🔥

Latency Percentile Explorer

This runs the framework's percentile logic (ported from lib/latency_stats.py): p50/p95/p99 over a sample of ingestion latencies, plus SLA status. Adjust the tail and SLA and regenerate.

1. Pipeline behaviour

Typical latency (s) 30

Tail severity 1.0

SLA (p95, s) 90

2. Latency distribution (200 batches)

–

p50 (s)

–

p95 (s)

–

p99 (s)

–

max (s)

–

Get the full accelerator — $990

Included with Ingestion Latency Framework

The complete Ingestion Latency Framework package includes:

Databricks Asset Bundle
Two latency measures
Percentile summaries
A dependency-free stats library

Ready for the complete Ingestion Latency Framework?

Get the full product, documentation, examples, and future updates.

Get the full product →