Known issues for the CDP Private Cloud Data Services 1.5.4

Known Issues in Management Console 1.5.4

OPSX-5147: OOM when retrieving size of Binary File

Sometimes, diagnostics bundle collection fails to complete due to OOM issues.

Limit the time range for the diagnostics bundle.

OPSX-5148: Diagnostics Collection from UI w/ Default No Time Limit Should Not Invoke Timestamp Filtering

When the diagnostics collection is triggered through the UI, by default, "No Time Limit" is selected. Filtering of logs by timestamp is still observed.

No workaround available.

DOCS-20088/OPSX-4781: Vault pods may take long time to be ready during upgrades from 1.5.2 to 1.5.3

The 'vault-0' pod takes longer time to attach volume in some upgrade cases than usual. Due to the excess time taken the cluster upgrade may fail. But, usually in 15 minutes the volume can attach automatically and the pod would start running. In that case, the user can resume the upgrade.

No workaround available.

OPSX-5155: OS Upgrade | Pods are not starting after the OS upgrade from RHEL 8.6 to 8.8

After an OS upgrade and start of the ECS service, pods fail to come up due to stale state.

Restart the ECS cluster.

OPSX-5055: ECS upgrade failed at Unseal Vault step

During an ECS upgrade from 1.5.2 to 1.5.4 release, the vault pod fails to start due to an error caused by the Longhorn volume unable to attach to the host. The error is as below:

Warning FailedAttachVolume 3m16s (x166 over 5h26m) attachdetach-controller AttachVolume.Attach failed for volume "pvc-0ba86385-9064-4ef9-9019-71976b4902a5" : rpc error: code = Internal desc = volume pvc-0ba86385-9064-4ef9-9019-71976b4902a5 failed to attach to node host-1.cloudera.com with attachmentID csi-7659ab0e6655d308d2316536269de47b4e66062539f135bf6012bfc8b41fc345: the volume is currently attached to different node host-2.cloudera.com

Follow below steps provided by SUSE to ensure the Longhorn volume is correctly attached to the node where the vault pod is running.

# Find out the volume name that is failing to attach to the vault pod. 
For e.g. pvc-bc73e7d3-c7e7-468a-b8e0-afdb8033e40b from the pod logs.
kubectl edit volumeattachments.longhorn.io -n longhorn-system pvc-bc73e7d3-c7e7-468a-b8e0-afdb8033e40b

# Update the "spec:" section of the volumeattachment and replace 
attachmentTickets section with {} as shown below and save.
spec:
 attachmentTickets: {}
 volume: pvc-bc73e7d3-c7e7-468a-b8e0-afdb8033e40b

# scale down the vault statefulset to 0 and scale it back up.
kubectl scale sts vault --replicas=0 -n vault-system
kubectl scale sts vault --replicas=1 -n vault-system

OPSX-4308: Display error in UI if listEnvironments failed

On the Environments page, if the listEnvironments API call fails, the error is hidden, and instead no environments are displayed, even though they do exist. This can be due to vault issues or connectivity issues.

No workaround available but the register environment page shows the error.

OPSX-4684: Start ECS command shows green(finished) even though start docker server failed on one of the hosts

The Docker service starts, but one or more Docker roles fail to start because the corresponding host is unhealthy.

Ensure the host is healthy. Start the the Docker role on the host.

OPSX-735: Kerberos service should handle Cloudera Manager downtime

The Cloudera Manager Server in the base cluster operates to generate Kerberos principals for Private Cloud. If there is downtime, you may observe Kerberos-related errors.

Resolve downtime on Cloudera Manager. If you encounter Kerberos errors, you can retry the operation (such as retrying creation of the Virtual Warehouse).

Known Issues in Management Console 1.5.3

OPSX-4754 [ECS Restart Stability] DaemonSet rollout process is stuck post rolling restart where DaemonSet kube-system/rke2-canal has not finished or progressed for at least 15 minutes

On RHEL 9.x, an ECS service DaemonSet rollout health alert appears in the Cloudera Manager after an ECS installation and a rolling restart.

To fix the DaemonSet rollout issue:

1. Edit the DaemonSet rke2-canal configuration file by running the following command:

   KUBECTL -n kube-system edit ds/rke2-canal

   Change the value of felixIptablesBackend from auto to Legacy and save the DaemonSet rke2-canal configuration file.

2. Reboot each node one-by-one.

3. Check to see if any of the nodes are cordoned off. If so, uncordon them:

   [root@host-1 ~]# $KUBECTL get nodes
                     NAME STATUS ROLES AGE VERSION
                     host-1.ecs-restart1.kcloud.cloudera.com Ready,SchedulingDisabled control-plane,etcd,master 17h v1.26.10+rke2r1
                     host-2.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     host-3.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     host-4.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     [root@host-1 ~]# $KUBECTL uncordon host-1.ecs-restart1.kcloud.cloudera.com
                     node/host-1.ecs-restart1.kcloud.cloudera.com uncordoned
                     [root@host-1 ~]# $KUBECTL get nodes
                     NAME STATUS ROLES AGE VERSION
                     host-1.ecs-restart1.kcloud.cloudera.com Ready control-plane,etcd,master 17h v1.26.10+rke2r1
                     host-2.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     host-3.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     host-4.ecs-restart1.kcloud.cloudera.com Ready <none> 17h v1.26.10+rke2r1
                     [root@host-1 ~]#

4. Ensure that the Vault is unsealed. , To unseal the vault in Cloudera Manager navigate to Clusters > ECS > <***ECS SERVICES***> such as ECS-1 or ECS-2 > Actions > Unseal Vault.
5. Wait for five to six minutes.

6. Check for longhorn pods that fail to come up on any of the hosts:

   [root@host-1 ~]# kubectl -o wide get pods -n longhorn-system | grep -v "Running" | grep -v "Completed"
   NAMESPACE                                NAME                                                              READY   
   STATUS                  RESTARTS         AGE     IP             NODE
   NOMINATED NODE   READINESS GATES
   longhorn-system                          longhorn-csi-plugin-frwnw                                         2/3     
   CrashLoopBackOff        14 (3m51s ago)   6h20m   10.x.x.x     host-1.upgr-ecs-ext.kcloud.cloudera.com    <none>
             <none>
   longhorn-system                          longhorn-manager-lgzmb                                            0/1     
   CrashLoopBackOff        7 (97s ago)      6h24m   10.x.x.x     host-1.upgr-ecs-ext.kcloud.cloudera.com    <none>
             <none> 

7. Reboot the host (In this case the host is: host-1.upgr-ecs-ext.kcloud.cloudera.com).
8. Wait for 15-30 minutes for pods to come up.
9. Post ECS reboot, if you notice buildkit pods in the following CrashLoopBackOff state, then delete those buildkit pods:

   [root@host-1 ~]# kubectl -o wide get pods | grep -v "Running" | grep -v "Completed"
   NAMESPACE                                NAME                                                              READY   STATUS                  RESTARTS         AGE     IP             NODE                                       NOMINATED NODE   READINESS GATES
   quasar-skl2-host-1                          buildkit-2jdmw                                         2/3     CrashLoopBackOff        14 (3m51s ago)   6h20m   10.x.x.x     host-1.upgr-ecs-ext.kcloud.cloudera.com    <none>           <none>
   quasar-skl2-host-1                         buildkit-k20smc                                            0/1     CrashLoopBackOff        7 (97s ago)      6h24m   10.x.x.x     host-2.upgr-ecs-ext.kcloud.cloudera.com    <none>           <none>  

   You can delete the above buildkit pods by one of the following ways:

   • On the Cloudera Manager UI, navigate to Clusters > ECS > <***ECS SERVICES***> such as ECS-1 or ECS-2 > Web UI > ECS Web UI > Delete.

   • Run the following command to delete all such buildkit pods:

     [root@host-1 ~]# kubectl delete pod buildkit-2jdmw -n quasar-skl2-host-1

     Wait for the buildkit pods to start back up.

OPSAPS-69892: kube-proxy failure causing issues with cluster

After rebooting/restarting an ECS agent node, the kube-proxy Linux process may not start due to a race condition in the kubelet. When this happens, ECS cluster networking and other services – such as Vault, DNS, authentication, Longhorn storage, etc. – are affected. At the Kubernetes pod level, errors such as "connection refused", "connection timed out" and "i/o timeout" may be observed. If you suspect possible networking issues in your ECS cluster, checking kube-proxy is a good first step.

To fix this issue, perform the following steps on all of the affected nodes:

1. To identify which agent needs to be restarted, check the status of each kube-proxy pod to make sure it is in the "ready" state by running the following command on each host in the cluster.

   kubectl describe pod [***POD-NAME***] -n kube-system

   Here, [***POD-NAME***] should have a format such as: kube-proxy-<hostname>.

   In the Conditions section of the describe pod output, confirm that the "ready" condition is "True".

   Conditions:
     Type              Status
     Initialized       True 
     Ready             True 
     ContainersReady   True 
     PodScheduled      True 

   Another option is to run the following command:

   kubectl get pods -n kube-system -l component=kube-proxy -o go-template='{{range .items}}
   {{.metadata.name}}{{"\n"}}{{"  "}}{{range .status.conditions}}{{ if eq .type "Ready" }}
   Ready:{{.status}}{{"\n\n"}}{{end}}{{end}}{{end}}'

   The sample output displays the status of all of the kube-proxy pods in the cluster:

   
   kube-proxy-host-1.cloudera.com
     Ready:True
   
   kube-proxy-host-2.cloudera.com
     Ready:True
   
   kube-proxy-host-3.cloudera.com
     Ready:True

2. If the "ready" state is False, kube-proxy is not functioning properly, regardless of whether the kube-proxy process is running on that host or not. On each of the affected nodes, run the following command to delete the kube-proxy pod manifest:

    rm /var/lib/rancher/rke2/agent/pod-manifests/kube-proxy.yaml

3. Start the agent role.

   After the agent role is started, you may not immediately see the kube-proxy process running, but a new kube-proxy process should start shortly. Check the pod status to make sure it is ready. After all of the problem agents have been restarted, the cluster may complain that the vault is sealed – if so, unseal it. At this point, the Control Plane should be functioning properly.

Additional details about this issue are available here: https://www.suse.com/support/kb/doc/?id=000021284

Known Issues in Management Console 1.5.2

OPSAPS-68923: CM - After CM upgrade from 7.9.5 to 7.11.3.x ECS cluster showing stale config

After Cloudera Manager upgrade from 7.9.5 to 7.11.3.x, an ECS 1.5.0 cluster may show a stale config to add ""limit_fds": 1048576"

This can be ignored – no restart of the ECS cluster is necessary. When the ECS 1.5.0 cluster is upgraded to 1.5.2, the stale config will be resolved.

OPSX-4594: [ECS Restart Stability] Post rolling restart few volumes are in detached state (vault being one of them)

After rolling restart there may be some volumes in detached state.

1. Open the Longhorn UI to view the detached volumes.
2. Perform the following operations for each volume in a detached state:
   1. Identify the workload name and type from the volume details.
   2. Identify the workload and number of replicas using kubectl or the Kubernetes UI.
   3. Scale the workload down to 0.
   4. Wait for the pods associated with the workload to fully terminate.
   5. Scale up the workload up to the number of replicas it had originally.

To prevent this issue, use the Longhorn UI to set the number of replicas for the volume to at least 3.

OPSAPS-68558: [7.9.5->7.11.3.2] CM upgrade failed with BeanCreationException: Error creating bean with name 'com.cloudera.server.cmf.TrialState'

After upgrading the Cloudera Manager package, the Cloudera Manager Server does not start. An error about "Active Commands" is shown in the Cloudera Manager Server log.

This may happen when the Private Cloud Data Services Control Plane is actively issuing requests to Cloudera Manager while an upgrade is being performed.

Before upgrading Cloudera Manager make sure there are no active commands. If there are any active commands, wait for them to complete before starting a Cloudera Manager upgrade.

If Cloudera Manager restart fails after upgrade due to an active getClientConfig command, check the Cloudera Manager server log for a "There are 1 active commands of type GetClientConfigFiles" error. This may block a Cloudera Manager restart after upgrade. Use the following steps to resolve this issue:

1. Login to Cloudera Manager database.
2. Search for any active GetClientConfigFiles command in the COMMANDS table.

   UPDATE COMMANDS SET active=0,success=false,state='CANCELLED' where command_id=<command_id>;  

3. Delete these entries, including foreign key dependencies, in the following tables:
   • PROCESSES
   • PROCESSES_DETAIL
   • COMMANDS_DETAIL

   cm=> DELETE FROM COMMANDS where command_id=1546340765;
   ERROR:  update or delete on table "commands" violates foreign key constraint "fk_process_command" on table "processes"
   DETAIL:  Key (command_id)=(1546340765) is still referenced from table "processes".
   cm=>
   cm=> DELETE FROM processes where command_id=1546340765;
   ERROR:  update or delete on table "processes" violates foreign key constraint "fk_processes_detail_process" on table "processes_detail"
   DETAIL:  Key (process_id)=(1546340766) is still referenced from table "processes_detail".
   cm=>
   cm=>                                        ^
   cm=> DELETE FROM processes_detail where process_id=1546340766;
   DELETE 1
   cm=> DELETE FROM processes where command_id=1546340765;
   DELETE 1
   cm=> DELETE FROM COMMANDS where command_id=1546340765;
   ERROR:  update or delete on table "commands" violates foreign key constraint "fk_commands_detail_command" on table "commands_detail"
   DETAIL:  Key (command_id)=(1546340765) is still referenced from table "commands_detail".
   cm=>
   cm=> DELETE FROM commands_detail where command_id=1546340765;
   DELETE 1
   cm=> DELETE FROM COMMANDS where command_id=1546340765;
   DELETE 1 

4. Restart the Cloudera Manager server.

OPSX-4392: Getting the real client IP address in the application

CML has a feature for adding the audit event for each user action (Monitoring User Events). In Private Cloud, instead of the client IP, we are getting the internal IP, which is logged into the internal DB.

In ECS, add the enable-real-ip configuration as true for the nginx ingress controller:

apiVersion: v1
data:
  allow-snippet-annotations: "true"
  enable-real-ip: "true"                   <<<<<<<<<< new config
kind: ConfigMap
metadata:
  annotations:
    meta.helm.sh/release-name: rke2-ingress-nginx
    meta.helm.sh/release-namespace: kube-system
  creationTimestamp: "2023-05-09T04:54:53Z"
  labels:
    app.kubernetes.io/component: controller
    app.kubernetes.io/instance: rke2-ingress-nginx
    app.kubernetes.io/managed-by: Helm
    app.kubernetes.io/name: rke2-ingress-nginx
    app.kubernetes.io/part-of: rke2-ingress-nginx
    app.kubernetes.io/version: 1.6.4
    helm.sh/chart: rke2-ingress-nginx-4.5.201
  name: rke2-ingress-nginx-controller
  namespace: kube-system
  resourceVersion: "162559439"
  uid: cca67b0c-bc05-4e1f-8439-7d44323f4624

In OCP, you may be able configure this using HAproxy with X-forward-for pass to OpenShift 4.

PULSE-944 and PULSE-941 Observability namespace not created after platform upgrade from 151 to 152

The Cloudera Observability namespace is not created after a platform upgrade from PvC DS 1.5.1 to PvC DS 1.5.2.

During the creation of the resource pool the Cloudera Observability namespace is provided by the CDP Private Cloud Service. If the provisioning flow is not completed, such as due to a timing difference between the start of the computeAPI pod and the call to the computeAPI pod by the service, the namespace is not created.

Trigger the Cloudera Observability namespace deployment by restarting the pvcservice pod.

PULSE-921 Observability namespace has no pods

The Cloudera Observability namespace should have the same number of pods and nodes. When the Cloudera Observability namespace has no pods the prometheus-node-exporter-1.6.0 helm release state becomes invalid and the CDP Private Cloud Service is unable to uninstall and reinstall the namespace. Also, as the Node Exporter is not installed into the Cloudera Observability namespace its metrics are unavailable when querying Prometheus in the control plane, for example the node_cpu_seconds_total metric.

Manually uninstall the invalid helm release with the --debug flag, verify that there are no helm releases listed by running -n observability -a, and then trigger the deployment process by restarting the pvcservice pod in the control plane.

PULSE-697 Add node-exporter to PvC DS

When expanding a cluster with new nodes and there is insufficient CPU and memory resources, the Node Exporter will encounter difficulties deploying new pods on the additional nodes.

To ensure sufficient resource allocation, such as when the Cloudera Observability namespace requires adjustment, delete the existing namespace and restart the pvcservice pod. This automatically initiates the creation of the Cloudera Observability namespace with the appropriate resource allocation.

PULSE-935 Longhorn volumes are over 90% of the capacity alerts on Prometheus volumes

Cloudera Manager displays the following alert about your Prometheus volumes: Concerning: Firing alerts for Longhorn: The actual used space of Longhorn volume is over 90% of the capacity.

Longhorn stores historical data as snapshots that are calculated with the active data for the volume's actual size. This size is therefore greater than the volume's nominal data value.

When the alert is displayed on the Cloudera Manager UI and it is related to Longhorn volumes used by Prometheus, ignore. For more information, see the Longhorn space consumption guidelines in the Longhorn documentation.

PULSE-937 Private-Key field change in Update Remote Write request does not reflect in enabling the metric flow

When using the Management Console UI for Remote Storage the Disable option does not deactivate the remote write configuration, even when the action returns a positive result message. Therefore, when disabling a remote storage configuration use the CLI client to disable the remote storage configuration directly from the API.

At this time when a remote storage configuration is incorrect, do not use the Edit or Disable option from the configuration’s Actions menu (ellipsis icon) to change its configuration. Instead, delete the remote storage's configuration from the configuration’s Actions menu with the Remove Configuration action and then re-create the remote write configuration with the Delete and Create operations of the API, using the CLI client.

PULSE-841 Disabling the remote write configuration logs an error in both cp prometheus and env prometheus

When a metric replication is set up between the cluster and Cloudera Observability and the connection is disabled or deleted, Prometheus writes an error message that states that it cannot replicate the metrics.

No workaround is required. After a few minutes the errors are no longer logged and Prometheus no longer tries to replicate the metrics.

PULSE-895 Disabling the remote write config in the UI is broken in cdp-pvc

The Remote Write Enable and Disable options in the Management Console’s User Interface do not work when a Remote Storage configuration is created with a requestSignerAuth type from either the HTTP API or using the CDP-CLI tool.

At this time, do not use the Enable or Disable options from the Remote Storage configuration’s Actions menu in the Management Console’s UI. Instead, enable or disable the configuration from the HTTP API or using the CDP-CLI tool.

PULSE-936 No Alert to prompt the metric flow being affected b/c of wrong private key configuration

A remote write alert was not triggered when the wrong private key was used in a Remote Storage configuration.

No workaround. Incorrect configuration settings, such as in this case where a bad private key was used, may block the forwarding of metrics. When creating a Remote Storage configuration you must carefully verify each configuration setting.

Known Issues in Management Console 1.5.1

External metadata databases are no longer supported on OCP

As of CDP Private Cloud Data Services 1.5.1, external Control Plane metadata databases are no longer supported. New installs require the use of an embedded Control Plane database. Upgrades from CDP Private Cloud Data Services 1.4.1 or 1.5.0 to 1.5.1 are supported, but there is currently no migration path from a previous external Control Plane database to the embedded Control Plane database. Upgrades from 1.4.0 or 1.5.0 with external Control Plane metadata databases also require additional steps, which are described in the CDP Private Cloud Data Services 1.5.1 upgrade topics.

DOCS-15855: Networking API is deprecated after upgrade to CDP Private Cloud Data Services 1.5.1 (K8s 1.24)

When the control plane is upgraded from 1.4.1 to 1.5.1, the Kubernates version changes to 1.24. The Livy pods running in existing Virtual Clusters (VCs) use a deprecated networking API for creating ingress for Spark driver pods. Because the old networking API is deprecated and does not exist in Kubernates 1.24, any new job run will not work for the existing VCs.

CDPQE-24295: Update docker client on docker.lab.eng.hortonworks machine

When you attempt to execute the Docker command to fetch the Cloudera-provided images into your air-gapped environment, you may encounter an issue where Docker pulls an incorrect version of the HAProxy image, especially if you are using an outdated Docker client. This situation arises due to the Cloudera registry containing images with multiple platform versions. Unfortunately, older Docker clients may lack the capability to retrieve the appropriate architecture version, such as amd64.

Update the Docker client. It has been demonstrated that Docker 20.10.5 and later versions have been successful in resolving this problem.

OPSX-4266: ECS upgrade from 1.5.0 to 1.5.1 is failing in Cadence schema update job

When upgrading from ECS 1.5.0 to 1.5.1, the CONTROL_PLANE_CANARY fails with the following error:

Firing alerts for Control Plane: Job did not complete in time, Job failed to complete. 

And the cdp-release-cdp-cadence-schema-update job fails.

Use the following steps to manually execute the job:

1. Export the job manifest into a file:

   kubectl get job cdp-release-cdp-cadence-schema-update -n <cdp> -o yaml > job.yaml

2. Delete the cdp-release-cdp-cadence-schema-update job:

   kubectl delete job cdp-release-cdp-cadence-schema-update -n <cdp>

3. Remove runtime information from the manifest, such as:

   resourceVersion
   uid
   selector
     matchLabels
       controller-uid
   labels
     controller-uid
   status section

4. Create the job:

   kubectl apply -f job.yaml

OPSX-4076:

When you delete an environment after the backup event, the restore operation for the backup does not bring up the environment.

Create the environment manually.

OPSX-4024: CM truststore import into unified truststore should handle duplicate CommonNames

If multiple CA certificates with the exact same value for the Common Name field are present in the Cloudera Manager truststore when a Private Cloud Data Services cluster is installed, only one of them may be imported into the Data Services truststore. This may cause certificate errors if an incorrect/old certificate is imported.

Remove old certificates from the Cloudera Manager truststore, and ensure certificates have unique Common Names.

COMOPS-2822: OCP error x509: certificate signed by unknown authority

The error x509: certificate signed by unknown authority usually means that the Docker daemon that is used by Kubernetes on the managed cluster does not trust the self-signed certificate.

Usually the fix is to copy the certificate to the path below on all of the worker nodes in the cluster:

/etc/docker/certs.d/<your_registry_host_name>:<your_registry_host_port>/ca.crt

OPSX-3073 [ECS] First run command failed at setup storage step with error "Timed out waiting for local path storage to come up"

Pod stuck in pending state on host for a long time. Error in Role log related to CNI plugin:

Events:

   Type     Reason                  Age                   From     Message
  ----     ------                  ----                  ----     -------
  Warning  FailedCreatePodSandBox  3m5s (x269 over 61m)  kubelet  (combined from similar events): 
Failed to create pod sandbox: rpc error: code = Unknown desc = failed to setup network for sandbox 
"70427e9b26fb014750dfe4441fdfae96cb4d73e3256ff5673217602d503e806f": 
failed to find plugin "calico" in path [/opt/cni/bin] 

Delete the cni directory on the host failing to launch pods:

ssh root@ecs-ha1-p-7.vpc.cloudera.com rm -rf /var/lib/cni

Restart the canal pod running on that host:

kubectl get pods -n kube-system -o wide | grep  ecs-ha1-p-7.vpc.cloudera.com
kube-proxy-ecs-ha1-p-7.vpc.cloudera.com                 1/1     Running     0          11h   10.65.52.51    ecs-ha1-p-7.vpc.cloudera.com   <none>           <none>
rke2-canal-llkc9                                        2/2     Running     0          11h   10.65.52.51    ecs-ha1-p-7.vpc.cloudera.com   <none>           <none>
rke2-ingress-nginx-controller-dqtz8                     1/1     Running     0          11h   10.65.52.51    ecs-ha1-p-7.vpc.cloudera.com   <none>           <none>
kubectl delete pod rke2-canal-llkc9 -n kube-system

Known Issues in Management Console 1.5.0

Somehow the Rebuilding field inside volume.meta is set to true causing the volume to get stuck in attaching/detaching loop

This is a condition that can occur in ECS Longhorn storage.

Since the volume has only 1 replica in this case, we can:

1. Scale down the workload. The Longhorn volume will be detached.

2. Wait for the Longhorn volume to be detached.

3. SSH into the node that has the replica.

4. cd into the replica folder (for example, /longhorn/replicas/pvc-126d40e2-7bff-4679-a310-e444e84df267-1a5dc941).

5. Change the"Rebuilding" field from true to false in the volume.meta file.

6. Scale up the workload to attach the volume.

Known Issues in Management Console identified before 1.5.0

INSIGHT-2469: COE Insight from case 922848: Not able to connect to bit bucket

After installing CML on an ECS cluster, users were not able to connect the internal bitbucket repo.

Workaround:

In this case the MTU of the ECS virtual network interfaces were larger than that of host external interface, which may cause the network requests from ECS containers to get truncated.

The Container Network Interface (CNI) is a framework for dynamically configuring networking resources. CNI integrates smoothly with Kubenetes to enable the use of an overlay or underlay network to automatically configure the network between pods. Cloudera ECS uses Calico as the CNI network provider.

The MTU of the pods’ virtual network interface can be seen by running the ifconfig command.

The default MTU of the virtual network interfaces is 1450.

The MTU setting of the virtual interfaces is stored as a configmap in the kube-system namespace. To modify the MTU, edit the rke2-canal-config configmap.

$ /var/lib/rancher/rke2/bin/kubectl --kubeconfig 
 /etc/rancher/rke2/rke2.yaml --namespace kube-system 
 edit cm rke2-canal-config

Find the veth_mtu parameter in the YAML content. Modify the default value of 1450 to the required MTU size.

Next, restart the rke2-canal pods from the kube-system namespace. There will be rke2-canal pods for each ECS node.

After the pods are restarted, all subsequent new pods will use the new MTU setting. However, existing pods that are already running will remain on the old MTU setting. Restart all of the pods to apply the new MTU setting.

OPSX-2484: FileAlreadyExistsException during timestamp filtering

The timestamp filtering may result in FileAlreadyExistsException when there is a file with same name already existing in the tmp directory.

None

OPSX-2772: For Account Administrator user, update roles functionality should be disabled

An Account Administrator user holds the biggest set of privileges, and is not allowed to modify via current UI, even user try to modify permissions system doesn't support changing for account administrator.

CDP Private Cloud Data Services ECS Installation: Failed to perform First Run of services.

If an issue is encountered during the Install Control Plane step of Containerized Cluster First Run, installation will be re-attempted infinitely rather than the command failing.

Since the control plane is installed and uninstalled in a continuous cycle, it is often possible to address the cause of the failure while the command is still running, at which point the next attempted installation should succeed. If this is not successful, abort the First Run command, delete the Containerized Cluster, address the cause of the failure, and retry from the beginning of the Add Cluster wizard. Any nodes that are re-used must be cleaned before re-attempting installation.

Environment creation through the CDP CLI fails when the base cluster includes Ozone

Problem: Attempt to create an environment using the CDP command-line interface fails in a CDP Private Cloud Data Services deployment when the Private Cloud Base cluster is in a degraded state and includes Ozone service.

Workaround: Stopping the Ozone service temporarily in the Private Cloud Base cluster during environment creation prevents the control plane from using Ozone as a logging destination, and avoids this issue.

Filtering the diagnostic data by time range might result in a FileAlreadyExistsException

Problem:Filtering the collected diagnostic data might result in a FileAlreadyExistsException if the /tmp directory already contains a file by that name.

There is currently no workaround for this issue.

Kerberos service does not always handle Cloudera Manager downtime

Problem: The Cloudera Manager Server in the base cluster must be running to generate Kerberos principals for CDP Private Cloud. If there is downtime, you might observe Kerberos-related errors.

Resolve downtime issues on Cloudera Manager. If you encounter Kerberos errors, you can retry the concerned operation such as creating Virtual Warehouses.

Updating user roles for the admin user does not update privileges

In the Management Console, changing roles on the User Management page does not change privileges of the admin user.

None

Upgrade applies values that cannot be patched

If the size of a persistent volume claim in a Containerized Cluster is manually modified, subsequent upgrades of the cluster will fail.

None