1

I am new to AWS, I have auto scaling enabled for my Elastic Beanstalk based server. For some reason the below healthd process is almost fully consuming the cpu causing the auto scaler to add a new instance as I have set the expansion policy to add a new instance to the beanstalk environment when resource utilization is > 70.

healthd   20   0 1024648  43660   9876 S  75.7  1.1 121:16.03 ruby

I have two questions:

  1. How to avoid 5xx (network) errors when a new instance is added?
  2. What is the process Healthd needed for? why it is running when I didn't start it? and how can I prevent that process from draining the CPU?

Maybe the Load Balancer is starting to send traffic to the new instance before the application is completely up on that instance and this is why I am getting the network errors! How can I verify this is the cause and how to avoid it?

marc_s
  • 732,580
  • 175
  • 1,330
  • 1,459
A_Matar
  • 2,210
  • 3
  • 31
  • 53
  • `healthd` is related to the enhanced health reporting options associated with EB and other AWS solutions. Can you share some additional details on: your instance size, the platform you are running, your root volume config settings, and if you're leveraging any swapfile configs via .ebextensions? – dusthaines Jun 22 '21 at 16:39

0 Answers0