Issue
I am using r3.4xlarge instance with 1.2 TB io1 disk attached with 12000 IOPS. When the load increases (writing of data) I noticed there is a latency/lag in disk I/O . From the CloudWatch metrics it is clear that during peak VolumeWriteBytes is always stuck around ~103KB. Is there a way I can increase this?
Solution
The issue was because of the instance r3.4xlarge which is not EBS optimized by default. Also I haven't enabled EBS optimization for r3.4xlarge explicitly which was the reason for the latency as the EBS throughput is be restricted by the maximum network throughput of the instance which is also shared for EBS and EC2 network bandwidth since it is not EBS optimized by default.
AWS Document -> https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ebs-optimized.html
r3.4xlarge must have EBS optimization enabled to achieve the level of performance described in the document. When you enable EBS optimization for an instance that is not EBS-optimized by default, you pay an additional low, hourly fee for the dedicated capacity as well. So, yes this could potentially .
Answered By - Matt Answer Checked By - Mary Flores (WPSolving Volunteer)