Cost Optimization Guide

Real-time log filtering for lower latency decisions

Real-time filtering ensures noise is controlled immediately, not after monthly cost spikes.

Why this problem exists

Batch-style cleanup is too slow for rapidly changing traffic patterns.

Delayed filtering keeps noisy data flowing into expensive systems.

Real cost and impact

Real-time suppression prevents noise bursts from inflating the same-day ingestion profile.

Operational teams also gain faster access to cleaner signals.

Solutions (including alternatives)

  • Run rule evaluation in the ingestion path with deterministic drop and mask behavior.
  • Use low-latency routing so critical events still reach Datadog in near real time.
  • Mirror full retained data to S3 for later replay and audit needs.

How LogTrim solves it

LogTrim executes filtering, masking, and routing in one real-time data plane.

Teams enforce consistent policy with low added latency.

Example scenario

During a traffic surge, a team dropped repetitive success noise in real time and maintained clean alerting.

They avoided a reactive post-incident ingestion spike.

Reduce your costs with LogTrim

Start with high-noise categories, keep high-signal logs in Datadog, and archive full retention in S3.