CPU data graph

Options
Mauro Mequelussi-1595850
Mauro Mequelussi-1595850 Member ✭
in ❓Other questions
I've noticed that the instance's CPU utilization graph has changed.

What worried me was that before my average was usually no more than 10, now it's peaking at almost 100%.

Has anyone else noticed this?[Captura de Tela 2021-10-17 às 21.32.32.png]
· Share on Twitter

Comments

  • Sean Montgomery
    Sean Montgomery Administrator

    ADMIN
    Options
    Hi ,

    The previous graph before was averaging several components together which was making the usage look lower than it really was. Now it is breaking apart the two most important components and graphing their usage only. 

    I'll have the team double check to make sure the numbers are correct. This is definitely math heavy so worth a 2nd look.
    · Share on Twitter
  • Sean Montgomery
    Sean Montgomery Administrator

    ADMIN
    Options
    Hey ,

    I believe I found a discrepancy. Please give it another look. It should be lower now. I'm still going to have the team double check this graph.
    · Share on Twitter
  • Mauro Mequelussi-1595850
    Mauro Mequelussi-1595850 Member ✭
    Options
    [Captura de Tela 2021-10-18 às 11.58.41.png] It looks like it has improved yes, but it still looks bigger than the previous graphics.

    As soon as you can confirm with the team if it's correct, let me know.
    · Share on Twitter
  • Sean Montgomery
    Sean Montgomery Administrator

    ADMIN
    Options
     just confirmed everything is correct. 

    This looks pretty close to your previous usage. Remember before that graphs were getting averaged together. This was creating a false representation of the actual performance that mattered. The primary candidates for performance are the API layer and the database. Now those are being represented properly.
    · Share on Twitter
  • Mauro Mequelussi-1595850
    Mauro Mequelussi-1595850 Member ✭
    Options
    Has anything changed in the graphics again?

    Looks like the graphics are up again. As it only has the last 24 hours, it is difficult to analyze when it started to rise.

    Also, some plan to have graphs of what is being consumed, per workspace/api/function. I know you can see individual on each endpoint, but to understand what's consuming when working with multiple workspaces and endpoints.[Captura de Tela 2021-10-27 às 14.58.57.png] 
    · Share on Twitter
  • Sean Montgomery
    Sean Montgomery Administrator

    ADMIN
    Options
    Hi ,

    Nothing has changed. Everything is running properly.

    We do not have plans for CPU graphs on a per api level. That is easier said than done. Right now the CPU graphs are limited to the instance, as it is not trivial to see exact resource usage on a particular endpoint. 

    I agree it would be very useful, but it just isn't possible at the moment.

    In terms of performance of api endpoints, you should be looking at the request history and then looking at how long each endpoint is taking. Take a look at the screenshot below.

    [Image]
    · Share on Twitter

Categories