[jira] [Resolved] (HADOOP-17346) Fair call queue is defeated by abusive service principals

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

[jira] [Resolved] (HADOOP-17346) Fair call queue is defeated by abusive service principals

Steve Loughran (Jira)

     [ https://issues.apache.org/jira/browse/HADOOP-17346?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Eric Payne resolved HADOOP-17346.
    Fix Version/s: 3.2.3
       Resolution: Fixed

Thanks [~ahussein]. I committed to branch-3.1, branch-3.2, branch-3.3, and trunk.

> Fair call queue is defeated by abusive service principals
> ---------------------------------------------------------
>                 Key: HADOOP-17346
>                 URL: https://issues.apache.org/jira/browse/HADOOP-17346
>             Project: Hadoop Common
>          Issue Type: Bug
>          Components: common, ipc
>            Reporter: Ahmed Hussein
>            Assignee: Ahmed Hussein
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 3.3.1, 3.4.0, 3.1.5, 3.2.3
>         Attachments: HADOOP-17346-branch-3.1.001.patch, HADOOP-17346.branch-3.2.001.patch, HADOOP-17346.branch-3.3.001.patch
>          Time Spent: 50m
>  Remaining Estimate: 0h
> [~daryn] reported  that the FCQ prioritizes based on the full kerberos principal (ie. "user/host@realm") rather than short name (ie. "user") to prevent service principals like the DNs and NMs being de-prioritized since service principals are expected to be well behaved.  Notably the DNs contribute a significant but important load so the intent is not to de-prioritize all DNs because their sum total load is high relative to users.
> This has the unfortunate side effect of allowing misbehaving & non-critical service principals to abuse the FCQ. The gstorm/* principals are a prime example.   Each server is spamming opens as fast as possible which ensures that none of the gstorm servers can be de-prioritized because each principal is a fraction of the total load from all principals.
> The secondary and more devasting problem is other abusive non-service principals cannot be effectively de-prioritized.  The sum total of all gstorm load prevents other principals from surpassing the priority thresholds.  Principals stay in the highest priority queues which allows the abusive principals to overflow the entire call queue for extended periods of time.  Notably it prevents the FCQ from moderating the heavy create loads from p_gup @ DB which cause significant performance degradation.
> Prioritization should be based on short name with configurable exemptions for services like the DN/NM.
> [~daryn] suggested a solution that we applied on our clusters.

This message was sent by Atlassian Jira

To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]