Snowplow Micro Rate Limits and Event Processing Capacity

Assumptions: internal-only article (source sits in Internal → Husky Drafts); content migrated as-is, no fact-check against docs.snowplow.io; owner set to Juana Hernández, updated August 2026. Zendesk breadcrumb, TOC and footer nav omitted — Pylon renders those itself.

Article title: Snowplow Micro Rate Limits and Event Processing Capacity Type: Reference / orientation

1 — Internal note · /internal

Internal — Support Team Only. Migrated from Zendesk (Internal → Husky Drafts). Owner: Juana Hernández · Updated: August 2026

2 — Header 2 · /h2

Rate limit

3 — Text

Hosted Snowplow Micro has a maximum rate limit of 1500 requests per minute (approximately 25 requests per second on average). This limit is configured at the ingress level.

4 — Table · /table · header row on · 2 columns

5 — Callout · /callout · icon 🔑 · color yellow

**Key rule:** The rate limit applies to requests per minute, not total events. If your backend sends events in batches, each request counts as one against the limit, regardless of how many events it contains.

6 — Header 2 · /h2

When the limit is exceeded

7 — Text

When traffic exceeds the 1500 requests per minute threshold, Micro rejects events that exceed the limit. This is a graceful rejection at the ingress layer, meaning:

8 — Bulleted list · /bullet

- Micro does not crash or stop functioning
- Events within the rate limit are processed normally
- Events exceeding the limit are rejected and will not be stored

9 — Header 2 · /h2

Checking your traffic against the limit

10 — Text

To calculate whether your traffic will fit within Micro's capacity, divide your total events by the time period.

11 — Callout · /callout · icon ℹ · color blue

**Example:** 1.8 million events over 7 days = approximately 257,000 events per day, or ~179 events per minute on average. This is well within the 1500 req/min limit.

12 — Header 2 · /h2

Bursty traffic patterns

13 — Text

If your application sends traffic in bursts rather than evenly throughout the day, you could exceed the 1500 req/min limit during peak periods even if your daily average is well within limits. Monitor your traffic patterns to understand whether you have peak sending hours that could spike above the threshold.

14 — Header 2 · /h2

When to use Micro

15 — Text

Micro is designed for:

16 — Bulleted list · /bullet

- Development and testing environments
- Schema validation and tracker testing
- CI/CD test suites
- Debugging and inspection

17 — Callout · /callout · icon ⚠ · color yellow

**Warning:** Micro retains events for a rolling 7-day period only. It is not intended for production-level traffic or long-term data retention. If a customer needs production-grade event processing, point them to Snowplow's self-hosted pipeline.

18 — Header 2 · /h2

Performance notes

19 — Text

Testing has shown that Micro can handle approximately 2 million events (within the rate limits) without falling over. However, certain UI filters may become slower when filtering by multiple columns simultaneously, especially with large event volumes near the 7-day retention window.

Notes for the author

• If a "Snowplow Micro overview" or self-hosted pipeline article exists in Pylon, link it from block 17 using the + relative-link picker rather than pasting a URL.

• Block 4's table is new — it condenses figures already stated in the prose. Delete it if you'd rather keep the article strictly 1:1 with the Zendesk source.