As we enter 2023, many in the DevOps community are thinking along the same lines—how are we going to do more with less for the foreseeable future?
After substantial reductions in staff, funding, and revenue expectations over the past several months, this thinking applies both to the people and budget to execute a digital strategy.
However, even with these constraints, the pressure to execute the digital strategy hasn’t changed. Your customers still expect top-quality digital products round-the-clock, and your competitors are there to serve them if yours don’t fit the bill.
Here are some recommendations to equip your teams to execute a digital strategy even when working with fewer resources.
1. Don’t let institutional knowledge walk out the door
Whether it’s the result of turnover or a reduction in staff, the loss of institutional knowledge from departing employees can hold back the teams responsible for building and managing your applications.
For example, our users often find that only a select few infrastructure experts are capable of orchestrating their teams’ most complex application environments, such as those that involve working with multiple IaC tools, cloud platforms, and Kubernetes assets.
As a result, those experts end up orchestrating those environments manually every time someone on the team needs them.
Research suggests this problem is widespread. One survey of 500 developers and IT decision makers found that 85% of respondents have faced a barrier to implementing a DevOps framework, with a lack of skills in employees cited as the most common hurdle.
This request-and-fulfill model creates a lot of risk. What happens to that process if one of those skilled experts leaves the organization?
This is where we recommend a blueprint model for the most complex environments. By creating a repeatable blueprint for the infrastructure and environments needed to support specific use cases—including individual stages within the CI/CD pipeline—you essentially document that infrastructure expertise for future use.
In other words, the blueprint essentially turns that knowledge into an asset that your teams can use over time, regardless of the specific individuals on the team.
Additional reading on this topic: Navigating the DevOps Skills Gap
2. Look for opportunities to automate infrastructure processes
When working with smaller teams, automation is even more valuable.
Many of our users come to us to adapt as the environments required to support their development pipelines increasingly require the orchestration of multiple IaC and Kubernetes assets.
This not only requires the technical expertise described above—it also takes a substantial amount of manual work to complete.
This is the perfect example of a process that you can automate to help free up time for your teams.
Our users automate the orchestration of complete application requirements based on the IaC scripts, Helm Charts, and native Kubernetes manifests in their Git repositories—all in a matter of a few minutes.
Once configured, those environments are then repeatable via the blueprint model, further cutting down on the redundant manual work that prevents your staff from other, more valuable work.
DevOps teams have long embraced automation to accelerate their pipelines. As application infrastructure evolves to require complete environments, automation needs to evolve accordingly.
Learn more: Join our upcoming IaC Automation Summit
3. Take the guesswork out of cloud cost governance
Last year, a KPMG survey of 1,000 senior technology leaders found that about 67% said their investments in the public cloud have not delivered a significant return. A Wall Street Journal report on the report is full of quotes about “surprises” on the cloud bill.
Many of our users find that these surprises come from simple oversights or misunderstandings about the cloud services their teams deploy. Sometimes it could be that someone forgot to terminate a cloud service before taking off for a long weekend. In others it could be that an engineer deployed a much more expensive cloud instance type or storage service than needed for the workload, with little understanding of the financial impact.
Often, the problem is not a lack of standards to prevent this kind of waste. The challenge is enforcing those standards.
Cloud cost controls go hand-in-hand with the blueprint approach that we recommend for our users. Once you’ve created a repeatable blueprint for your environment, you can set role-based permissions to prevent any unauthorized modifications to that environment.
This allows you to pre-configure the cloud services to match the needs of the workload, automate teardown after a maximum duration, and automatically tag each environment for granular cost reporting.
End users can view and deploy the environment via self service without the ability to configure anything that doesn’t adhere to your standards.
Meanwhile, policies applied across your organization can establish standards across your infrastructure administrators—including restrictions on configurations, instance types, and cloud platforms.
At a time when you’re being asked to operate at high capacity while relying on limited resources, the priority is balancing productivity and control.