Issues with cloudwatch reports for CPU utilization metric

1

Hello Team, we observed that even though an instance is not being used, no one is logged in, and no one connecting through any IDE e.g. Visual Studio Code, still AWS CloudWatch shows CPU utilization kept around 3% . We also verified 'top' and 'w' on Unix console and it shows load average 0.00 for all of 2min, 5min, and 15min.

As we have implemented few lambda triggers to auto-stop instances which are not being used, this issue with CPU utilization reporting prevents us from optimally stopping the unused instances.

As this impacts our AWS usage billing, can you please suggest either of these:

  1. Is there any exising bug or known issue with CloudWatch reporting?
  2. Do we need to configure the CloudWatch reporting or instances with some additional parameters so as to be able to observe accurate results?
  3. Please redirect us to the technical team if this is not in scope of Account issues team.

Thanks.

  • Could you show us how you've configured cloudwatch to trigger the autosleep?

  • yes we can show , as it is repeatedly happening now a days. can some one from AWS team help to connect with us and we can show our issue. day or other we are seeing anomaly in billing.

Keine Antworten

Du bist nicht angemeldet. Anmelden um eine Antwort zu veröffentlichen.

Eine gute Antwort beantwortet die Frage klar, gibt konstruktives Feedback und fördert die berufliche Weiterentwicklung des Fragenstellers.

Richtlinien für die Beantwortung von Fragen