These open source options will have to meet the same level of developer experience that other services offer. Without this, developers will choose the options that make them the most productive versions of themselves, rather than considering lock in too closely. In 2022, expect more options on the market that match commercial or proprietary software for developer experience while still being fully open source.

5. Skills around data will be at a premium

As I mentioned before, developers don't care about databases right up until something breaks, and has to be fixed quickly. At this point, skills around database management and understanding how things really work become incredibly valuable. The challenge for companies is whether they try to keep those skills internally, or rely on third party providers instead.

For those that are more risk-averse, holding on to DBAs just in case is an option. However, most DBAs already see the writing on the wall for their roles if they stay focused solely on databases. They are either reskilling around the cloud alongside their DBA knowledge, or looking at joining companies that are focused solely on databases as their work. This makes the market value for those roles difficult to gauge - DBA skills on their own are going down in value as companies reduce their internal teams, but those with data skills in general alongside cloud are more valuable than ever.

The other side impact of this is that security for databases continues to be a problem. Default configurations for many of the most popular databases continue to be insecure. Developers can easily overlook those changes, leading to lost data or ransomware attacks. These will continue in 2022.

Overall, data will continue to be what companies use to differentiate their offerings. However, this approach will need support to work over time. Across cloud deployments and new services, developers will need help in managing and maintaining their growth over time. Over the next year, companies will find some of the issues that come up as they scale around those data strategies. Getting access to the right support, skills and open source projects will help overcome those hurdles.



Donnie Berkholz 

Donnie Berkholz is Senior Vice President Product at Percona and brings nearly twenty years of experience in open source development and product strategy to the company. His previous roles include leading product development strategies as Vice President, Products at container software company Docker, driving the DevOps transformation and creating platform products at travel-tech leader CWT, and roles at developer analyst firm RedMonk and 451 Research, where he advised companies on product strategy and development in the software developer, data scientist and DevOps spaces.

Published Thursday, January 20, 2022 7:33 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<January 2022>