Mike Metral

TensorFlow v1.10.1 GPU Only (CUDA 9.2, cuDNN 7.2) Python 3.6.4, Ubuntu 18.04

1

[Tensorflow v1.10.1 - GPU Support Only - CUDA 9.2 - cuDNN 7.2 - Python 3.6.4](https://github.com/metral/tensorflow-wheels/tree/master/gpu/tf-v1.10.1/cuda-9.2.148_396.37_cudnn-7.2.1-py3.6.4)

Consider adding missing envvar configs for the aws-k8s-cni

3

### Problem description The following configs that are present in the [upstream docs](https://git.io/Jvlah) are currently not available for CNI config: - ENI_CONFIG_ANNOTATION_DEF - ENI_CONFIG_LABEL_DEF - AWS_VPC_K8S_CNI_RANDOMIZESNAT - AWS_VPC_K8S_CNI_EXCLUDE_SNAT_CIDRS - MINIMUM_IP_TARGET...

In previews, adding tags to roles is triggering a replace of an existing cluster and LaunchConfig since the respective role `.arns` used by each on updates is perceived to change,...

Stringify the kubeconfig generated instead of returning an Output<t>

2

### Problem description The default kubeconfig file is returned as a type of `Output`. To use this in providers, users have to take the additional step of stringifying the object....

Changing the instanceType in an existing NodeGroup results in a 400 failure to delete the LaunchConfig since it is attached to an active ASG

24

When a NodeGroup is stood up with a given instance type, e.g. `t2.medium`, and then on a future update is changed to say `t3.large`, results in the following error: ```...

Add a couple of examples demonstrating: - [ ] Using full Service spec to set & create a public LB - [ ] Talk to defaults done (e.g. names, metadata,...

File: [docs/reference/pkg/nodejs/pulumi/aws/eks/_index.md](https://www.pulumi.com/docs/reference/pkg/nodejs/pulumi/aws/eks/) The "Enabling Control Plane Logging" example in this file is incorrect, as AWS manages the log group and logs for the control plane. User's cannot provide their own...