Release 0.1.4 (#6)

* Added OKE module with examples for flannel and native cni * changed some parameters * Changed variables names, added preconditions, etc * Finalized the first version of OKE Module * fix: not enforcing TF < 1.3.0 * doc: updates * fix: Operator host added * Added virtual node pools * Added multiple oke examples * feat: clusters added * feat: initial templates * Modified oke examples and readmes * chore: template updated * doc: updates * fix: attributes renamed * fix: examples updated * feat: IAM example added * doc: updates * solved issue with key verification and variables name * solved naming on variables * Modified the native examples * fix: examples updated per new bastion module interface * fix: iam template updated * doc: updates * fix: var name updated * doc: updates * doc: updates * fix: ssh issue * feat: oss module basics * fix: bastion module called remotely * feat: cloud-init basics * feat: cloud-init script cane be provided inline or as file * feat: default_ssh_public_key_path and ssh_public_key_path are overloaded, taking a key string in addition to a key file. * chore: folders renamed * feat: default_ssh_public_key_path and ssh_public_key_path are overloaded, also taking a key string in addition to a key file. * fix: clusters examples updated * fix: clusters updated * doc: updates for clusters * chore: OSS removed * Update README.md * doc: updates * doc: release notes and version bump * chore: vars set with type any * fix: dependency variables strongly typed * chore: dependency variables quoted * doc: updated * doc: updates * doc: examples updated * fic: examples updated * doc: updates * doc: release notes updated --------- Signed-off-by: Andre Correa <andre.correa@oracle.com> Co-authored-by: Ionut Sturzu <ionut.sturzu@oracle.com> Co-authored-by: gfeodoro <gfeodoro@GFEODORO-RO.ro.oracle.com>
oci-landing-zones · May 17, 2024 · 48ad16f · 48ad16f
1 parent 594faa6
commit 48ad16f
Show file tree

Hide file tree

Showing 82 changed files with 3,944 additions and 763 deletions.
diff --git a/.gitignore b/.gitignore
@@ -12,4 +12,7 @@
 marketplace-images/**/*.txt
 platform-images/**/*.txt
 userdata
-mounting-block-volumes.txt
+mounting-block-volumes.txt
+**/examples/clusters/*
+**/creds
+**/clean-setup
diff --git a/README.md b/README.md
@@ -6,7 +6,7 @@ This repository contains Terraform modules for managing workload resources in OC
 
 The following modules are available:
 - [CIS Compute & Storage](./cis-compute-storage/)
-- OKE (Oracle Kubernetes Engine) - soon
+- [OKE (Oracle Kubernetes Engine)](./cis-oke/)
 
 Helper modules:
 - [Platform Images](./platform-images/) - aids in finding OCI Platform images. Use it to obtain image information for provisioning a Compute instance.

diff --git a/RELEASE-NOTES.md b/RELEASE-NOTES.md
@@ -1,3 +1,13 @@
+# May 15, 2024 Release Notes - 0.1.4
+
+## New
+1. OKE module added, supporting basic and enhanced clusters, with managed node pools and virtual node pools. See [OKE module](./cis-oke/README.md) for details.
+
+## Updates
+1. Compute module can now manage cluster networks and compute clusters. See [Clusters](./cis-compute-storage/README.md#clusters-1) for details.
+2. Compute module now supports cloud-init scripts passed in as a file or as a string in [Terraform heredoc style](https://developer.hashicorp.com/terraform/language/expressions/strings#heredoc-strings). See [Compute](./cis-compute-storage/README.md#compute-1) for details.
+3. Compute module now supports SSH public keys passed in as a file or as a string.
+
 # February 29, 2024 Release Notes - 0.1.3
 
 ## Updates

diff --git a/cis-compute-storage/README.md b/cis-compute-storage/README.md
diff --git a/cis-compute-storage/SPEC.md b/cis-compute-storage/SPEC.md
diff --git a/cis-compute-storage/cluster_instances_configuration.tf b/cis-compute-storage/cluster_instances_configuration.tf
diff --git a/cis-compute-storage/clusters.tf b/cis-compute-storage/clusters.tf
@@ -0,0 +1,86 @@
+# Copyright (c) 2024, Oracle and/or its affiliates. All rights reserved.
+# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl.
+
+data "oci_identity_availability_domains" "cluster_ads" {
+  for_each       = var.clusters_configuration != null ? (var.clusters_configuration.clusters != null ? var.clusters_configuration.clusters : {}) : {}
+  compartment_id = each.value.compartment_id != null ? (length(regexall("^ocid1.*$", each.value.compartment_id)) > 0 ? each.value.compartment_id : var.compartments_dependency[each.value.compartment_id].id) : (length(regexall("^ocid1.*$", var.clusters_configuration.default_compartment_id)) > 0 ? var.clusters_configuration.default_compartment_id : var.compartments_dependency[var.clusters_configuration.default_compartment_id].id)
+}
+
+locals {
+  cluster_networks = { for k, v in (var.clusters_configuration != null ? (var.clusters_configuration.clusters != null ? var.clusters_configuration.clusters : {}) : {}) : k => v if lower(coalesce(v.type,"cluster_network")) == "cluster_network"}
+  compute_clusters = { for k, v in (var.clusters_configuration != null ? (var.clusters_configuration.clusters != null ? var.clusters_configuration.clusters : {}) : {}) : k => v if lower(coalesce(v.type,"cluster_network")) != "cluster_network"}
+}
+
+resource "oci_core_cluster_network" "these" {
+  for_each = local.cluster_networks
+    lifecycle {
+      ## Check 1: cluster_network_settings is required for clusters of type "cluster_network".
+      precondition {
+        condition = lower(coalesce(each.value.type,"cluster_network")) == "cluster_network" && each.value.cluster_network_settings == null ? false : true
+        error_message = "VALIDATION FAILURE in cluster \"${each.key}\": \"cluster_network_settings\" is required for clusters of type \"${lower(coalesce(each.value.type,"cluster_network"))}\"."
+      }
+    }  
+    #Required
+    compartment_id = each.value.compartment_id != null ? (length(regexall("^ocid1.*$", each.value.compartment_id)) > 0 ? each.value.compartment_id : var.compartments_dependency[each.value.compartment_id].id) : (length(regexall("^ocid1.*$", var.clusters_configuration.default_compartment_id)) > 0 ? var.clusters_configuration.default_compartment_id : var.compartments_dependency[var.clusters_configuration.default_compartment_id].id)
+    instance_pools {
+        #Required
+        instance_configuration_id = contains(keys(oci_core_instance_configuration.these),each.value.cluster_network_settings.instance_configuration_id) ? oci_core_instance_configuration.these[each.value.cluster_network_settings.instance_configuration_id].id : (length(regexall("^ocid1.*$", each.value.cluster_network_settings.instance_configuration_id)) > 0 ? each.value.cluster_network_settings.instance_configuration_id : null)
+        size = each.value.cluster_network_settings.instance_pool != null ? coalesce(each.value.cluster_network_settings.instance_pool.size,1) : 1
+
+        #Optional
+        display_name = each.value.cluster_network_settings.instance_pool != null ? each.value.cluster_network_settings.instance_pool.name : null
+        defined_tags = each.value.defined_tags != null ? each.value.defined_tags : var.clusters_configuration.default_defined_tags
+        freeform_tags = each.value.freeform_tags != null ? each.value.freeform_tags : var.clusters_configuration.default_freeform_tags
+    }
+    placement_configuration {
+      #Required
+      availability_domain  = data.oci_identity_availability_domains.cluster_ads[each.key].availability_domains[coalesce(each.value.availability_domain,1) - 1].name
+      primary_vnic_subnets {
+        #Required
+        subnet_id = length(regexall("^ocid1.*$", each.value.cluster_network_settings.networking.subnet_id)) > 0 ? each.value.cluster_network_settings.networking.subnet_id : var.network_dependency["subnets"][each.value.cluster_network_settings.networking.subnet_id].id
+        is_assign_ipv6ip = coalesce(each.value.cluster_network_settings.networking.ipv6_enable,false)
+        dynamic "ipv6address_ipv6subnet_cidr_pair_details" {
+          for_each = coalesce(each.value.cluster_network_settings.networking.ipv6_enable,false) ? coalesce(each.value.cluster_network_settings.networking.ipv6_subnet_cidrs,[]) : []
+          content {
+            ipv6subnet_cidr = each.key
+          }
+        }
+      }
+      dynamic "secondary_vnic_subnets" {
+        for_each = each.value.cluster_network_settings.networking.secondary_vnic_settings != null ? [1] : []
+        content {
+          subnet_id = length(regexall("^ocid1.*$", each.value.cluster_network_settings.networking.secondary_vnic_settings.subnet_id)) > 0 ? each.value.cluster_network_settings.networking.secondary_vnic_settings.subnet_id : var.network_dependency["subnets"][each.value.cluster_network_settings.networking.secondary_vnic_settings.subnet_id].id
+          display_name = each.value.cluster_network_settings.networking.secondary_vnic_settings.name
+          is_assign_ipv6ip = coalesce(each.value.cluster_network_settings.networking.secondary_vnic_settings.ipv6_enable,false)
+          dynamic "ipv6address_ipv6subnet_cidr_pair_details" {
+            for_each = coalesce(each.value.cluster_network_settings.networking.secondary_vnic_settings.ipv6_enable,false) ? coalesce(each.value.cluster_network_settings.networking.secondary_vnic_settings.ipv6_subnet_cidrs,[]) : []
+            content {
+              ipv6subnet_cidr = each.key
+            }
+          }
+        }
+      }
+    }
+
+    #Optional
+    # cluster_configuration {
+    #     #Required
+    #     hpc_island_id = oci_core_hpc_island.test_hpc_island.id
+
+    #     #Optional
+    #     network_block_ids = var.cluster_network_cluster_configuration_network_block_ids
+    # }
+
+    display_name = each.value.name
+    defined_tags = each.value.defined_tags != null ? each.value.defined_tags : var.clusters_configuration.default_defined_tags
+    freeform_tags = each.value.freeform_tags != null ? each.value.freeform_tags : var.clusters_configuration.default_freeform_tags
+}
+
+resource "oci_core_compute_cluster" "these" {
+  for_each = local.compute_clusters
+    compartment_id = each.value.compartment_id != null ? (length(regexall("^ocid1.*$", each.value.compartment_id)) > 0 ? each.value.compartment_id : var.compartments_dependency[each.value.compartment_id].id) : (length(regexall("^ocid1.*$", var.clusters_configuration.default_compartment_id)) > 0 ? var.clusters_configuration.default_compartment_id : var.compartments_dependency[var.clusters_configuration.default_compartment_id].id)
+    availability_domain = data.oci_identity_availability_domains.cluster_ads[each.key].availability_domains[coalesce(each.value.availability_domain, 1) - 1].name
+    display_name = each.value.name
+    defined_tags = each.value.defined_tags != null ? each.value.defined_tags : var.clusters_configuration.default_defined_tags
+    freeform_tags = each.value.freeform_tags != null ? each.value.freeform_tags : var.clusters_configuration.default_freeform_tags
+}
diff --git a/cis-compute-storage/compute.tf b/cis-compute-storage/compute.tf
@@ -170,9 +170,10 @@ resource "oci_core_instance" "these" {
       }
     }
     metadata = {
-      ssh_authorized_keys = each.value.ssh_public_key_path != null ? file(each.value.ssh_public_key_path) : file(var.instances_configuration.default_ssh_public_key_path)
-    #  user_data           = contains(keys(data.template_cloudinit_config.config),each.key) ? data.template_cloudinit_config.config[each.key].rendered : null
+      ssh_authorized_keys = each.value.ssh_public_key_path != null ? (fileexists(each.value.ssh_public_key_path) ? file(each.value.ssh_public_key_path) : each.value.ssh_public_key_path) : var.instances_configuration.default_ssh_public_key_path != null ? (fileexists(var.instances_configuration.default_ssh_public_key_path) ? file(var.instances_configuration.default_ssh_public_key_path) : var.instances_configuration.default_ssh_public_key_path): null
+      user_data           = contains(keys(data.template_file.cloud_config),each.key) ? base64encode(data.template_file.cloud_config[each.key].rendered) : null
     }
+    compute_cluster_id   = each.value.cluster_id != null ? (contains(keys(oci_core_compute_cluster.these),each.value.cluster_id) ? oci_core_compute_cluster.these[each.value.cluster_id].id : (length(regexall("^ocid1.*$", each.value.cluster_id)) > 0 ? each.value.cluster_id : null)) : null
 }
 
 resource "oci_core_volume_backup_policy_assignment" "these_boot_volumes" {
@@ -181,6 +182,11 @@ resource "oci_core_volume_backup_policy_assignment" "these_boot_volumes" {
     policy_id = local.oracle_backup_policies[lower(each.value.boot_volume != null ? each.value.boot_volume.backup_policy : "bronze")]
 }
 
+data "template_file" "cloud_config" {
+  for_each = var.instances_configuration != null ? {for k, v in var.instances_configuration["instances"] : k => v if v.cloud_init != null || var.instances_configuration.default_cloud_init_heredoc_script != null || var.instances_configuration.default_cloud_init_script_file != null} : {}
+    template = coalesce(try(each.value.cloud_init.heredoc_script,null), try(file(try(each.value.cloud_init.script_file,null)),null), var.instances_configuration.default_cloud_init_heredoc_script, try(file(var.instances_configuration.default_cloud_init_script_file),null), "__void__")
+}
+
 /* data "template_file" "block_volumes_templates" {
   for_each = var.instances_configuration != null ? {for k, v in var.instances_configuration["instances"] : k => v if v.device_mounting != null} : {}
     template = file("${path.module}/userdata/linux_mount.sh")
@@ -190,19 +196,7 @@ resource "oci_core_volume_backup_policy_assignment" "these_boot_volumes" {
       block_vol_att_type = each.value.device_mounting.emulation_type != null ? lower(each.value.device_mounting.emulation_type) : "paravirtualized"
     }
 }
-
-data "template_cloudinit_config" "config" {
-  for_each      = var.instances_configuration != null ? {for k, v in var.instances_configuration["instances"] : k => v if v.device_mounting != null} : {}
-    gzip          = false
-    base64_encode = true
-
-    # Main cloud-config configuration file.
-    part {
-      filename     = "cloudinit.sh"
-      content_type = "text/x-shellscript"
-      content      = data.template_file.block_volumes_templates[each.key].rendered
-    }
-  } */
+*/  
 
 data "oci_core_vnic_attachments" "these" {
   for_each = var.instances_configuration != null ? var.instances_configuration["instances"] : {}

diff --git a/cis-compute-storage/examples/cluster-networks/README.md b/cis-compute-storage/examples/cluster-networks/README.md
@@ -0,0 +1,40 @@
+# CIS Cluster Network Example
+
+## Introduction
+This example shows how to deploy an RDMA cluster network in OCI using the [cis-compute-storage module](../../). It deploys one Compute instance, one cluster instance configuration, and one cluster network with the characteristics described below. Refer to [input.auto.tfvars.template](./input.auto.tfvars.template) for the variables configuration.
+
+A [cluster network](https://docs.oracle.com/en-us/iaas/Content/Compute/Tasks/managingclusternetworks.htm) is a pool of high performance computing (HPC) instances that are connected with a high-bandwidth, ultra low-latency network. They're designed for highly demanding parallel computing jobs.
+
+### Compute Instance
+- The deployed Compute instance is used as a template for the cluster instance configuration.
+- The Compute instance shape is "BM.Optimized3.36".
+- The Compute instance is created in the compartment and subnet specified by *default_compartment_id* and *default_subnet_id* attributes, respectively, within *instances_configuration* variable.
+
+Note that you must provide the image *name* and *publisher_name* for provisioning the Compute instance. Use the [marketplace-images module](../../../marketplace-images/) to obtain Marketplace images information based on a search filter. It will also return the image OCID that can be used instead of the image name/publisher pair.
+
+### Cluster Instance Configuration
+- A cluster instance configuration is created based on the Compute instance. This is indicated by *template_instance_id* attribute within *cluster_instances_configuration* variable.
+- The cluster instance configuration is created in the compartment specified by *default_compartment_id* attribute within *cluster_instances_configuration* variable.
+
+### Cluster Instance Pool
+- A cluster instance pool is created based on the provided cluster instance configuration.
+
+### RDMA Cluster Network
+- An RDMA cluster network with one cluster instance pool of size 1.
+- The cluster instance pool size is specified by *instance_pool size* attribute within *clusters_configuration* variable.
+- The cluster network is created in the compartment specified by *default_compartment_id* attribute within *clusters_configuration* variable.
+
+## Using this example
+1. Rename *input.auto.tfvars.template* to *\<project-name\>.auto.tfvars*, where *\<project-name\>* is any name of your choice.
+
+2. Within *\<project-name\>.auto.tfvars*, provide tenancy connectivity information and adjust the input variables, by making the appropriate substitutions:
+   - Replace \<REPLACE-WITH-\*\> placeholders with appropriate values. 
+
+Refer to [cis-compute-storage module README.md](../../README.md) for overall attributes usage.
+
+3. In this folder, run the typical Terraform workflow:
+```
+terraform init
+terraform plan -out plan.out
+terraform apply plan.out
+```
diff --git a/cis-compute-storage/examples/cluster-networks/input.auto.tfvars.template b/cis-compute-storage/examples/cluster-networks/input.auto.tfvars.template
@@ -0,0 +1,90 @@
+# Copyright (c) 2023 Oracle and/or its affiliates.
+# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl.
+
+#--------------------------------------------------------------------------------------------------------------------------------------
+# 1. Rename this file to <project-name>.auto.tfvars, where <project-name> is a name of your choice.
+# 2. Provide values for "Tenancy Connectivity Variables".
+# 3. Replace <REPLACE-WITH-*> placeholders with appropriate values.
+#--------------------------------------------------------------------------------------------------------------------------------------
+
+#---------------------------------------
+# Tenancy Connectivity Variables
+#---------------------------------------
+
+tenancy_ocid         = "<tenancy OCID>"            # Get this from OCI Console (after logging in, go to top-right-most menu item and click option "Tenancy: <your tenancy name>").
+user_ocid            = "<user OCID>"               # Get this from OCI Console (after logging in, go to top-right-most menu item and click option "My profile").
+fingerprint          = "<PEM key fingerprint>"     # The fingerprint can be gathered from your user account. In the "My profile page, click "API keys" on the menu in left hand side.
+private_key_path     = "<path to the private key>" # This is the full path on your local system to the API signing private key.
+private_key_password = ""                          # This is the password that protects the private key, if any.
+region               = "<your tenancy region>"     # The region name.
+
+#---------------------------------------
+# Input variables
+#---------------------------------------
+
+# This variable defines the RDMA clusters configuration. 
+# The clusters attribute allows for the definition of an arbitrary number of clusters.
+clusters_configuration = {
+  default_compartment_id = "<REPLACE-WITH-CLUSTER-COMPARTMENT-OCID>"
+  clusters = {
+    "CLUSTER-NETWORK" = {
+      name = "basic-cluster-network"
+      type = "cluster_network"
+      availability_domain = 2
+      cluster_network_settings = {
+        instance_configuration_id = "INSTANCE-CONFIG"
+        instance_pool = {
+          size = 1
+        }
+        networking = {
+          subnet_id = "<REPLACE-WITH-CLUSTER-SUBNET-OCID>"
+          ipv6_enable = true
+        }
+      }
+    }  
+  }
+}
+
+# This variable defines the instances configuration that clusters are built on.
+# The configurations attribute allows for the definition of an arbitrary number of configurations.
+# Each configuration must provide a template_instance_id attribute, which value must be the key of a Compute instance within the instances_configuration variable above.
+cluster_instances_configuration = {
+  default_compartment_id = "<REPLACE-WITH-CLUSTER-COMPARTMENT-OCID>"
+  configurations = {
+    INSTANCE-CONFIG = {
+      name = "cluster-instance-configuration"
+      template_instance_id = "CLUSTER-NETWORK-INSTANCE" 
+    }
+  }
+}
+
+# This variable defines the Compute instances used as templates for the cluster.
+# The instances attribute allows for the definition of an arbitrary number on Compute instances.
+# Cluster networks require that Compute instance shapes is one of: "BM.Optimized3.36", "BM.HPC2.36", "BM.GPU.A100-v2.8", "BM.GPU4.8"
+instances_configuration = {
+  default_compartment_id = "<REPLACE-WITH-INSTANCE-COMPARTMENT-OCID>" 
+  default_subnet_id = "<REPLACE-WITH-INSTANCE-SUBNET-OCID>" 
+  default_ssh_public_key_path = "~/.ssh/id_rsa.pub"
+  instances = {
+    CLUSTER-NETWORK-INSTANCE = {
+      shape = "BM.Optimized3.36"
+      name  = "BM.Optimized3.36 Template Instance"
+      placement = {
+        availability_domain = 2
+        fault_domain = 2
+      }
+      boot_volume = {
+        size = 120
+        preserve_on_instance_deletion = false
+      }
+      networking = {
+        hostname  = "bm-optimized336-template-instance"
+        network_security_groups = ["<REPLACE-WITH-INSTANCE-NSG-OCID>"]
+      }
+      image = {
+        name = "Oracle Linux 7 STIG" # Marketplace image
+        publisher_name = "Oracle Linux"
+      }
+    }
+  }
+}
diff --git a/cis-compute-storage/examples/cluster-networks/main.tf b/cis-compute-storage/examples/cluster-networks/main.tf
@@ -0,0 +1,14 @@
+
+# Copyright (c) 2023 Oracle and/or its affiliates.
+# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl.
+
+module "cluster_networks" {
+  source = "../.."
+  providers = {
+    oci = oci
+    oci.block_volumes_replication_region = oci
+  }
+  clusters_configuration = var.clusters_configuration
+  cluster_instances_configuration = var.cluster_instances_configuration
+  instances_configuration = var.instances_configuration
+}
diff --git a/cis-compute-storage/examples/cluster-networks/providers.tf b/cis-compute-storage/examples/cluster-networks/providers.tf
@@ -0,0 +1,19 @@
+# Copyright (c) 2023 Oracle and/or its affiliates.
+# Licensed under the Universal Permissive License v 1.0 as shown at https://oss.oracle.com/licenses/upl.
+
+provider "oci" {
+  region               = var.region
+  tenancy_ocid         = var.tenancy_ocid
+  user_ocid            = var.user_ocid
+  fingerprint          = var.fingerprint
+  private_key_path     = var.private_key_path
+  private_key_password = var.private_key_password
+}
+
+terraform {
+   required_providers {
+     oci = {
+       source = "oracle/oci"
+     }
+   }
+}