Can I find which 'labels' there are for 'Runs waiting to start'?

Answered

Raymond Kivisto

• October 02, 2018 21:26

In DevOptics 'Run Insights', when I see there are 'Runs waiting to start', is there any way to drill down to see which specific 'labels' those runs are waiting for?

The reason I'd like this data is so I can know in real time which type of agents may be lacking in my build agent pool.

I believe I can get the historical data over a time period in the 'Runs per node label' view, under the 'Average queue time' and 'Average queue length' columns.

0

• 

  Alexander Tacho

  • October 03, 2018 09:19

  Correct you can get historic data on "Runs per node label" screen.

   

  Comparing Runs waiting to start with idle executors can help you identify a general problem with how you assign work to executors. But there is no way to drill down to see where the inefficiencies are. 

  What is also interesting in the screenshot the expected runs waiting to start at this time in day historically is 3. 

  Do you have a sense of why there are so many "runs waiting to start"?

   

  That said, we are aware of the gap of drill-down options and are exploring it at the moment. I will record this as an RFE but would love to have additionally more in depth conversation to best solve your needs.

   

  Thanks,
  Alex

  DevOptics PM

  0
• 

  Raymond Kivisto

  • October 03, 2018 13:54

  Thanks very much! I think the historical data is great and key to help me understand the build agent pool utilization for the non-ephemeral agents. I can always get the live information by looking at each one of the Jenkins Masters to see what is in the build queue.

  0

Please sign in to leave a comment.