Update backup to mount the same BV as bastion

Author: arnaudfroidmont

autoscaling/tf_init/bastion_update.tf

@@ -57,7 +57,6 @@ resource "local_file" "inventory" {
     login_block = var.login_block,
     scratch_nfs_type = local.scratch_nfs_type,
     bastion_mount_ip = var.bastion_mount_ip,
-    backup_mount_ip = var.backup_mount_ip,
     login_mount_ip = var.login_mount_ip,
     cluster_mount_ip = local.mount_ip,
     cluster_name = local.cluster_name,

autoscaling/tf_init/inventory.tpl

@@ -44,7 +44,6 @@ bastion_block = ${bastion_block}
 login_block = ${login_block} 
 scratch_nfs_type = ${scratch_nfs_type}
 bastion_mount_ip = ${bastion_mount_ip}
-backup_mount_ip = ${backup_mount_ip}
 login_mount_ip = ${login_mount_ip}
 cluster_mount_ip = ${cluster_mount_ip}
 autoscaling = true

bastion.tf

@@ -15,6 +15,7 @@ resource "oci_core_volume_attachment" "bastion_volume_attachment" {
   instance_id     = oci_core_instance.bastion.id
   display_name    = "${local.cluster_name}-bastion-volume-attachment"
   device          = "/dev/oracleoci/oraclevdb"
+  is_shareable    = true
 } 
 
 resource "oci_core_volume_backup_policy" "bastion_boot_volume_backup_policy" {
@@ -267,7 +268,6 @@ resource "null_resource" "cluster" {
       login_block = var.login_block, 
       scratch_nfs_type = local.scratch_nfs_type,
       bastion_mount_ip = local.bastion_mount_ip,
-      backup_mount_ip = local.backup_mount_ip,
       login_mount_ip = local.login_mount_ip,
       cluster_mount_ip = local.mount_ip,
       autoscaling = var.autoscaling,
@@ -394,7 +394,6 @@ resource "null_resource" "cluster" {
       login_block = var.login_block, 
       scratch_nfs_type = local.scratch_nfs_type,
       bastion_mount_ip = local.bastion_mount_ip,
-      backup_mount_ip = local.backup_mount_ip,
       login_mount_ip = local.login_mount_ip,
       cluster_mount_ip = local.mount_ip,
       scratch_nfs_type_cluster = var.scratch_nfs_type_cluster,

inventory.tpl

@@ -38,7 +38,6 @@ bastion_block = ${bastion_block}
 login_block = ${login_block}
 scratch_nfs_type = ${scratch_nfs_type}
 bastion_mount_ip = ${bastion_mount_ip}
-backup_mount_ip = ${backup_mount_ip}
 login_mount_ip = ${login_mount_ip}
 cluster_mount_ip = ${cluster_mount_ip}
 autoscaling = ${autoscaling}

locals.tf

@@ -44,7 +44,6 @@ locals {
   is_instance_pool_flex_shape = length(regexall(".*VM.*.*Flex$", var.instance_pool_shape)) > 0 ? [local.instance_pool_ocpus]:[]
 
   bastion_mount_ip = var.bastion_block ? element(concat(oci_core_volume_attachment.bastion_volume_attachment.*.ipv4, [""]), 0) : "none"
-  backup_mount_ip = var.bastion_block && var.slurm_ha ? element(concat(oci_core_volume_attachment.backup_volume_attachment.*.ipv4, [""]), 0) : "none"
   login_mount_ip = var.login_block ? element(concat(oci_core_volume_attachment.login_volume_attachment.*.ipv4, [""]), 0) : "none"
 
   scratch_nfs_type = var.cluster_network ? var.scratch_nfs_type_cluster : var.scratch_nfs_type_pool 

playbooks/roles/iscsi/tasks/el.yml

@@ -23,6 +23,7 @@
   filesystem:
     dev: '{{ target["devicenodes"][0] }}'
     fstype: xfs
+  when: not ('slurm_backup' in group_names)
 
 - name: Mount local volume
   mount:

playbooks/site.yml

@@ -178,8 +178,8 @@
 - hosts: slurm_backup
   become: true
   vars:
-    iscsi_ip: "{{ backup_mount_ip }}"
-    local_path: "/mnt/block"
+    iscsi_ip: "{{ bastion_mount_ip }}"
+    local_path: "/mnt/nfs_backup"
   tasks: 
     - include_role: 
         name: iscsi

slurm_ha.tf

@@ -1,20 +1,11 @@
-resource "oci_core_volume" "backup_volume" { 
-  count = var.bastion_block && var.slurm_ha ? 1 : 0
-  availability_domain = var.bastion_ad
-  compartment_id = var.targetCompartment
-  display_name = "${local.cluster_name}-backup-volume"
-  size_in_gbs = var.bastion_block_volume_size
-  vpus_per_gb = split(".", var.bastion_block_volume_performance)[0]
-} 
-
-
 resource "oci_core_volume_attachment" "backup_volume_attachment" { 
   count = var.bastion_block && var.slurm_ha ? 1 : 0
   attachment_type = "iscsi"
-  volume_id       = oci_core_volume.backup_volume[0].id
+  volume_id       = oci_core_volume.bastion_volume[0].id
   instance_id     = oci_core_instance.backup[0].id
   display_name    = "${local.cluster_name}-backup-volume-attachment"
   device          = "/dev/oracleoci/oraclevdb"
+  is_shareable    = true
 } 
 
 resource "oci_core_instance" "backup" {
@@ -60,7 +51,7 @@ resource "oci_core_instance" "backup" {
 
 resource "null_resource" "backup" { 
   count = var.slurm_ha ? 1 : 0
-  depends_on = [oci_core_instance.backup, oci_core_volume_attachment.backup_volume_attachment ] 
+  depends_on = [oci_core_instance.backup] 
   triggers = { 
     backup = oci_core_instance.backup[0].id
   } 
@@ -186,7 +177,7 @@ resource "null_resource" "backup" {
 }
 resource "null_resource" "cluster_backup" { 
   count = var.slurm_ha ? 1 : 0
-  depends_on = [null_resource.backup, oci_core_compute_cluster.compute_cluster, oci_core_cluster_network.cluster_network, oci_core_instance.backup, oci_core_volume_attachment.backup_volume_attachment ] 
+  depends_on = [null_resource.backup, oci_core_compute_cluster.compute_cluster, oci_core_cluster_network.cluster_network, oci_core_instance.backup ] 
   triggers = { 
     cluster_instances = join(", ", local.cluster_instances_names)
   } 
@@ -231,7 +222,6 @@ resource "null_resource" "cluster_backup" {
       login_block = var.login_block, 
       scratch_nfs_type = local.scratch_nfs_type,
       bastion_mount_ip = local.bastion_mount_ip,
-      backup_mount_ip = local.backup_mount_ip,
       login_mount_ip = local.login_mount_ip,
       cluster_mount_ip = local.mount_ip,
       autoscaling = var.autoscaling,
@@ -358,7 +348,6 @@ resource "null_resource" "cluster_backup" {
       login_block = var.login_block, 
       scratch_nfs_type = local.scratch_nfs_type,
       bastion_mount_ip = local.bastion_mount_ip,
-      backup_mount_ip = local.backup_mount_ip,
       login_mount_ip = local.login_mount_ip,
       cluster_mount_ip = local.mount_ip,
       scratch_nfs_type_cluster = var.scratch_nfs_type_cluster,

variables.tf

@@ -93,8 +93,8 @@ variable "marketplace_version_id" {
        "4" = "OL7.9-OFED5.0-2.1.8.0-RHCK-20210709"
        "HPC_OL7" = "OracleLinux-7-OCA-RHCK-OFED-5.8-3.0.7.0-2023.12.04-1"
        "HPC_OL8" = "OracleLinux-8-OCA-RHCK-OFED-5.8-3.0.7.0-2023.12.04-1"
-       "GPU_OL7" = "OracleLinux-7-OCA-RHCK-OFED-5.8-3.0.7.0-GPU-535-2023.12.04-1"
-       "GPU_OL8" = "OracleLinux-8-OCA-RHCK-OFED-5.8-3.0.7.0-GPU-535-2023.12.04-1"
+       "GPU_OL7" = "OracleLinux-7-OCA-RHCK-OFED-5.8-3.0.7.0-GPU-535-2023.12.04-2"
+       "GPU_OL8" = "OracleLinux-8-OCA-RHCK-OFED-5.8-3.0.7.0-GPU-535-2023.12.04-2"
   }
 }