search

oracle / terraform-kubernetes-installer

Terraform Installer for Kubernetes on Oracle Cloud Infrastructure
Other
179 stars 118 forks source link

Metadata size limit of 32000 bytes being hit. #189

Open eclectiqus opened 6 years ago

eclectiqus commented 6 years ago

Terraform Version

terraform -v Terraform v0.11.6 + provider.null v1.0.0 + provider.oci (unversioned) + provider.random v1.2.0 + provider.template v1.0.0 + provider.tls v1.1.0

OCI Provider Version

terraform-provider-oci_v2.1.3 2018/04/09 00:12:08 [INFO] terraform-provider-oci 2.1.3 This binary is a plugin. These are not meant to be executed directly. Please execute the program that consumes these plugins, which will load any plugins automatically

Terraform Installer for Kubernetes Version

git rev-parse master df6e012aa8b154dc97b368cb8455ded9a5aa2b83

Input Variables

# Values of non-sensitive input variables

Description of issue:

This is my first time trying to setup the terraform k8s installer. I managed to get around the issue reported elsewhere of the null OCI images references/lookup. Now the error I get is about reaching the metadata size limit of 32000 bytes.

Error message: Error: Error applying plan:

3 error(s) occurred:

I'm not sure where I would either be able to decrease the metadata or increase the limit to 33000 bytes.

garthy commented 6 years ago

Hi can you post your input variables? (The non sensitive bits)

eclectiqus commented 6 years ago

Sure. It's pretty vanilla from the example file though.

region = "us-ashburn-1" cloud_controller_user_ocid = "ocid1.tenancy.oc1..aaaaaaaafzkf23c5nxd47p7kmwkaevnpghc5umfdmrrizgtec3kbocaz5srq"

etcdShape = "VM.Standard1.2" k8sMasterShape = "VM.Standard1.8" k8sWorkerShape = "VM.Standard1.8"

etcdAd1Count = "1" etcdAd2Count = "1" etcdAd3Count = "1"

k8sMasterAd1Count = "1" k8sMasterAd2Count = "1" k8sMasterAd3Count = "1"

k8sWorkerAd1Count = "1" k8sWorkerAd2Count = "1" k8sWorkerAd3Count = "1"

etcdLBShape = "100Mbps" k8sMasterLBShape = "100Mbps"

etcd_ssh_ingress = "10.0.0.0/16" etcd_ssh_ingress = "0.0.0.0/0" etcd_cluster_ingress = "10.0.0.0/16" master_ssh_ingress = "0.0.0.0/0" worker_ssh_ingress = "0.0.0.0/0" master_https_ingress = "0.0.0.0/0" worker_nodeport_ingress = "0.0.0.0/0" worker_nodeport_ingress = "10.0.0.0/16"

control_plane_subnet_access = "public" k8s_master_lb_access = "public" natInstanceShape = "VM.Standard1.2" nat_instance_ad1_enabled = "true" nat_instance_ad2_enabled = "false" nat_instance_ad3_enabled = "true" nat_ssh_ingress = "0.0.0.0/0" public_subnet_http_ingress = "0.0.0.0/0" public_subnet_https_ingress = "0.0.0.0/0"

worker_iscsi_volume_create = true worker_iscsi_volume_size = 100

etcd_iscsi_volume_create = true etcd_iscsi_volume_size = 50

eclectiqus commented 6 years ago

odd. I've refreshed my copy of the project, saw a bunch of changed files. And now I'm getting an error that VM.Standard1.8 is not found or not authorized:

Error: Error applying plan:

3 error(s) occurred:

eclectiqus commented 6 years ago

I notched the shape down to standard1.2 and viola, it worked!

Now scripts/cluster-check.sh is failing which is strange, but probably because the authorized public key is somehow not showing up on the nodes.

benofben commented 6 years ago

I'm seeing the same metadata size issue. I'm trying to stuff a 90k python script into cloud-init and we don't seem to allow that:

Error: Error applying plan:

1 error(s) occurred:

* oci_core_instance.utility: 1 error(s) occurred:

* oci_core_instance.utility: Service error:InvalidParameter. Metadata size is 174986 bytes and cannot be larger than 32000 bytes. http status code: 400. Opc request id: d3aeda8c3dd723c3a37994c57465db7e/038961F6E91429A60AF47B0BE9DD76AF/4EA48E454B9ACE92F7ABBB81F0A27A4A

What would it take to bump this limit?