Skip to content
Calcrivo

Log Ingestion Calculator

Calculate daily log ingestion volume from the number of log sources, events per second per source, and average event size.

Inputs

sources

Number of distinct hosts/services/pods emitting logs.

events/sec

Average log events emitted per second by a single source.

bytes

Average size of one log event/line, including metadata.

Ingestion Rate

43.20GB/day

Ingestion Rate

0.042TB/day

Projected Monthly Volume

1,296.0GB/month

Total Events per Day

86,400,000events

Step by step

  1. Total events/sec: sources × events/sec per source

    200 × 5

    = 1,000 events/sec

  2. Bytes/sec: total events/sec × avg bytes

    1,000 × 500

    = 500,000 bytes/sec

  3. GB/day: bytes/sec × 86400 ÷ 1e9

    500,000 × 86400 ÷ 1e9

    = 43.20 GB/day

How it works

Log ingestion volume scales with the number of sources, how frequently each source emits events, and the size of each event: rate_GB_per_day = sources × events_per_sec × avg_event_bytes × 86400 ÷ 1e9. This is the foundational sizing input for every downstream stage of a logging pipeline — buffers, shippers, indices and long-term storage should all be sized against this rate.

Formula

rate_GB_per_day = sources × events_per_sec × avg_event_bytes × 86400 / 1e9

sources
Number of log-emitting hosts/services
events_per_sec
Average events per second per source
avg_event_bytes
Average log event size (bytes)

Frequently Asked Questions

How do I measure events per second per source accurately?

Sample actual log output over a representative window (including peak traffic periods, not just average) using `wc -l` on rotated logs divided by the time window, or query your existing log shipper's own ingestion metrics.

Does this rate stay constant throughout the day?

Rarely — most systems see peak ingestion during business hours or traffic spikes; size buffers and shippers against peak rate, not average rate, to avoid backpressure during busy periods.

How does average event size vary by log type?

Plain text access logs are often 100-300 bytes; structured JSON application logs with stack traces or context fields commonly run 500 bytes to several KB per event.

What's a good next step after estimating ingestion rate?

Feed this rate into buffer sizing (Fluentd/Fluent Bit), index storage sizing (OpenSearch), and observability cost calculators to get a full picture of pipeline capacity and spend.

You might also need