Conversation
✅ Deploy Preview for redpanda-docs-preview ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
Co-authored-by: Gellért Peresztegi-Nagy <gellert.nagy@redpanda.com>
|
|
||
| To identify which clients are actively connected and generating traffic, see <<view-connected-client-details>>. | ||
|
|
||
| // TODO: Verify the exact quota_rule label values for user-based quotas in v26.1. |
There was a problem hiding this comment.
@pgellert Would you mind confirming that the label values are as described here?
There was a problem hiding this comment.
Yeah, good point, they do change. These are the set of values that are possible (a bunch of new rule values added, and the old cluster_ ones don't exist anymore): https://github.com/redpanda-data/redpanda/blob/a8aa17c80aaa60d6c4ac55e43c50eaf7da43372c/src/v/kafka/server/client_quota_translator.cc#L111-L133
There was a problem hiding this comment.
@pgellert according to our metrics docs generation tool + Claude Code, the label name also changed (redpanda_quota_rule, alongside redpanda_quota_type), I went ahead and added that in this doc but if you could please also double check, that would be helpful.
There was a problem hiding this comment.
Thanks, the label names are indeed redpanda_quota_type and redpanda_quota_type. They have always been that, though, so that was a typo from earlier, rather than a change in 26.1.
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
| Use user-based quotas when: | ||
|
|
||
| * You operate a multi-tenant environment, such as SaaS platforms or enterprises with departments. | ||
| * You require isolation between users or tenants, including for compliance requirements. |
There was a problem hiding this comment.
"* You require isolation between users or tenants, to avoid noisy neighbor issues."
There was a problem hiding this comment.
@mattschumpert Updated this line, and also updated the intro to this doc: https://deploy-preview-1596--redpanda-docs-preview.netlify.app/current/manage/cluster-maintenance/about-throughput-quotas/
Redpanda uses throughput quotas to limit the rate of produce and consume requests from clients. Understanding how quotas work helps you prevent individual clients from disproportionately consuming resources and causing performance degradation for other clients (also known as the "noisy-neighbor" problem), and ensure fair resource sharing across users and applications.
How does this sound?
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
modules/manage/pages/cluster-maintenance/manage-throughput.adoc
Outdated
Show resolved
Hide resolved
|
|
||
| == Throughput control overview | ||
|
|
||
| Redpanda provides two types of throughput limits: broker-wide limits configured using cluster properties and client quotas configured using Kafka API. |
There was a problem hiding this comment.
Maybe format these two as definition lists? Easier to consume quickly.
|
|
||
| Redpanda provides two types of throughput limits: broker-wide limits configured using cluster properties and client quotas configured using Kafka API. | ||
|
|
||
| This page covers client quotas, which enable per-user and per-client rate limiting with fine-grained control through entity hierarchy and precedence rules. For information about broker-wide limits, see xref:manage:cluster-maintenance/manage-throughput.adoc#broker-wide-throughput-limits[Broker-wide throughput limits]. |
There was a problem hiding this comment.
This confused me: "This page"...we don't typically use that to describe the content, do we? I see that client quotas is a sub-topic within Quota entities, but there are also other topics covered here too, like user-based quotas.
Is there an issue here with heading hierarchy?
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
| * User-only quotas share across client_ids: When you configure a quota at the user level (for example, `/config/users/alice`) without specifying a `client_id` dimension, all connections from that user share a single quota bucket regardless of their `client_id`. | ||
| * Client-id-only quotas share across users: When you configure a quota at the `client_id` level (for example, `/config/clients/app-1`) without specifying a user dimension, all users connecting with that `client_id` share a single quota bucket. | ||
| * Combined quotas for per-(user, client_id) tracking: To get independent quota tracking per (user, client_id) combination, configure quotas that include both dimensions, such as `/config/users/<user>/clients/<client-id>` or `/config/users/<default>/clients/<default>`. | ||
| * Prefix-based quotas group clients: Client ID prefix quotas (for example, `/config/client-id-prefix/app-`) cause all clients whose `client_id` starts with that prefix to share a quota bucket. |
There was a problem hiding this comment.
Wondering if each of these bullets applies to a specific level shown above. If so, you could use numbers to help readers quickly make the match/connection. I could be all wrong here, but if not, it sure would help me to quickly make the connection between the bullets here and the various levels above.
|
|
||
| Then `alice` connecting with `client_id=app-1` and `alice` connecting with `client_id=app-2` share the same 10 MB/s limit. Their combined throughput cannot exceed 10 MB/s. | ||
|
|
||
| To give each of `alice`'s clients an independent 10 MB/s limit, configure: |
There was a problem hiding this comment.
| To give each of `alice`'s clients an independent 10 MB/s limit, configure: | |
| To give each of `alice`'s` clients an independent 10 MB/s limit, configure: |
There was a problem hiding this comment.
Was missing the closing delimiter, but is this what you intended?
modules/manage/pages/cluster-maintenance/about-throughput-quotas.adoc
Outdated
Show resolved
Hide resolved
|
|
||
| This quota applies to all users who don't have a more specific quota configured. Each user is tracked independently: `alice` gets their own 10 MB/s bucket, `bob` gets their own 10 MB/s bucket, and so on. | ||
|
|
||
| Within each user, all `client_id` values share that user's bucket. `alice` connecting with `client_id=app-1` and `alice` connecting with `client_id=app-2` share the same 10 MB/s limit, while `bob`'s connections have a separate 10 MB/s limit. |
There was a problem hiding this comment.
| Within each user, all `client_id` values share that user's bucket. `alice` connecting with `client_id=app-1` and `alice` connecting with `client_id=app-2` share the same 10 MB/s limit, while `bob`'s connections have a separate 10 MB/s limit. | |
| Within each user, all `client_id` values share that user's bucket. `alice` connecting with `client_id=app-1` and `alice` connecting with `client_id=app-2` share the same 10 MB/s limit, while `bob`'s` connections have a separate 10 MB/s limit. |
There was a problem hiding this comment.
Again, not sure if this is what you intended--just adding a closing delimiter.
|
|
||
| | `consumer_byte_rate` | ||
| | Limit throughput of fetch requests | ||
| This limits user `alice` to 2 MB/s for produce requests, regardless of which `client_id` they use. |
There was a problem hiding this comment.
singular user, but plural pronoun. I see you are consistent with this, but not sure why.
| rpk cluster quotas alter --add consumer_byte_rate=1000000 --name user=alice --name client-id=consumer-1 | ||
| ---- | ||
|
|
||
| User `alice` using `client_id="consumer-1"` is limited to a 1 MB/s fetch rate. The same user with a different `client_id` would use a different quota (or fall back to less specific matches). |
There was a problem hiding this comment.
What is the significance of the double quotation marks here? Seems superfluous.
| rpk cluster quotas alter --add producer_byte_rate=3000000 --name user=bob --name client-id-prefix=app1- | ||
| ---- | ||
|
|
||
| All clients used by user `bob` with `client_id` starting with "app1-" share a combined 3 MB/s quota. |
There was a problem hiding this comment.
Sometimes you use the delimiters, sometime the double quotation marks, sometimes both. Not going to mark every instance. Just be consistent, whatever you decide.
|
|
||
| To identify which clients are actively connected and generating traffic, see <<view-connected-client-details>>. | ||
|
|
||
| Quota metrics use the `redpanda_quota_rule` label to identify which quota was applied to a request. The label distinguishes between different entity types (user, client, or combinations). See the label values in xref:reference:public-metrics-reference.adoc#redpanda-kafka-quotas-client-quota-throughput[]. |
There was a problem hiding this comment.
link is not rendering
CC knows why!
The anchor exists but uses underscores, not hyphens. The xref uses redpanda-kafka-quotas-client-quota-throughput but the actual section heading is redpanda_kafka_quotas_client_quota_throughput.
Antora auto-generates anchors from section headings by replacing spaces with hyphens — but here the heading itself contains underscores (it's a metric name), so the anchor is also underscores.
The fix is to change the xref to match:
xref:reference:public-metrics-reference.adoc#redpanda_kafka_quotas_client_quota_throughput[]
Feediver1
left a comment
Feediver1
left a comment
There was a problem hiding this comment.
Left several comments. Please fix all format/tagging errors.
Co-authored-by: Joyce Fee <102751339+Feediver1@users.noreply.github.com>
4 participants
Description
This pull request introduces significant improvements to the documentation for configuring and understanding throughput quotas in Redpanda, with a focus on clarifying the distinction between user-based and client-based quotas, explaining quota precedence and tracking, and providing detailed, actionable configuration instructions.
A new conceptual guide has been added, and the practical guide has been expanded and reorganized for clarity and completeness.
Key documentation improvements:
Conceptual clarity and new guide
about-throughput-quotas.adocthat explains the differences between user-based and client-based quotas, entity hierarchy, precedence rules, and quota tracking behavior. This guide helps users choose the right quota type for their use case and understand how Redpanda enforces quotas.nav.adoc) to include the new conceptual guide for easy discovery.Enhanced practical configuration instructions
manage-throughput.adocwith clear learning objectives, detailed descriptions of both user-based and client-based quotas, and step-by-step instructions for configuring, viewing, and deleting quotas usingrpk. The guide now covers combined (user, client) quotas, default quotas, and best practices for multi-tenant environments. [1] [2]Improved quota monitoring and enforcement explanation
user=""vs.user=<default>), and how throughput is tracked independently for each (user, client_id) pair. [1] [2]User experience and structure
References: [1] [2] [3] [4] [5] [6] [7]
Resolves https://redpandadata.atlassian.net/browse/
Review deadline: 9 Mar
Page previews
Manage Throughput
NEW: About Throughput Quotas
Checks