We have 256vCPU quota and using just 8, but we are not able to initiate worked node of t3a.medium

It sounds like you're encountering a vCPU limit issue despite believing you have a higher quota. This discrepancy could be due to several factors:

1. You might be looking at the overall vCPU quota for your account, but AWS maintains separate vCPU quotas for different instance type families. The t3a.medium instances belong to the "Standard (A, C, D, H, I, M, R, T, Z) Spot Instance Request" quota category, which might have a lower limit than your overall vCPU quota.

2. Your current limit for the specific instance family in the Mumbai region appears to be set at 8 vCPUs, which explains why you can't launch additional instances after using those 8 vCPUs.

To resolve this issue, you can:

1. Request a service quota increase specifically for the instance type family you're using (t3a instances) in the Mumbai region through the AWS Support Center.

2. Consider using different instance types that fall under a different vCPU quota category if you need immediate capacity.

3. Terminate any unused EC2 instances to free up vCPU capacity within your current quota.

4. Review your EKS node group or capacity provider configuration to potentially use instance types with different vCPU requirements if appropriate for your workload.

You'll need to specifically request an increase for the Standard instance family quota in the Mumbai region, not just rely on the overall vCPU quota you believe you have.
Sources
Troubleshooting Amazon ECS Managed Instances - Amazon Elastic Container Service
Seeing An error occurred (VcpuLimitExceeded) in slurm_resume.log when I fail to run a job, or in clustermgtd.log, when I fail to create a cluster - AWS ParallelCluster

"Trusted Advisor - Service Limits" is the great place to start. Review your service limits, including vCPU limits. See steps below.

Solution:

• Check Current Limits:
  • Use AWS Trusted Advisor > Service Limits dashboard
  • Navigate to "EC2 Service Limits" section
  • Look for "Running On-Demand Standard (A, C, D, H, I, M, R, T, Z) instances"
  • Verify current usage and limits for Mumbai region
• Understanding the Problem:
  • Different instance families have separate vCPU quotas
  • t3a.medium falls under "Standard instances" quota
  • Current limit appears to be 8 vCPUs in Mumbai region
  • Overall account vCPU quota (256) differs from instance family quotas
• Resolution Steps:
  • a) Short-term options:
    • Terminate unused EC2 instances
    • Use different instance types from other families
    • Review node group configuration
  • b) Long-term solution:
    • Request quota increase through AWS Support Center
    • Specifically request increase for Standard instances in Mumbai
    • Include business justification and expected usage
• Best Practices:
  • Regularly monitor Trusted Advisor for limit warnings
  • Plan capacity requirements in advance
  • Consider multi-AZ deployment
  • Implement proper instance type diversity

For immediate assistance, open AWS Support case with details from Trusted Advisor dashboard.