EC2

This page describes the Optimization Pack for AWS EC2.

Metrics

CPU and Networking

Name	Unit	Description
cpu_util	percent	The average CPU utilization % across all the CPUs (i.e., how much time on average the CPUs are busy doing work)
network_in_bytes_details	bytes/s	The number of inbound network packets in bytes per second broken down by network device (e.g., wlp4s0)
network_out_bytes_details	bytes/s	The number of outbound network packets in bytes per second broken down by network device (e.g., eth01)

Disks

Name	Unit	Description
disk_read_bytes	bytes/s	The number of bytes per second read across all disks
disk_write_bytes	bytes/s	The number of bytes per second written across all disks
aws_ec2_disk_iops_reads	ops/s	The per second average number of EBS IO disk-read operations summed across all disks
aws_ec2_disk_iops_writes	ops/s	The per second average number of EBS IO disk-write operations summed across all disks
aws_ec2_disk_iops	ops/s	The per second average number of EBS IO disk operations summed across all disks

CPU Credits

Name	Unit	Description
aws_ec2_credits_cpu_available	credits	The number of earned CPU credits that an instance has accrued since it was launched or started. Credits are accrued in the credit balance after they are earned, and removed from the credit balance when they are spent
aws_ec2_credits_cpu_used	credits	The number of CPU credits spent by the instance for CPU utilization

Disk Credits

Name	Unit	Description
aws_ec2_ebs_credits_io_util	percent	The percentage of I/O credits remaining in the burst bucket
aws_ec2_ebs_credits_bytes_util	percent	The percentage of throughput credits remaining in the burst bucket

Pricing

Name	Unit	Description
aws_ec2_price	dollars	AWS EC2 hourly instance price (on-demand)

Parameters

Notice: for the following parameters to take effect, the instance needs to be stopped and changes need to be applied before restarting the instance.

Name	Unit	Type	Default Value	Domain	Restart	Description
aws_ec2_instance_type		Categorical	m5	c5,c5d,c5a,c6g,c6gd,c6gd, r5,r5d,r5a,r5ad,r6g,r6gd, m5,m5d,m5a,m5ad,m6g,m6gd, t3,t3a, a1,z1d	yes	Instance types comprise varying combinations of CPU, memory, storage, and networking capacity, optimized to fit different use cases
aws_ec2_instance_size		Ordinal	large	nano, micro,small,medium,large, xlarge,2xlarge,4xlarge,8xlarge, 9xlarge,12xlarge,16xlarge, 18xlarge,24xlarge	yes

Constraints

The following table shows a sample of constraints that are required in the definition of the study, depending on the tuned parameters.

Notice that AWS does not support all combinations of instance types and sizes, so it is better to specify them beforehand in your constraints to avoid unnecessary experiment failures.

Instance size

Instance size is an ordinal parameter, this means you constraint it by using a 0-based index, as in this example:

aws.aws_ec2_instance_type === "c5" && (aws.aws_ec2_instance_size === 0 || aws.aws_ec2_instance_size === 2) || aws.aws_ec2_instance_size === "m5" && (aws.aws_ec2_instance_size === 0 || aws.aws_ec2_instance_size === 1)

Domain of aws_ec2_instance_size here is: [4xlarge, 8xlarge, 9xlarge]. Since "c5" instance type does not support 8xlarge instance size, and "m5" instance family does not support the 9xlarge one, this option is enforced with a constraint that allows only c5.4xlarge, c5.9xlarge, m5.4xlarge, and m5.8xlarge configurations