9da2fc4b8b
Rather than assuming a short sleep will work, we instead wait until netcat is listening of the socket. We've also configured the netcat listener to persist after the first connection, which allows Vault and us to check the connection without the process closing. As we implemented this we also ran into AWS issues in us-east-1 and us-west-2, so we've changed our deploy regions until those issues are resolved. Signed-off-by: Ryan Cragun <me@ryan.ec>
37 lines
828 B
HCL
37 lines
828 B
HCL
# Copyright (c) HashiCorp, Inc.
|
|
# SPDX-License-Identifier: BUSL-1.1
|
|
|
|
variable "vault_instance_count" {
|
|
type = number
|
|
description = "How many vault instances are in the cluster"
|
|
}
|
|
|
|
variable "vault_pods" {
|
|
type = list(object({
|
|
name = string
|
|
namespace = string
|
|
}))
|
|
description = "The vault instances for the cluster to verify"
|
|
}
|
|
|
|
variable "vault_bin_path" {
|
|
type = string
|
|
description = "The path to the vault binary"
|
|
default = "/bin/vault"
|
|
}
|
|
|
|
variable "vault_root_token" {
|
|
type = string
|
|
description = "The vault root token"
|
|
}
|
|
|
|
variable "kubeconfig_base64" {
|
|
type = string
|
|
description = "The base64 encoded version of the Kubernetes configuration file"
|
|
}
|
|
|
|
variable "context_name" {
|
|
type = string
|
|
description = "The name of the k8s context for Vault"
|
|
}
|