Skip to content

Add NVMe and 16000 IOPS to requirements #1343

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Aug 27, 2025
Merged

Add NVMe and 16000 IOPS to requirements #1343

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
de1eb5e
Add NVMe and 16000 IOPS to requirements
JakeSCahill Aug 26, 2025
80ab4dd
Apply suggestions from code review
JakeSCahill Aug 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions modules/deploy/partials/requirements.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -134,6 +134,12 @@ endif::[]

*Requirements*:

- NVMe (Non-Volatile Memory Express) drives are required for production deployments. NVMe drives provide the high throughput and low latency needed for optimal Redpanda performance.

- At least 16,000 IOPS (Input/Output Operations Per Second).
+
See also: xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].

Comment on lines +139 to +142
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Verification agent

🧩 Analysis chain

Make the IOPS threshold explicitly “per broker node” and align phrasing with self-test.

Calling out the unit of enforcement (“per broker node”) prevents misconfiguration when users size storage for multi-broker clusters. Also, reference the self-test here to drive a single source of truth for how to validate.

-- At least 16,000 IOPS (Input/Output Operations Per Second).
-+
-See also: xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].
+- At least 16,000 IOPS (Input/Output Operations Per Second) per broker node.
++
+Validate your storage with the xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].

Verification script to catch lingering inconsistencies elsewhere (for example, pages that still call NVMe “recommended” or omit the 16,000 IOPS figure):

🏁 Script executed:

#!/bin/bash
# Find potentially outdated NVMe phrasing
rg -n -C2 -iP 'nvme.*recommend|recommend.*nvme|nvme recommended'

# Find IOPS thresholds and confirm consistency with 16,000
rg -n -C2 -iP '\b(?:16,?000)\b.*iops|iops.*\b(?:16,?000)\b'

# Scan for any allowance of NFS on data or Tiered Storage cache
rg -n -C2 -iP 'nfs.*(supported|support)'

Length of output: 5152

Clarify IOPS and NVMe requirements across documentation

The storage requirements should explicitly reference “per broker node” for IOPS and align all NVMe phrasing with the production requirement. Please update the following:

• In modules/deploy/partials/requirements.adoc (lines 139–142):

- At least 16,000 IOPS (Input/Output Operations Per Second).
- See also: xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].
+ At least 16,000 IOPS (Input/Output Operations Per Second) per broker node.
+
+ Validate your storage with the xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].

• In Kubernetes guides, change “Local NVMe disks, which is recommended for best performance” to reflect that NVMe is required for production:
– modules/deploy/pages/redpanda/kubernetes/gke-guide.adoc (line 37)
– modules/deploy/pages/redpanda/kubernetes/aks-guide.adoc (line 38)
– modules/deploy/pages/redpanda/kubernetes/eks-guide.adoc (line 254)

These edits ensure a single source of truth for storage sizing and drive home the requirement for NVMe and per-node IOPS thresholds.

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change
- At least 16,000 IOPS (Input/Output Operations Per Second).
+
See also: xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].
At least 16,000 IOPS (Input/Output Operations Per Second) per broker node.
Validate your storage with the xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].
🤖 Prompt for AI Agents 
In modules/deploy/partials/requirements.adoc around lines 139–142, clarify that
the IOPS requirement is "at least 16,000 IOPS per broker node" and update the
sentence to reference per-node IOPS and link to the existing cluster diagnostics
self-test; additionally, in the Kubernetes guides update the phrasing to state
that Local NVMe disks are "required for production" (not just recommended) in
these files and lines: modules/deploy/pages/redpanda/kubernetes/gke-guide.adoc
(line 37), modules/deploy/pages/redpanda/kubernetes/aks-guide.adoc (line 38),
and modules/deploy/pages/redpanda/kubernetes/eks-guide.adoc (line 254), ensuring
wording is consistent across docs and aligns with production requirements.

- An XFS or ext4 file system.
+
The Redpanda data directory (`/var/lib/redpanda/data`) and the Tiered Storage cache must be mounted on an XFS or ext4 file system.
Expand Down
4 changes: 3 additions & 1 deletion modules/deploy/partials/self-test.adoc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,4 +2,6 @@

To understand the performance capabilities of your Redpanda cluster, Redpanda offers built-in self-test features that evaluate the performance of both disk and network operations.

For more information, see xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].
For more information, see xref:manage:cluster-maintenance/cluster-diagnostics.adoc#self-test[Disk and network self-test benchmarks].

When using the storage bandwidth test, ensure that your results show at least 16,000 IOPS (Input/Output Operations Per Second) for production environments. If your test results are below this threshold, your storage may not be suitable for production Redpanda workloads.