Observability Migration With 60% Lower Cost and Zero Gaps Using Grafana LGTM

We moved 2,000 dashboards and every alert from Datadog to a self-managed Grafana LGTM stack in 10 weeks, with no gaps in monitoring across 150 microservices.

  • Industry E-Commerce & Retail
  • Region Europe
  • Platform Grafana LGTM Stack

60%

lower annual observability cost

0

monitoring gaps across 150 services

10 weeks

from Datadog to Grafana, fully retired

Observability Migration With 60% Lower Cost and Zero Gaps Using Grafana LGTM
Observability Migration With 60% Lower Cost and Zero Gaps Using Grafana LGTM
O V E R V I E W

2,000 dashboards and every alert had to move without a blind spot

Our client is a mid-size e-commerce company running a storefront and order platform made up of 150 microservices. Its engineering and SRE teams relied on Datadog as the single source of truth for metrics, logs, traces, and alerting. As the platform grew, renewal costs grew faster, and leadership wanted an observability stack the business controlled rather than rented.

 

Seasonal sales drive a large share of revenue, so one blind spot on a peak-traffic day could mean failed checkouts that customers found before engineers did. The client partnered with Ksolves to leave Datadog without a single gap in visibility.

C H A L L E N G E

01

Escalating Datadog Renewal Costs Spend rose with every new service and every increase in telemetry volume, becoming one of the fastest-growing lines in the infrastructure budget.

02

Lock-In Through Proprietary Instrumentation Services used Datadog-specific agents and libraries, so changing the backend meant touching code in all 150 microservices.

03

2,000 Dashboards in a Proprietary Format Rebuilding years of operational knowledge by hand would take months and bring in query errors nobody notices until an incident.

04

On-Call Alerts That Could Not Be Lost The on-call rotation depended on Datadog monitors, and one dropped or mistranslated alert would become a silent production failure.

05

No Room for Blind Spots During Peak Sales A big-bang cutover risked losing visibility at exactly the moment traffic and revenue were highest.

06

Correlated Signals at Risk Engineers debug by pivoting between metrics, logs, and traces, and a new stack that broke that link would slow every investigation.

S O L U T I O N

Run both stacks side by side, and switch off Datadog only after the new one proves itself

Ksolves, an AI-first DevOps services company, designed the migration around one rule: the new stack had to prove itself before the old one was switched off. Datadog and the Grafana LGTM stack ran in parallel on the same telemetry while dashboards and alerts were converted automatically and checked against live data. Each fix below answers the challenge with the same number.

  1. 01 Self-managed Grafana LGTM stack Grafana Mimir for metrics, Loki for logs, and Tempo for traces sit behind one Grafana instance, replacing usage-based licensing with infrastructure the client owns.
  2. 02 OpenTelemetry re-instrumentation Services emit OpenTelemetry data through a Collector that exports to both Datadog and LGTM during the parallel run, so the backend became a configuration choice instead of a code dependency.
  3. 03 Automated dashboard conversion Conversion tooling turned Datadog dashboard definitions into Grafana JSON and rewrote queries in PromQL and LogQL, with each result compared side by side against its original.
  4. 04 Alert translation and parity testing Datadog monitors became Grafana Alerting rules, and firing behavior was compared on both platforms, so no alert was retired until its replacement proved equivalent.
  5. 05 Service-by-service cutover Each service moved only after its dashboards and alerts were verified, with Datadog kept live as a safety net until the final service had migrated.
  6. 06 Cross-signal correlation Trace IDs are linked into Loki log lines and metrics connect to traces, so engineers can jump from a latency spike to the exact trace and log line in one workflow.

Architecture

stream-dfd

Technology Stack

Category Technology
Platform Grafana (dashboards and Alerting)
Metrics Store Grafana Mimir
Log Store Grafana Loki
Trace Store Grafana Tempo
Integration OpenTelemetry SDKs and Collector

D E L I V E R Y

How telemetry flows from 150 services to one Grafana view

All 150 microservices now emit metrics, logs, and traces through OpenTelemetry into Mimir, Loki, and Tempo, which Grafana queries with PromQL, LogQL, and TraceQL. Grafana Alerting evaluates the converted rules and routes notifications to the on-call rotation. The move ran in six steps: instrument, dual-ship, convert, validate parity, cut over service by service, and retire Datadog. Because telemetry is vendor-neutral, new signals, backends, or services can be added without re-instrumenting code.

R E S U L T S

From a vendor-locked bill to a stack the client owns, with no gap in visibility

Ksolves replaced a rising Datadog bill with a self-managed Grafana LGTM stack fed by OpenTelemetry. Every dashboard and alert was converted and verified before its service moved, and no service went dark. The client now owns its telemetry pipeline and can add signals, backends, or services without re-instrumenting a line of code.

  • 60% Lower Annual Observability Cost Datadog renewals rose with every new service and telemetry increase. Annual spend is now 60% lower on the self-managed Grafana LGTM stack.
  • 0 Monitoring Gaps Across 150 Services A cutover had risked blind spots during peak sales. None were recorded at any point in the migration.
  • 10 Weeks from Datadog to Grafana A migration of metrics, logs, traces, and alerts had looked like a long, high-risk programme. It was complete in 10 weeks.
  • 2,000 Dashboards Converted, Not Rebuilt Dashboards were locked in a proprietary format. All 2,000 were converted automatically and validated against live data.
  • 150 Services on Vendor-Neutral Telemetry Instrumentation was tied to Datadog agents. All 150 services now emit OpenTelemetry data, so a future backend change needs configuration, not code.

T E A M

Meet the Engineers on Your Project

Every engagement is staffed by senior engineers from our DevOps team, who will design, build, and support your platform.

Harshit Nagar, DevOps Engineer

Harshit Nagar

DevOps Engineer

Shobhit Agarwal , DevOps Engineer

Shobhit Agarwal

DevOps Engineer

Prashant Bhargava, DevOps Engineer

Prashant Bhargava

DevOps Engineer

Tired of Escalating Datadog Renewal Bills?

Learn how our DevOps team completed a zero-downtime, parallel-run migration to self-managed Grafana LGTM across 150 services.

Global Presence
Follow Us
Copyright 2026© Ksolves.com | All Rights Reserved
Ksolves USP