Merge branch 'master' into multi-cnv-workload

conf/README.md

@@ -349,6 +349,7 @@ higher priority).
 * `continue_upgrade_after_checks_even_if_not_healthy` -  if set to true Rook will continue the OSD daemon upgrade process even if the PGs are not clean.
 * `upgrade_osd_requires_healthy_pgs` - If set to true OSD upgrade process won't start until PGs are healthy.
 * `workaround_mark_disks_as_ssd` - WORKAROUND: mark disks as SSD (not rotational - `0` in `/sys/block/*d*/queue/rotational`)
+* `node_labels` - Comma-separated labels to be applied to the nodes in the cluster, e.g. 'cluster.ocs.openshift.io/openshift-storage="",node-role.kubernetes.io/infra=""', default - empty string
 
 #### UPGRADE
 
@@ -416,6 +417,8 @@ Configuration specific to external Ceph cluster
 * `external_cluster_details` - base64 encoded data of json output from exporter script
 * `rgw_secure` - boolean parameter which defines if external Ceph cluster RGW is secured using SSL
 * `rgw_cert_ca` - url pointing to CA certificate used to sign certificate for RGW with SSL
+* `use_rbd_namespace` - boolean parameter to use RBD namespace in pool
+* `rbd_namespace` - Name of RBD namespace to use in pool
 
 ##### login
 

conf/deployment/aws/rosa_hcp_1az_3w_m5.12x.yaml

@@ -26,3 +26,4 @@ ENV_DATA:
   ms_env_type: "staging"
   addon_name: "ocs-converged"
   persistent-monitoring: false
+  node_labels: cluster.ocs.openshift.io/openshift-storage=""

conf/deployment/aws/rosa_hcp_1az_6w_m5.2x.yaml

@@ -26,3 +26,4 @@ ENV_DATA:
   ms_env_type: "staging"
   addon_name: "ocs-converged"
   persistent-monitoring: false
+  node_labels: cluster.ocs.openshift.io/openshift-storage=""

conf/deployment/vsphere/ipi_1az_rhcos_vsan_3m_6w_intransit_encryption.yaml

@@ -0,0 +1,16 @@
+---
+# This config is suppose to work on most of DCs we have.
+DEPLOYMENT:
+  allow_lower_instance_requirements: false
+ENV_DATA:
+  platform: 'vsphere'
+  deployment_type: 'ipi'
+  worker_replicas: 6
+  master_replicas: 3
+  master_num_cpus: '16'
+  master_memory: '65536'
+  fio_storageutilization_min_mbps: 10.0
+  in_transit_encryption: true
+REPORTING:
+  polarion:
+    deployment_id: 'OCS-4690'

conf/ocsci/external_rhcs_with_rbd_namespace.yaml

@@ -0,0 +1,4 @@
+# Config file to use RBD namespace in a pool
+---
+EXTERNAL_MODE:
+  use_rbd_namespace: True

ocs_ci/deployment/acm.py

@@ -86,6 +86,8 @@ def __init__(self):
         self.dr_only_list = []
 
     def deploy(self):
+        # Download subctl binary in any case.
+        self.download_binary()
         if self.source == "upstream":
             self.deploy_upstream()
         elif self.source == "downstream":
@@ -94,7 +96,6 @@ def deploy(self):
             raise Exception(f"The Submariner source: {self.source} is not recognized")
 
     def deploy_upstream(self):
-        self.download_binary()
         self.submariner_configure_upstream()
 
     def deploy_downstream(self):

ocs_ci/deployment/helpers/external_cluster_helpers.py

@@ -7,18 +7,20 @@
 import logging
 import re
 import tempfile
+import uuid
 
 from ocs_ci.framework import config
 from ocs_ci.ocs import defaults, constants
 from ocs_ci.ocs.ocp import OCP
 from ocs_ci.ocs.exceptions import (
+    ExternalClusterCephfsMissing,
+    ExternalClusterCephSSHAuthDetailsMissing,
     ExternalClusterExporterRunFailed,
+    ExternalClusterRBDNamespaceCreationFailed,
     ExternalClusterRGWEndPointMissing,
     ExternalClusterRGWEndPointPortMissing,
-    ExternalClusterCephSSHAuthDetailsMissing,
-    ExternalClusterObjectStoreUserCreationFailed,
-    ExternalClusterCephfsMissing,
     ExternalClusterNodeRoleNotFound,
+    ExternalClusterObjectStoreUserCreationFailed,
 )
 from ocs_ci.ocs.resources import pod
 from ocs_ci.ocs.resources.csv import get_csv_name_start_with_prefix
@@ -133,6 +135,12 @@ def get_external_cluster_details(self):
             ceph_user = config.EXTERNAL_MODE["run_as_user"]
             params = f"{params} --run-as-user {ceph_user}"
 
+        if config.EXTERNAL_MODE.get("use_rbd_namespace"):
+            rbd_namespace = config.EXTERNAL_MODE.get(
+                "rbd_namespace"
+            ) or self.create_rbd_namespace(rbd=rbd_name)
+            params = f"{params} --rados-namespace {rbd_namespace}"
+
         out = self.run_exporter_script(params=params)
 
         # encode the exporter script output to base64
@@ -476,6 +484,30 @@ def is_rgw_user_exists(self, user):
         logger.debug(f"RGW users: {rgw_user_list}")
         return True if user in rgw_user_list else False
 
+    def create_rbd_namespace(self, rbd, namespace=None):
+        """
+        Create RBD namespace
+
+        Args:
+            rbd (str): RBD pool name where namespace has to create
+            namespace (str): Name of RBD namespace
+
+        Returns:
+            str: RBD Namepsace name
+
+        Raises:
+            ExternalClusterRBDNamespaceCreationFailed: In case fails to create RBD namespace
+
+        """
+        namespace = namespace or f"rbd_namespace_{uuid.uuid4().hex[:8]}"
+        logger.info(f"creating RBD namespace {namespace}")
+        cmd = f"rbd namespace create {rbd}/{namespace}"
+        retcode, out, err = self.rhcs_conn.exec_cmd(cmd)
+        if retcode != 0:
+            logger.error(f"Failed to create RBD namespace in {rbd}. Error: {err}")
+            raise ExternalClusterRBDNamespaceCreationFailed
+        return namespace
+
 
 def get_exporter_script_from_configmap():
     """

ocs_ci/deployment/helpers/odf_deployment_helpers.py

@@ -37,4 +37,7 @@ def get_required_csvs():
     if ocs_version >= version.VERSION_4_17:
         operators_4_17_additions = [defaults.CEPHCSI_OPERATOR]
         ocs_operator_names.extend(operators_4_17_additions)
+    if ocs_version >= version.VERSION_4_18:
+        operators_4_18_additions = [defaults.ODF_DEPENDENCIES]
+        ocs_operator_names.extend(operators_4_18_additions)
     return ocs_operator_names

ocs_ci/deployment/hosted_cluster.py

@@ -402,6 +402,9 @@ def deploy_ocp(
         if not config.ENV_DATA["platform"].lower() in HCI_PROVIDER_CLIENT_PLATFORMS:
             raise ProviderModeNotFoundException()
 
+        if not config.ENV_DATA.get("deploy_acm_hub_cluster", True):
+            deploy_acm_hub = False
+
         self.deploy_dependencies(
             deploy_acm_hub, deploy_cnv, deploy_metallb, download_hcp_binary
         )
@@ -917,7 +920,7 @@ def get_onboarding_key(self):
              str: onboarding token key
         """
         secret_ocp_obj = ocp.OCP(
-            kind=constants.SECRET, namespace=constants.OPENSHIFT_STORAGE_NAMESPACE
+            kind=constants.SECRET, namespace=config.ENV_DATA["cluster_namespace"]
         )
 
         key = (
@@ -1160,7 +1163,7 @@ def get_provider_address(self):
         """
         Get the provider address
         """
-        ocp = OCP(namespace=constants.OPENSHIFT_STORAGE_NAMESPACE)
+        ocp = OCP(namespace=config.ENV_DATA["cluster_namespace"])
         storage_provider_endpoint = ocp.exec_oc_cmd(
             (
                 "get storageclusters.ocs.openshift.io -o jsonpath={'.items[*].status.storageProviderEndpoint'}"
@@ -1210,7 +1213,7 @@ def storage_claim_exists_cephfs(self):
         else:
             ocp = OCP(
                 kind=constants.STORAGECLAIM,
-                namespace=constants.OPENSHIFT_STORAGE_NAMESPACE,
+                namespace=config.ENV_DATA["cluster_namespace"],
                 cluster_kubeconfig=self.cluster_kubeconfig,
             )
 
@@ -1297,7 +1300,7 @@ def storage_claim_exists_rbd(self):
         else:
             ocp = OCP(
                 kind=constants.STORAGECLAIM,
-                namespace=constants.OPENSHIFT_STORAGE_NAMESPACE,
+                namespace=config.ENV_DATA["cluster_namespace"],
                 cluster_kubeconfig=self.cluster_kubeconfig,
             )
 

ocs_ci/deployment/ibmcloud.py

@@ -19,6 +19,7 @@
     LeftoversExistError,
     VolumesExistError,
 )
+from ocs_ci.ocs.resources.backingstore import get_backingstore
 from ocs_ci.ocs.resources.pvc import (
     scale_down_pods_and_remove_pvcs,
 )
@@ -203,6 +204,7 @@ def destroy_cluster(self, log_level="DEBUG"):
         resource_group = self.get_resource_group()
         if resource_group:
             try:
+                self.delete_bucket()
                 scale_down_pods_and_remove_pvcs(self.DEFAULT_STORAGECLASS)
             except Exception as err:
                 logger.warning(
@@ -235,6 +237,25 @@ def destroy_cluster(self, log_level="DEBUG"):
             logger.info("Force cleaning up Service IDs and Account Policies leftovers")
             ibmcloud.cleanup_policies_and_service_ids(self.cluster_name)
 
+    def delete_bucket(self):
+        """
+        Deletes the COS bucket
+        """
+        api_key = config.AUTH["ibmcloud"]["api_key"]
+        service_instance_id = config.AUTH["ibmcloud"]["cos_instance_crn"]
+        endpoint_url = constants.IBM_COS_GEO_ENDPOINT_TEMPLATE.format(
+            config.ENV_DATA.get("region", "us-east").lower()
+        )
+        backingstore = get_backingstore()
+        bucket_name = backingstore["spec"]["ibmCos"]["targetBucket"]
+        logger.debug(f"bucket name from backingstore: {bucket_name}")
+        cos = ibmcloud.IBMCloudObjectStorage(
+            api_key=api_key,
+            service_instance_id=service_instance_id,
+            endpoint_url=endpoint_url,
+        )
+        cos.delete_bucket(bucket_name=bucket_name)
+
     def manually_create_iam_for_vpc(self):
         """
         Manually specify the IAM secrets for the cloud provider

ocs_ci/deployment/provider_client/storage_client_deployment.py

@@ -271,7 +271,7 @@ def provider_and_native_client_installation(
         if self.ocs_version >= version.VERSION_4_16:
             # Validate native client is created in openshift-storage namespace
             self.deployment.wait_for_csv(
-                self.ocs_client_operator, constants.OPENSHIFT_STORAGE_NAMESPACE
+                self.ocs_client_operator, config.ENV_DATA["cluster_namespace"]
             )
 
             # Verify native storageclient is created successfully

ocs_ci/deployment/rosa.py

@@ -100,6 +100,11 @@ def deploy(self, log_level=""):
             machinepool_details.wait_replicas_ready(
                 target_replicas=config.ENV_DATA["worker_replicas"], timeout=1200
             )
+            if node_labels := config.ENV_DATA.get("node_labels"):
+                if machinepool_id := config.ENV_DATA.get("machine_pool"):
+                    rosa.label_nodes(
+                        self.cluster_name, machinepool_id, node_labels, rewrite=False
+                    )
 
         logger.info("generate kubeconfig and kubeadmin-password files")
         if config.ENV_DATA["ms_env_type"] == "staging":

ocs_ci/framework/__init__.py

@@ -493,7 +493,7 @@ def __init__(self):
             except ClusterNotFoundException:
                 # if no provider is available then set the switch to current index so that
                 # no switch happens and code runs on current cluster
-                logger.DEBUG("No provider was found - using current cluster")
+                logger.debug("No provider was found - using current cluster")
                 switch_index = config.cur_index
             super().__init__(switch_index)
 
@@ -509,7 +509,7 @@ def __init__(self):
             except ClusterNotFoundException:
                 # if no provider is available then set the switch to current index so that
                 # no switch happens and code runs on current cluster
-                logger.DEBUG("No Consumer was found - using current cluster")
+                logger.debug("No Consumer was found - using current cluster")
                 switch_index = config.cur_index
             super().__init__(switch_index)
 

ocs_ci/framework/conf/default_config.yaml

@@ -279,6 +279,9 @@ ENV_DATA:
   #RDR Green field
   rdr_osd_deployment_mode: "greenfield"
 
+  # Label nodes with specific labels, used for example fot ODF deployment on ROSA HCP
+  node_labels: ""
+
   # Assisted Installer related settings
 
 # This section is related to upgrade
@@ -288,6 +291,10 @@ UPGRADE:
   ocp_upgrade_path: "quay.io/openshift-release-dev/ocp-release"
   ocp_arch: "x86_64"
   upgrade_logging_channel: "4.18"
+  # None value means that value in Rook operator config is used.
+  # Otherwise it is changed to the provided value before ODF upgrade.
+  csi_rbd_plugin_update_strategy_max_unavailable: null
+  csi_cephfs_plugin_update_strategy_max_unavailable: null
 
 # This section stores secret and uploaded from home dir or s3
 # for entry into this section, please email ecosystem team

ocs_ci/framework/pytest_customization/marks.py

@@ -384,6 +384,11 @@
     reason="Azure KV config required to run the test.",
 )
 
+rosa_hcp_required = pytest.mark.skipif(
+    config.ENV_DATA["platform"].lower() != ROSA_HCP_PLATFORM,
+    reason="Test runs ONLY on ROSA HCP cluster",
+)
+
 external_mode_required = pytest.mark.skipif(
     config.DEPLOYMENT.get("external_mode") is not True,
     reason="Test will run on External Mode cluster only",