How OpenObserve Replaced Overprovisioned Clusters with Full Cost Visibility Across AWS and Azure
in pre-production infrastructure spend
Increased utilization from 24% through intelligent infrastructure optimization
Full cost visibility and unified observability across all clusters (AWS EKS and Azure AKS)
Reduced operational overhead by consolidating 7 environments into one
About OpenObserve
OpenObserve (O2) is an open-source observability platform for logs, metrics, traces, and frontend monitoring — positioned as a cost-effective alternative to Datadog, Splunk, and Elasticsearch, with claimed 140x lower storage costs and a single binary deployment. It operates in the Software Development / DevOps space, specifically cloud-native observability. It has over 6000+ active deployments.
The Challenge: Cost Visibility and Manual Overhead at Scale
As OpenObserve's infrastructure grew to span seven Kubernetes clusters across two cloud providers, the overhead of manual management began to create compounding problems for the engineering team.
Unpredictable Kubernetes Environment Costs
OpenObserve's Kubernetes clusters ran on a static setup: fixed node sizes, fixed resource requests, and little dynamic adjustment to actual demand. With average cluster utilization sitting at just 24%, they were overprovisioning for peak capacity and paying for it regardless of usage, while test environment bills regularly exceeded production spend. Without visibility into the cause, waste was hard to catch: engineers would spin up resources, forget to scale back down, and those costs would quietly accumulate.
Manual Processes That Didn't Scale
The team's existing approach to cost optimization relied on custom scripts and AI-assisted analysis to determine resource allocations, but the process was slow and required significant engineering attention. When workloads shifted, engineers had to manually open cloud consoles, assess utilization, and update capacity. There were no autoscaling feedback loops, no optimization tooling, and no reliable signal linking usage to actual cost, so engineering time was consistently diverted from product development.
Fragmented Pre-Production Environments
OpenObserve was running dedicated clusters for each lower environment type, including dev, test, and pentest environments. Managing multiple isolated clusters for pre-production workloads added operational complexity without delivering proportional value, and it made cost attribution and control even harder.
The Solution: Automated Kubernetes Optimization Across Clouds
After identifying the need for a smarter approach to Kubernetes resource management, OpenObserve evaluated DevZero as a solution to bring automation and cost visibility across its multi-cloud infrastructure. Onboarding was fast, and the team had clusters connected and delivered insights within days.
Fast Implementation, Immediate Clarity
The team started by connecting two lower-environment clusters to DevZero, then rolled out across all seven clusters once they confirmed the integration was working as expected. The entire process was straightforward, and the DevZero team was available for support within minutes when questions came up.
“The DevZero team made the onboarding experience easy and fast. We started with two clusters in our test environments and then added everything else. Within two weeks, we had a solid picture of our infrastructure costs.”
Automated Vertical Scaling and Node Pool Management
DevZero took over the resource management work that had previously required manual intervention. The platform automatically handles vertical upscaling of workloads, managing node pools and adjusting resource allocations based on actual demand, without engineering involvement.
“Before DevZero, we had to manually manage nodes based on workload changes: update resource requests, provision additional nodes, and handle all of it by hand. Now DevZero manages all of that automatically. It saves a significant amount of time and energy that we can put toward other work.”
Unified Non-Production Environment Management
One of the more significant structural changes DevZero enabled was the consolidation of OpenObserve's non-production environments. What had previously been separate clusters for dev, test, pentest, and other pre-production environments is now a single cluster managed by DevZero, with workloads organized by node pools and resource boundaries maintained automatically.
A Single Pane of Glass Across AWS and Azure
Prior to DevZero, assessing the state of OpenObserve's infrastructure required jumping between AWS, Azure, and Lens. DevZero replaced that fragmented workflow with a unified dashboard that surfaces cost and utilization insights across all clusters in one place.
“I used to have to open AWS, Azure, and Lens just to understand what was going on with our infrastructure. Now I go directly to the DevZero platform and have everything I need in one view. It's made a real difference in how I work day to day.”
The Results: Infrastructure Managed, Engineering Unblocked
A Clearer Picture of Infrastructure Spend
With full cost visibility across all seven clusters, the OpenObserve team can now identify and address waste proactively rather than discovering it after the fact. What had previously required manual investigation and scripted analysis is now surfaced automatically, giving the team a reliable foundation for infrastructure decisions and confident forecasting going forward.
Cost Reduction in Pre-Production Infrastructure Spend
By consolidating previously fragmented dev, test, and pentest clusters into a single DevZero-managed cluster, OpenObserve cut its pre-production infrastructure spend by 42% month over month. Beyond the cost impact, the consolidation reduced the operational overhead of managing multiple isolated environments and created a cleaner foundation for cost attribution across the team.
Engineering Time Returned to Product Work
With DevZero managing node provisioning, vertical scaling, and resource allocation automatically, the OpenObserve engineering team no longer spends time on infrastructure firefighting. That capacity is now directed toward product development and issue resolution.
“With DevZero, the team is now focused on product development instead of troubleshooting infrastructure problems caused by resource constraints. The hours we were losing to infrastructure management are now going into shipping products.”
Confidence in Infrastructure Going Forward
With DevZero managing their Kubernetes infrastructure, OpenObserve now has the tooling and visibility to scale confidently. Average cluster utilization has climbed from 24% to 76%; resource management, once requiring constant manual oversight, now runs automatically; and the team has a clear view of their infrastructure costs at any given time.