Skip to content

Exceeded resource group quota limitation if request tokens exceeded 500ms #8349

@nolouch

Description

@nolouch

Bug Report

What did you do?

User use resource control.

What did you expect to see?

no error report

What did you see instead?

meet exceed resource group quota limitation, but RU usage is below the RU settings.

image

image

image

Event 1 17:40:19.765 17:40:19.765 Coming a request and finding that the local tokens are insufficient, send a notification to the thread that acquires tokens.
Event 2 17:40:19.765 ~ 17:40:20.198 17:40:20.198 The request keep retrying, but the local tokens haven't refreshed yet, continuously logging the same events as before during this retry period. After retrying 500ms timeout,the report failed error to the applications.
Event 3 17:40:20.263 ~17:40:20.265 17:40:20.263 The thread responsible for fetching Tokens received the notification and started to send requests for tokens.17:40:20.265 Obtain new tokens authorization

See the above table. Theoretically, Event 1 should immediately trigger Event 3. After Event 3 succeeds, then enough tokens are obtained during the retry period of Event 2, the request can continue. However, it is possible that since the current event-driven system is similar to a single-threaded event loop, in some cases, the processing delay of a certain message exceeds 500 ms, leading to a failure in obtaining tokens and resulting in an error.

What version of PD are you using (pd-server -V)?

7.5.2

Metadata

Metadata

Assignees

No one assigned

    Labels

    affects-7.1This bug affects the 7.1.x(LTS) versions.affects-7.5This bug affects the 7.5.x(LTS) versions.affects-8.1This bug affects the 8.1.x(LTS) versions.affects-8.2report/customerCustomers have encountered this bug.severity/majortype/bugThe issue is confirmed as a bug.

    Type

    No type

    Projects

    Status

    Closed

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions