I was debuging an EC2 volume IO issue, from other logs I can find the issue happened around 20:10 CST,
However, I notice the Cloudwatch metrics “Disk Reads” and “Disk Operations” from the specified EC2 instances is 0 all the time.
After check-in AWS docs, I found:
- If the volume type is EBS, we should check disk operation from the volume cloudwatch monitor.
- The Disk reads and disk operations metric on EC2 apply to local instances storage type (if have).
- If there is no instances storage attached to the EC2, the “Disk Reads” and “Disk Operations” will be 0.
- Different type of instances may have or not have instances storage, refer Instance Storage Lifecycle
Good to note it here.