Xuan Gong commented on YARN-611:

Here is my Proposal:
We can make this resetCountPolicy choosable (If users have another 
requirements, we can implement more policies for them). Currently, we will 
provider WindowsSlideAMRetryCountResetPolicy. To use this policy, the users 
need to initiate it by passing a parameter which is used to define period of 
time in milliseconds that AM retry count will be reset. And they can put this 
policy into ApplicationSubmissionContext. In that case, the RMApp and 
RMAppAttempt can get this policy to use. Also, we need to change the way that 
we are using to decide whether this AppAttempt is lastRetry. We can use : 
maxAppAttempts == (getNumFailedAppAttempts() + 1 - this.attemptResetCount) to 
do the calculation. 
Note: getNumFailedAppAttempts() will calculate how many previous attempts are 
really failed (excluding the preemption, nm resync, hardware error and rm 

this.attemptResetCount is used to track the number of failure that we should 
reset. In every resetCountPolicy, we should provide a way to calculate this 
number on time. For WindowsSlideAMRetryCountResetPolicy, after AM successfully 
run a period of time, we can set this.attemptResetCount as the number of really 
failed previous attempts.

Also, we need to provide a way to re-build the  this.attemptResetCount value 
when RM restart/failover happens

> Add an AM retry count reset window to YARN RM
> ---------------------------------------------
>                 Key: YARN-611
>                 URL: https://issues.apache.org/jira/browse/YARN-611
>             Project: Hadoop YARN
>          Issue Type: Bug
>          Components: resourcemanager
>    Affects Versions: 2.0.3-alpha
>            Reporter: Chris Riccomini
>            Assignee: Xuan Gong
> YARN currently has the following config:
> yarn.resourcemanager.am.max-retries
> This config defaults to 2, and defines how many times to retry a "failed" AM 
> before failing the whole YARN job. YARN counts an AM as failed if the node 
> that it was running on dies (the NM will timeout, which counts as a failure 
> for the AM), or if the AM dies.
> This configuration is insufficient for long running (or infinitely running) 
> YARN jobs, since the machine (or NM) that the AM is running on will 
> eventually need to be restarted (or the machine/NM will fail). In such an 
> event, the AM has not done anything wrong, but this is counted as a "failure" 
> by the RM. Since the retry count for the AM is never reset, eventually, at 
> some point, the number of machine/NM failures will result in the AM failure 
> count going above the configured value for 
> yarn.resourcemanager.am.max-retries. Once this happens, the RM will mark the 
> job as failed, and shut it down. This behavior is not ideal.
> I propose that we add a second configuration:
> yarn.resourcemanager.am.retry-count-window-ms
> This configuration would define a window of time that would define when an AM 
> is "well behaved", and it's safe to reset its failure count back to zero. 
> Every time an AM fails the RmAppImpl would check the last time that the AM 
> failed. If the last failure was less than retry-count-window-ms ago, and the 
> new failure count is > max-retries, then the job should fail. If the AM has 
> never failed, the retry count is < max-retries, or if the last failure was 
> OUTSIDE the retry-count-window-ms, then the job should be restarted. 
> Additionally, if the last failure was outside the retry-count-window-ms, then 
> the failure count should be set back to 0.
> This would give developers a way to have well-behaved AMs run forever, while 
> still failing mis-behaving AMs after a short period of time.
> I think the work to be done here is to change the RmAppImpl to actually look 
> at app.attempts, and see if there have been more than max-retries failures in 
> the last retry-count-window-ms milliseconds. If there have, then the job 
> should fail, if not, then the job should go forward. Additionally, we might 
> also need to add an endTime in either RMAppAttemptImpl or 
> RMAppFailedAttemptEvent, so that the RmAppImpl can check the time of the 
> failure.
> Thoughts?

This message was sent by Atlassian JIRA

Reply via email to