Configuring global rate limiting

レート制限でインスタンスの安定性を改善する

このページの内容

お困りですか?

アトラシアン コミュニティをご利用ください。

コミュニティに質問

Global rate limiting lets you control the volume of requests reaching your instance by classifying traffic into two categories and applying separate rate limits to each. This protects instance stability during periods of heavy traffic.

Global rate limiting is available in Confluence Data Center 10.2.13 and later. To check your version, go to Administration menu > General Configuration > System Information.

How global rate limiting classifies requests

Global rate limiting classifies all incoming requests into two categories: 
 

リクエスト タイプ
 

説明
 

Automation Request

Traffic from automation scripts or bots, including integrations, scheduled jobs, and background API requests.

Application Request

Traffic generated by users through browser or mobile interactions, including page navigation, form submissions, and UI-driven actions.


Automation and Application requests are configured separately, each with its own settings UI. The definitions of Requests allowed and Max requests remain the same as in standard rate limiting — see Improving instance stability with rate limiting for details.

You can exempt specific users or service accounts from rate limiting in both sections. Exemptions override global rate limiting settings.

Service accounts can't log in to Confluence directly, so they don't appear in the Application request exemption list. Add service account exemptions in the Automation request section instead.

How to turn on global rate limiting

To turn on global rate limiting:

  1. Confluence を停止します。

  2. Add the com.atlassian.ratelimiting.global-rate-limiter.enabled system property and set it to true:

    -Dcom.atlassian.ratelimiting.global-rate-limiter.enabled=true


    The way you add system properties depends on how you run Confluence. See Configuring System Properties for more information.


  3. Confluence を再起動します。

What happens when users are rate limited

When users hit their rate limit, the behaviour depends on the request type:

  1. Automation requests: The request fails with an HTTP 429 error (too many requests).

  2. Application requests: The user sees a Rate Limiter Error notification in Confluence.

To see which users and service accounts are currently rate limited, go to the rate limiting settings page and open the List of limited accounts tab. This list covers all users and service accounts across the entire cluster.


最終更新日 2026 年 6 月 2 日

この内容はお役に立ちましたか?

はい
いいえ
この記事についてのフィードバックを送信する
Powered by Confluence and Scroll Viewport.