BigBasket, a TATA Enterprise, migrates to groundcover to unify, deepen, and extend visibility
BigBasket consolidates logs, metrics, and traces, under a single observability platform, with a groundcover’s unique cost structure, which enables the expansion of observability across additional environments and teams.
%201.webp){kind=small}
.jpg)
"We cut our costs in half and actually increased our observability coverage in testing environments. We now have full coverage where we previously had to limit it due to cost concerns."
About BigBasket
bigbasket.com is India’s largest online food and grocery store. With over 20,000 SKUs and over 1,000 brands in their catalog.
As an e-commerce leader in a high-demand market, scalability, real-time performance, and system reliability are mission-critical to its operations. To handle the growing demand, BigBasket’s core infrastructure is built on AWS and Kubernetes, allowing its microservices-based platform to scale dynamically.
The Problem: Observability challenges at scale
Despite using multiple observability tools, BigBasket faced major challenges in maintaining full-stack visibility because of fragmented observability across environments. BigBasket previously relied on a mix of open source tools and a commercial license arrangement with a major legacy APM provider with large market presence.
Notably, the team did not have tracing. Their existing tool stack did not provide an adequate solution in this field, nor did they find one that did.
The use of multiple solutions meant their engineering teams had to constantly jump between tools. In addition, the incumbent APM provider’s license-based pricing model forced them to limit the number of users with access to application metrics and limit the metrics ingestion to keep costs low. This all made observability a growing challenge which ultimately boiled down to:
- Lack of correlation & visibility gaps: With multiple observability tools, correlating issues was an even bigger challenge than it already is. In addition, the team had to limit the volume of metrics sent to the incumbent APM provider in order to avoid escalating costs. This created blind spots and made it challenging for the team to get to the bottom of issues.
- Unpredictable costs & scaling limitations: Data volume-based pricing of incumbent solutions created a need to constantly monitor costs to avoid huge spikes. As part of the remediation, BigBasket had to limit its coverage (across systems and people), often sacrificing observability depth and leaving certain workloads under-monitored. Cost was also the reason for being unable to leverage the native distributed tracing capability provided by the incumbent observability partners, only due to unviable costs.
Why groundcover?
As BigBasket’s infrastructure grew in complexity, their existing observability stack forced the team to make trade-offs between visibility and financial viability. They needed a single, scalable solution that could unify logs, metrics, and traces across all workloads, while keeping costs predictable and performance high.
BigBasket chose groundcover for the granularity achieved with eBPF’s kernel-level visibility and the platform being Kubernetes-native. groundcover’s ability to deliver full-stack observability without requiring complex instrumentation was also a huge advantage. This allowed teams to see every request, capture every log, and debug faster, all in a single platform.
- Unified observability & deeper granularity: Not only did they get a unified platform for all their observability data, but they were able to see traces’ payloads, log attributes were added automatically, the network map offered visibility into external services, and much more.
- Everything out-of-the-box with eBPF auto-instrumentation: groundcover’s sensor is powered by eBPF, which not only offers kernel-level visibility, it does so with zero instrumentation. This was a huge time-saver for the company.
- Predictable & cost-effective pricing model: groundcover’s “bring your own cloud” (BYOC) model allowed BigBasket to scale observability freely without worrying about per-GB or per-trace fees. This flexibility enabled wider logging coverage, allowing teams to enable debug logs when necessary, without breaking the budget.
“With groundcover, we have a single source of truth for end-to-end observability, across our modern Kubernetes workloads as well as other platforms. Observability is no longer a bottleneck - it’s a competitive advantage.”
- Sushant Gulati, Senior Engineering Manager, BigBasket
The Impact
By implementing groundcover, BigBasket transformed its observability approach, moving from a fragmented, high-cost monitoring stack to a unified, cost-effective, and scalable solution. Previously, diagnosing incidents meant switching between multiple tools, dealing with missing logs and metrics, and managing unpredictable costs due to volume-based pricing models. Now, with real-time tracing, full log and metrics coverage, and optimized query performance, teams across DevOps, support, and engineering can collaborate seamlessly to detect, investigate, and resolve issues faster than ever before. groundcover also offers the ability to correlate critical metrics, logs and traces, to investigate the anatomy of an incident.
The result is a more resilient, efficient, and cost-stable infrastructure that supports BigBasket’s rapidly growing customer base. Whether handling high-traffic promotional events, scaling microservices, or monitoring legacy workloads, groundcover has enabled BigBasket to achieve:
- Faster incident resolution with full-stack visibility: With groundcover’s unified logs, traces, and real-time metrics, BigBasket reduced S1 incident resolution time by quickly identifying root causes. No more jumping between tools - teams could now analyze service interactions, database queries, and infrastructure performance in a single platform.
- Stable observability costs and increased coverage: With self-hosted storage instead of volume-based pricing, BigBasket expanded log coverage without cost overruns. High volume logs, previously disabled due to ingestion costs, could now be enabled on-demand - providing deeper insights during critical events.
- Full team adoption driven by the removal of per-user prices: Prior to groundcover, observability was primarily restricted to DevOps engineers due to per-user pricing. Now, with a flat subscription cost that isn’t affected by the number of users with full or limited access to the platform, support teams, QA engineers, and developers use groundcover daily.
- Proactive and responsive support - BigBasket feels that the groundcover team takes feedback seriously and delivers features and fixes very quickly, often before issues arise. A partnership was created that was built on recurring product discussions, which helped continuously deliver on their expectations from an observability solution. The groundcover team has also delivered a lot of improvements and new features very quickly based on the client feedback, allowing the company to scale further.
The Future
While eBPF provided immediate observability across all workloads, BigBasket has adopted OpenTelemetry for their microservices workloads. With groundcover’s native OTel support, BigBasket can seamlessly merge eBPF insights with developer-instrumented traces, ensuring long-term flexibility in its observability strategy.
As BigBasket continues to expand, ensuring seamless monitoring across multiple AWS regions is a priority. groundcover’s scalability ensures that new workloads - whether in additional Kubernetes clusters or new EC2 instances - will automatically inherit observability coverage, minimizing operational overhead.
By consolidating its observability stack, optimizing costs, and enabling real-time insights across all services, BigBasket has ensured that millions of grocery deliveries run reliably - empowering the future of online grocery shopping in India.
Sign up for Updates
Keep up with all things cloud-native observability.
