Granulate, an Intel Company, announced the upcoming launch of its latest free cost-reduction solution, gMaestro, a continuous workload and pod rightsizing tool for Kubernetes cost optimization.
We're all familiar with the process of QA within the software development cycle. Developers build a product and send it to QA engineers, who test and bless it before pushing it into the world. After release, a different team of SREs with their own toolset then monitor for issues and bugs. Now, a new level of customer expectations for speed and reliability have pushed businesses further toward delivering rapid product iterations and innovations to keep up with customer demands. This leaves little time to run the traditional development process. QA can no longer act as a major, individual step.
Fortunately, modern, automated infrastructure as code (IAC)-built toolchains that deliver continuous observability now let SRE teams watch the entire build pipeline from the first piece of code through release. This enables a whole new speed within the delivery cycle and merges the traditional roles of QA and monitoring.
Helping "the Human in the Middle"
Without a proper QA process, any change dev teams make to digital systems can have cascading effects on the infrastructure. This only further stresses the SRE team to find and fix issues after deployment. And changes are happening faster than ever. As a result, we meet the juxtaposition between the need for QA and the limited time to work the process into the development cycle as a standalone step.
Observability (mining deep data from distributed systems) delivers the data necessary to eliminate traditional QA, but that isn't enough. Humans still need help. When you apply AI to this observability data (intelligent observability), teams can analyze data at machine speed. This lets DevOps practitioners and SREs view the entire product lifecycle, from early development to daily performance, through the lens of quality.
The use of intelligent observability helps teams find the needle in the proverbial haystack of data — the root causes of issues within digital systems — instantly. It also helps identify actionable ways to quickly resolve a new product's impacts on the infrastructure. Without this capability, we revert to the old way of doing things where the dev team has QA find the needle instead. This new continuous learning and intelligent collaboration creates a merging of traditional QA and monitoring for a CI/CD pipeline that actually works.
Integrating observability with AI into the development cycle creates an opportunity to monitor expected outcomes much closer, enabling "the human in the middle" to spot change almost instantly. If the system starts behaving dramatically differently after a deployment, SREs and DevOps practitioners can see it and intervene immediately, without the need to wait for a QA team. If nothing changes or the system improves, they know there's no need to remediate a deployment.
In cases where a change in performance is expected, it was traditionally incumbent on the developer to modify the unit tests or to communicate the change to the QA team. Now, AI- and ML-based systems' change tolerance reduces IT teams' effort. For example, if you're monitoring a KPI with an adaptive thresholding algorithm, you can simply let the algorithm re-train and learn the new behavior instead of relying on the dev team to communicate the expected change in performance to QA.
Merged Systems Support DevOps Three Ways
A merged system of QA and monitoring throughout the development cycle also aligns with the DevOps Three Ways principles. We look at the First Way: flow/system thinking, the Second Way: amplifying feedback loops, and the Third Way: creating a culture of continual experimentation and learning, as the guiding principles behind DevOps practices.
The merging of traditional QA and monitoring supports the First Way — flow/system thinking — by building a holistic system view of the development process with the elimination of siloed workflows. This creates quality throughout development and delivery because the system is never optimized for local efficiency only or passed onto the next step with a known issue.
A merge also supports the Second Way — amplifying feedback loops — by giving IT teams clear, consistent feedback throughout the development and delivery process. As traditional QA and monitoring merge, the need to loop feedback through multiple teams with various processes and priorities evaporates.
This merge perhaps has the greatest impact on the Third Way. As DevOps practitioners focus on the holistic product cycle versus quick development that's passed off to QA, they can learn from bugs and build constant improvement into their process. This also gives them room to experiment and take risks. Infusing quality into the development process itself means they won't hand over garbage to the QA team — no matter how "out there" the forthcoming release might be.
Integrating observability with AI into the development cycle allows teams to not only see into systems as they're being built, but also identify actionable ways to resolve a new product's impacts on the overall infrastructure. As DevOps practitioners and SREs balance change, these actionable insights empower the merging of traditional QA and monitoring for a whole new speed of delivery — delivering better customer experiences and giving your business the ability to launch competitive, innovative services faster than ever.