Merge pull request red-hat-storage#10494 from petr-balogh/add-ibmclou…

…d-roks-multizone-support

Add multi-zone deployment for ROKS IBM Cloud

conf/deployment/ibmcloud/ibm_cloud_managed_3az_0m_3w_vpc_cluster.yaml

@@ -0,0 +1,30 @@
+# This is the config for IBM cloud ROKS deployment
+# for 3 AZ setup.
+---
+ENV_DATA:
+  platform: 'ibm_cloud'
+  deployment_type: 'managed'
+  region: 'eu-de'
+  zone: 'eu-de-3'
+  provider: "vpc-gen2"
+  worker_instance_type: "bx2.16x64"
+  master_replicas: 0
+  worker_replicas: 3
+  worker_availability_zones:
+    - 'eu-de-1'
+    - 'eu-de-2'
+    - 'eu-de-3'
+  rhel_workers: true
+  # Following values needs to be set in separate config and passed to ocs-ci in
+  # order to deploy OCP/OCS cluster on IBM Cloud
+  # vpc_id: VPC ID PLACEHOLDER
+  # subnet_id: SUBNET ID PLACEHOLDER
+  # cos_instance: COS INSTANCE PLACEHOLDER
+#AUTH:
+#  ibmcloud:
+#    api_key: IBM CLOUD API KEY PLACEHOLDER
+#    account_id: ACCOUNT ID PLACEHOLDER
+#    ibm_cos_access_key_id: KEY PLACEHOLDER
+#    ibm_cos_secret_access_key: SECRET PLACEHOLDER
+#DEPLOYMENT:
+#  ocs_secret_dockerconfigjson: BASE64 OF QUAY SECRET PLACEHOLDER

conf/deployment/ibmcloud/ibm_cloud_vpc_cluster.yaml

@@ -10,6 +10,8 @@ ENV_DATA:
   worker_instance_type: "bx2.16x64"
   master_replicas: 0
   worker_replicas: 3
+  worker_availability_zones:
+    - 'eu-de-3'
   rhel_workers: true
   # Following values needs to be set in separate config and passed to ocs-ci in
   # order to deploy OCP/OCS cluster on IBM Cloud

docs/supported_platforms.md

@@ -86,6 +86,13 @@ But an example is pasted also here:
 ENV_DATA:
   vpc_id: VPC ID PLACEHOLDER
   subnet_id: SUBNET ID PLACEHOLDER
+  # Subnet id from dict below will be used if user defines worker_availability_zones
+  # This is for multi zone deployment support. So If you have worker_availability_zones and
+  # subnet_ids_per_zone defined in credential conf it will be used instead of subnet_id.
+  subnet_ids_per_zone:
+    "eu-de-1": "PLACEHOLDER_SUBNET_ID"
+    "eu-de-2": "PLACEHOLDER_SUBNET_ID"
+    "eu-de-3": "PLACEHOLDER_SUBNET_ID"
   cos_instance: COS INSTANCE PLACEHOLDER
 AUTH:
   ibmcloud:

ocs_ci/ocs/node.py

@@ -1697,7 +1697,6 @@ def verify_all_nodes_created():
         expected_num_nodes += 3
     else:
         expected_num_nodes += config.ENV_DATA.get("infra_replicas", 0)
-
     existing_num_nodes = len(get_all_nodes())
 
     # Some nodes will take time to create due to the issue https://issues.redhat.com/browse/SDA-6346

ocs_ci/utility/ibmcloud.py

@@ -189,16 +189,22 @@ def create_cluster(cluster_name):
 
     """
     provider = config.ENV_DATA["provider"]
+    worker_availability_zones = config.ENV_DATA.get("worker_availability_zones", [])
+    worker_zones_number = len(worker_availability_zones)
     zone = config.ENV_DATA["zone"]
     flavor = config.ENV_DATA["worker_instance_type"]
     worker_replicas = config.ENV_DATA["worker_replicas"]
+    if worker_zones_number > 1:
+        worker_replicas = 2
     ocp_version = get_ibmcloud_ocp_version()
-
     cmd = (
         f"ibmcloud ks cluster create {provider} --name {cluster_name}"
         f" --flavor {flavor}  --workers {worker_replicas}"
         f" --kube-version {ocp_version}"
     )
+    # Reloading correct number of worker replica for later usage.
+    if worker_zones_number > 1:
+        worker_replicas = int(config.ENV_DATA["worker_replicas"] / worker_zones_number)
     if provider == "vpc-gen2":
         semantic_ocp_version = util_version.get_semantic_ocp_version_from_config()
         if semantic_ocp_version >= util_version.VERSION_4_15:
@@ -215,11 +221,27 @@ def create_cluster(cluster_name):
     cluster_info = get_cluster_details(cluster_name)
     # Create metadata file to store the cluster name
     cluster_info["clusterName"] = cluster_name
-    cluster_info["clusterID"] = cluster_info["id"]
+    cluster_id = cluster_info["id"]
+    cluster_info["clusterID"] = cluster_id
     cluster_path = config.ENV_DATA["cluster_path"]
     metadata_file = os.path.join(cluster_path, "metadata.json")
     with open(metadata_file, "w+") as f:
         json.dump(cluster_info, f)
+    for worker_zone in worker_availability_zones:
+        if worker_zone == zone:
+            continue
+        subnet = config.ENV_DATA["subnet_ids_per_zone"][worker_zone]
+        cmd = (
+            f"ibmcloud oc zone add {provider} --subnet-id {subnet}  "
+            f"--cluster {cluster_id} --zone {worker_zone} --worker-pool default"
+        )
+        run_ibmcloud_cmd(cmd)
+    if worker_zones_number > 1:
+        cmd = (
+            f"ibmcloud ks worker-pool resize --cluster {cluster_name} --worker-pool "
+            f"default --size-per-zone {worker_replicas}"
+        )
+        run_ibmcloud_cmd(cmd)
     # Temporary increased timeout to 10 hours cause of issue with deployment on
     # IBM cloud
     timeout = 36000