0

I am trying to create two cloudwatch alarm on my Elastic LoadBalancer. Suppose there is 3 instance attached to my elb.

If one instance goes unhealthy then I should trigger an alarm saying it's a WARNING and if all instance is unhealthy then I should trigger an alarm saying it's a CRITICAL.

I am experimenting with cloud watch metrics MAX, MIN, AVG and I am not getting right condition to put into.

This is what I am doing right now.

For CRITICAL if maximum no of healthy host < 1 .

This works but is taking a long time for creating that alarm.

For WARNING if maximum no of unhealthy host > 0.

I dont have much confidence in this because I am not able to verify this .

Colwin
  • 2,655
  • 3
  • 25
  • 25
amulamul
  • 372
  • 3
  • 14
  • how much time its taking for creating these alarms ? – Amit May 16 '17 at 05:08
  • @AmitK I don't have exact time but I guess it more than 10 minutes. Do you have idea what might causing this to happen ? – amulamul May 16 '17 at 14:12
  • aws don't push these metrics before 5 mins, if you want you can enable fine grained metrics which update in 1 min but there is extra charge for it. also read http://docs.aws.amazon.com/elasticloadbalancing/latest/classic/elb-cloudwatch-metrics.html#ViewingDataUsingCloudWatch – Amit May 17 '17 at 03:11
  • It was waiting for instance be to unhealthy before alarm got triggered which was 10 calls each in every 5 sec interval. That what was taking that time. I validated and everything works fine now ! – amulamul May 23 '17 at 18:28

0 Answers0