Migration guide

From a legacy APM vendor

Teams on a per-host, per-custom-metric, per-sample priced enterprise SaaS APM.

Estimated time: 4–6 weeks for a 50-service shop

Why teams switch

  • Bill grew faster than the infra it monitors
  • Cardinality penalty pushed teams to hide useful labels
  • Data lives off-prem with no straightforward export path
  • New product surfaces (RUM, profiling) priced as separate add-ons

What transfers as-is

  • OpenTelemetry instrumentation — Unimoni speaks OTLP/HTTP natively
  • Distributed traces, metrics, and logs in one ingest endpoint
  • Existing dashboards via screenshot + manual rebuild (most teams clean up at this stage anyway)

What to adapt

  • Vendor-specific APM agents — replace with the language-specific Unimoni auto-instrumentation pack
  • Runbooks referencing the old UI — replaceable via the AI-assist chat in the cabinet

Step by step

  • 1.Add the Unimoni auto-instrumentation pack alongside the existing APM agent
  • 2.Compare metrics on one service for two weeks
  • 3.Rebuild the dashboards that actually drive decisions
  • 4.Move on-call alerts; keep the legacy agent for 30 days as a safety net
  • 5.Cancel the legacy contract on the next renewal cycle