Note the 25,000 ops/sec clustering threshold alongside 25 GB

[mich-elle-luna]

The page stated the data-size trigger for clustering (25 GB) but left the throughput trigger without a concrete number. Add the 25,000 ops/sec threshold to the throughput bullet, paralleling the size bullet, and mention it in the page description.

---

Note

Low Risk
Documentation-only wording changes with no runtime or security impact.

Overview
Documents the throughput trigger for Redis Cloud clustering so it matches the existing size guidance (25 GB).

The page description now mentions high-throughput databases (25,000 ops/sec and higher) in addition to very large databases. The CPU/throughput example bullet adds an explicit rule: use multiple shards when throughput reaches 25,000 ops/sec, while keeping the explanation that clustering spreads operational load across instances.

^{Reviewed by Cursor Bugbot for commit 88a779c. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

DOC-6453

[github-actions]

Staging links:
https://redis.io/docs/staging/DOC-6453/operate/rc/databases/configuration/clustering/

[github-actions]

🧠 Redis Memory

Found 5 related items from repository history (5 new this commit):

• PR — DOC-6874: Document smooth scaling prerequisites on Cloud clustering page 🆕

  Adds a Smooth scaling section covering prerequisites (new DB, Redis hashing policy, DB version 8.4+, cluster version 8.0.18+, HA required),

• PR — RS: Explained replication_oom_threshold_percent in A-A memory limits docs 🆕

• PR — RS: Copied A-A clustering limitation to the planning doc 🆕

• Commit — @LiorKogan 🆕

  @LiorKogan

• PR — RS: Clarified 80% memory limit for Active-Active DB eviction is per-shard 🆕

Memory updated at 88a779c