cortexlabs/cortex

instance type g5.xlarge is not supported by cortex

dreamflasher opened this issue · 2 comments

dreamflasher commented

Version

0.42

Description

error: cluster.yaml: node_groups: index 0: instance_type: instance type g5.xlarge is not supported by cortex

Configuration

# cluster.yaml

# cluster name
cluster_name: cortex-sr

# AWS region
region: us-west-2

# list of cluster node groups;
node_groups:
  - name: ng-gpu
    instance_type: g5.2xlarge
    min_instances: 1
    max_instances: 5
    instance_volume_size: 50
    instance_volume_type: gp3
    spot: true
  # ...

# API load balancer type [nlb | elb]
api_load_balancer_type: nlb

# API load balancer scheme [internet-facing | internal]
api_load_balancer_scheme: internet-facing

# operator load balancer scheme [internet-facing | internal]
# note: if using "internal", you must configure VPC Peering to connect your CLI to your cluster operator
operator_load_balancer_scheme: internet-facing

# restrict access to APIs by cidr blocks/ip address ranges
api_load_balancer_cidr_white_list: [0.0.0.0/0]

# restrict access to the Operator by cidr blocks/ip address ranges
operator_load_balancer_cidr_white_list: [0.0.0.0/0]

# list of IAM policies to attach to your Cortex APIs
iam_policy_arns: ["arn:aws:iam::aws:policy/AmazonS3FullAccess"]

# instance type for prometheus (use an instance with more memory for clusters exceeding 300 nodes or 300 pods)
prometheus_instance_type: "t3.medium"

Steps to reproduce

  1. Start a g5 cluster

Expected behavior

Cluster starts

Actual behavior

error: cluster.yaml: node_groups: index 0: instance_type: instance type g5.xlarge is not supported by cortex

Suggested solution

Please support g5

deliahu commented

Unfortunately we are still waiting for a new release of the AWS CNI which supports g5 instances. Once that happens, it should be straightforward to add support in Cortex.

oborchers commented

@deliahu: Is there any update on this one?

  • 👀1