Support Questions
Find answers, ask questions, and share your expertise

Yarn Failed Applications Metric

Explorer

Hi, I have a question about the YARN metrics about failed (and killed) applications. I just can't understand what the plot is trying to say to me! The Applications Failed (Cumulative) plot is in application/second, what does it mean ? How should I interpret it ? E.g. 15 May I had 0.02 application/second fails, how many applications did actually fail on that day ? Why there's not a clearer plot like Applications Running (Cumulative) which has on the y-axis the number of applications that are actually running in the cluster ? Is there a way to have a similar plot for the failed and killed applications ?

 

Thanks in advance for the kind support.