Preparing storage

Prepare storage, including the required persistent volumes (PVs) and persistent volume claim (PVCs) for the operator, Application Engine, IBM Business Automation Navigator, IBM FileNet® Content Manager,  New in 20.0.3  Intelligent Task Prioritization,  New in 20.0.3  Workforce Insights, Java™ Message Service (JMS), IBM Process Federation Server, Elasticsearch, and IBM Business Automation Workflow.

Procedure

  1. Prepare storage for the operator.
    The operator requires a persistent volume (PV) and persistent volume claim (PVC) to store the Db2® JDBC driver. The following example illustrates the procedure using Network File System (NFS). An existing NFS server is required before you create PVs and PVCs.
    1. Create a directory for the operator on the NFS server.
      • If you are using Db2, run the following command on the NFS server: 
        mkdir -p NFS_storage_directory/db/jdbc/db2
chown -R :65534 NFS_storage_directory/db
chmod -R g+rw NFS_storage_directory/db
      •  New in 20.0.3  If you are using Oracle, run the following command on the NFS server: 
        mkdir -p NFS_storage_directory/db/jdbc/oracle
chown -R :65534 NFS_storage_directory/db
chmod -R g+rw NFS_storage_directory/db
      •  New in 20.0.3  If you are using PostgreSQL, run the following command on the NFS server: 
        mkdir -p NFS_storage_directory/db/jdbc/postgresql
chown -R :65534 NFS_storage_directory/db
chmod -R g+rw NFS_storage_directory/db
    2. Copy the JDBC JAR files to the new operator directory.
      • If you are using Db2, copy db2jcc4.jar and db2jcc_license_cu.jar to NFS_storage_directory/db/jdbc/db2.
      •  New in 20.0.3  If you are using Oracle, copy ojbdc8.jar to NFS_storage_directory/db/jdbc/oracle.
      •  New in 20.0.3  If you are using PostgreSQL, copy a PostgreSQL JDBC 4.2 driver JAR file to NFS_storage_directory/db/jdbc/postgresql.
    3. In the /etc/exports configuration file, add the following line at the end: 
      <NFS_STORAGE_DIRECTORY> *(rw,sync,no_subtree_check)
    4. After editing and saving the /etc/exports configuration file, you must restart the NFS service for the changes to take effect. The following example shows restarting the NFS server on RHEL 7: 
      systemctl stop nfs
systemctl stop rpcbind
systemctl start rpcbind
systemctl start nfs
    5. To create the shared PV and PVC that are required by the operator, create a file that is named operator-shared-pv-pvc.yaml with the following contents: 
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: operator-shared-pv
spec:
  storageClassName: "ecm-openshift-nfs"
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 1Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/db
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
---
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: operator-shared-pvc
  annotations:
    volume.beta.kubernetes.io/storage-class: "ecm-openshift-nfs"
  labels:
    app.kubernetes.io/instance: ibm-dba
    app.kubernetes.io/managed-by: ibm-dba
    app.kubernetes.io/name: ibm-dba
spec:
  accessModes:
    - ReadWriteMany
  resources:
    requests:
      storage: 1Gi
      where
      • <NFS_STORAGE_DIRECTORY> is the operator folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
    6. Run the following command on the OpenShift master node: 
      oc apply -f operator-shared-pv-pvc.yaml
    7.  New in 20.0.3  If you are using Oracle or PostgreSQL, modify the baw_configuration section of the custom resource template under the database data configuration to use the custom JDBC drivers, custom JDBC PVC name, and JDBC driver files.
  2. Prepare storage for Application Engine, by following the instructions in Implementing storage.
  3. Prepare storage for IBM Business Automation Navigator, by following the instructions in Creating volumes and folders for deployment on Kubernetes.
    Important: The value for the baw_configuration[x].case.network_shared_directory_pvc (the persistent volume claim (PVC) name for the case network shared directory) parameter in the custom resource must be set to the same value as the IBM Business Automation Navigator pvc_for_icn_pluginstore parameter.
  4. Prepare storage for IBM FileNet Content Manager, by following the instructions in Configuring storage for the content services environment.
  5.  New in 20.0.3  Prepare storage for Intelligent Task Prioritization.
    The Intelligent Task Prioritization feature requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting baml_configuration.intelligent_task_prioritization.storage.use_dynamic_provisioning to true. The trained model files will use shared_configuration.storage_configuration.sc_fast_file_storage_classname in the PV and the log file will use shared_configuration.storage_configuration.sc_medium_file_storage_classname in the PV.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting baml_configuration.intelligent_task_prioritization.storage.use_dynamic_provisioning to false. Then, create two PVs and PVCs manually and set baml_configuration.intelligent_task_prioritization.storage.existing_pvc_for_logstore to the name of the log PVC and baml_configuration.intelligent_task_prioritization.storage.existing_pvc_for_trained_pipelines to the name of the trained pipelines PVC in the custom resource file.

    The following example illustrates the Intelligent Task Prioritization PV and PVCs creation procedure that uses NFS. An existing NFS server is required before you create PVs and PVCs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands: 
      mkdir -p <NFS_STORAGE_DIRECTORY>/baml-itp/logstore
mkdir -p <NFS_STORAGE_DIRECTORY>/baml-itp/trained-pipelines

chown -R :65534 <NFS_STORAGE_DIRECTORY>/baml-itp/logstore
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baml-itp/logstore
chown -R :65534 <NFS_STORAGE_DIRECTORY>/baml-itp/trained-pipelines
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baml-itp/trained-pipelines
    2. Create the PVs required by Intelligent Task Prioritization by saving the following YAML files on the OpenShift master node and then running the oc apply -f <YAML_FILE_NAME> command on each of the files in the following order:
      1. baml-itp-logstore-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baml-itp-logstore-pv
spec:
  storageClassName: baml-itp-logstore-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/baml-itp/logstore
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
      2. baml-itp-logstore-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baml-itp-logstore-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baml-itp-logstore-pv
  volumeName: baml-itp-logstore-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
      3. baml-itp-trained-pipelines-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baml-itp-trained-pipelines-pv
spec:
  storageClassName: baml-itp-trained-pipelines-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/baml-itp/trained-pipelines
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
      4. baml-itp-trained-pipelines-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baml-itp-trained-pipelines-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baml-itp-trained-pipelines-pv
  volumeName: baml-itp-trained-pipelines-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  6.  New in 20.0.3  Prepare storage for Workforce Insights.
    The Workforce Insights feature requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting baml_configuration.workforce_insights.storage.use_dynamic_provisioning to true. The log file will use shared_configuration.storage_configuration.sc_medium_file_storage_classname in the PV.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting baml_configuration.workforce_insights.storage.use_dynamic_provisioning to false. Then, create two PVs and PVCs manually and set baml_configuration.workforce_insights.storage.existing_pvc_for_logstore to the name of the log PVC in the custom resource file.

    The following example illustrates the Workforce Insights PV and PVCs creation procedure that uses NFS. An existing NFS server is required before you create PVs and PVCs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands: 
      mkdir -p <NFS_STORAGE_DIRECTORY>/baml-wfi/logstore

chown -R :65534 <NFS_STORAGE_DIRECTORY>/baml-wfi/logstore
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baml-wfi/logstore
    2. Create the PVs required by Workforce Insights by saving the following YAML files on the OpenShift master node and then running the oc apply -f <YAML_FILE_NAME> command on each of the files in the following order:
      1. baml-wfi-logstore-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baml-wfi-logstore-pv
spec:
  storageClassName: baml-wfi-logstore-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/baml-wfi/logstore
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
      2. baml-wfi-logstore-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baml-wfi-logstore-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baml-wfi-logstore-pv
  volumeName: baml-wfi-logstore-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  7. Prepare storage for the Java Message Service (JMS).
    The JMS component requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting baw_configuration[x].jms.storage.use_dynamic_provisioning to true and provide the storage class name of baw_configuration[x].jms.storage.storage_class in the custom resource file.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting baw_configuration[x].jms.storage.use_dynamic_provisioning to false. Then, create a PV manually and set baw_configuration[x].jms.storage.storage_class in the custom resource file to the value of the storageClassName property of your PV.

    The following example illustrates the JMS PV creation procedure that uses NFS. An existing NFS server is required before you create PVs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands: 
      mkdir -p <NFS_STORAGE_DIRECTORY>/jms
chown -R :65534 <NFS_STORAGE_DIRECTORY>/jms
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/jms
    2. Create the PV required by JMS by creating a YAML file called jms-pv.yaml on the OpenShift master node with the following contents: 
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: jms-pv-baw
spec:
  storageClassName: "jms-storage-class"
  accessModes:
  - ReadWriteOnce
  capacity:
    storage: 2Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/jms
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • accessModes is set to the same value as the jms.storage.access_modes property in the custom resource configuration file
      • <NFS_SERVER_IP> is the IP address of your NFS server
    3. Run the following command: 
      oc apply -f jms-pv.yaml
  8. Prepare storage for Process Federation Server.
    The Process Federation Server component requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting pfs_configuration.logs.storage.use_dynamic_provisioning to true and provide the storage class name of pfs_configuration.logs.storage.storage_class in the custom resource file.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting pfs_configuration.logs.storage.use_dynamic_provisioning to false. Then, create a PV manually and set pfs_configuration.logs.storage.storage_class in the custom resource file to the value of the storageClassName property of your PV.

    The following example illustrates the Process Federation Server PV creation procedure that uses NFS. An existing NFS server is required before you create PVs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands:
       New in 20.0.3  
      mkdir -p <NFS_STORAGE_DIRECTORY>/pfs-logs

chown -R :65534 <NFS_STORAGE_DIRECTORY>/pfs-logs
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/pfs-logs
       For 20.0.2  
      mkdir -p <NFS_STORAGE_DIRECTORY>/pfs-logs-0

chown -R :65534 <NFS_STORAGE_DIRECTORY>/pfs-logs-0
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/pfs-logs-0
    2. Create the PV required by Process Federation Server by creating a YAML file on the OpenShift master node.

       New in 20.0.3  Create the pfs-pv-pvc-logs.yam file with the following content:

      apiVersion: v1
kind: PersistentVolume
metadata:
  name: pfs-logs-pv
spec:
  storageClassName: "pfs-logs"
  accessModes:
  - ReadWriteOnce
  capacity:
    storage: 5Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/pfs-logs
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: pfs-logs-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: pfs-logs
  volumeName: pfs-logs-pv
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
       For 20.0.2  The number of PVs to be created depends on your pfs_configuration.replicas setting. Create the same number of PVs as replicas. Create the pfs-pv-pfs-logs-0.yaml file with the following content:
      ---
apiVersion: v1
kind: PersistentVolume
metadata:
  name: pfs-logs-pv
spec:
  storageClassName: "pfs-logs"
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/pfs-logs-0
    server: <NFS_SERVER_IP>
    persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
    3. Run the following command:
       New in 20.0.3  
      oc apply -f pfs-pv-pvc-logs.yaml
       For 20.0.2  
      oc apply -f pfs-pv-pfs-logs-0.yaml
  9. Prepare storage for Elasticsearch.
    The Elasticsearch component on Process Federation Server requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting elasticsearch_configuration.storage.use_dynamic_provisioning to true and provide the storage class name of elasticsearch_configuration.storage.storage_class in the custom resource file.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting elasticsearch_configuration.storage.use_dynamic_provisioning to false. Then, create a PV manually and set elasticsearch_configuration.storage.storage_class in the custom resource file to the value of the storageClassName property of your PV.

    The following example illustrates the Elasticsearch PV creation procedure that uses NFS. An existing NFS server is required before you create PVs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands: 
      mkdir -p <NFS_STORAGE_DIRECTORY>/pfs-es-0

chown -R :65534 <NFS_STORAGE_DIRECTORY>/pfs-es-0
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/pfs-es-0
    2. Create the PV required by Elasticsearch by creating a YAML file called pfs-pv-pfs-es-0.yaml on the OpenShift master node with the following contents: 
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: pfs-es-0
spec:
  storageClassName: "pfs-es"
  accessModes:
  - ReadWriteOnce
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/pfs-es-0
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
      • The numbers of PVs to be created depends on your elasticsearch_configuration. replicas setting. Create the same number of PVs as replicas.
    3. Run the following command: 
      oc apply -f pfs-pv-pfs-es-0.yaml
  10. Prepare storage for Business Automation Workflow.
    The Business Automation Workflow component requires a PV to be created before you can deploy. You have the following options, depending on whether your Kubernetes environment supports dynamic provisioning.
    • Option 1: If your environment supports dynamic provisioning:

      Enable dynamic provisioning by setting baw_configuration.storage.use_dynamic_provisioning to true. The dump file will use shared_configuration.storage_configuration.sc_slow_file_storage_classname in the PV and the log file will use shared_configuration.storage_configuration.sc_medium_file_storage_classname in the PV.

    • Option 2: If your environment does not support dynamic provisioning:

      Disable dynamic provisioning by setting baw_configuration.storage.use_dynamic_provisioning to false. Then, create the required PVs and PVCs manually and set baw_configuration.storage.existing_pvc_for_logstore to the name of the log PVC and baw_configuration.storage.existing_pvc_for_dumpstore to the name of the dump PVC in the custom resource file.  New in 20.0.3  Also set baw_configuration.storage.existing_pvc_for_filestore to the name of the file PVC.

    The following example illustrates the Business Automation Workflow PV and PVC creation procedure that uses NFS. An existing NFS server is required before you create PVs and PVCs.
    1. Create related folders on an NFS server. You must grant minimal privileges to the NFS server. Give the least privilege to the mounted directories using the following commands: 
      mkdir -p <NFS_STORAGE_DIRECTORY>/baw/logstore
mkdir -p <NFS_STORAGE_DIRECTORY>/baw/dumpstore
 New in 20.0.3  mkdir -p <NFS_STORAGE_DIRECTORY>/baw/filestore

chown -R :65534 <NFS_STORAGE_DIRECTORY>/baw/logstore
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baw/logstore
chown -R :65534 <NFS_STORAGE_DIRECTORY>/baw/dumpstore
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baw/dumpstore
 New in 20.0.3  
chown -R :65534 <NFS_STORAGE_DIRECTORY>/baw/filestore
 New in 20.0.3  
chmod -R g+rw <NFS_STORAGE_DIRECTORY>/baw/filestore
    2. Create the Business Automation Workflow required PVs by saving the following YAML files on the OpenShift master node and then running the oc apply -f <YAML_FILE_NAME> command on the files in the following order:
      1. baw-logstore-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baw-logstore-pv
spec:
  storageClassName: baw-logstore-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/baw/logstore
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      where
      • <NFS_STORAGE_DIRECTORY> is the storage folder on your NFS server
      • <NFS_SERVER_IP> is the IP address of your NFS server
      2. baw-logstore-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baw-logstore-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baw-logstore-pv
  volumeName: baw-logstore-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
      3. baw-dumpstore-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baw-dumpstore-pv
spec:
  storageClassName: baw-dumpstore-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: <NFS_STORAGE_DIRECTORY>/baw/dumpstore
    server: <NFS_SERVER_IP>
  persistentVolumeReclaimPolicy: Recycle
      4. baw-dumpstore-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baw-dumpstore-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baw-dumpstore-pv
  volumeName: baw-dumpstore-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
       New in 20.0.3  5. baw-filestore-pv.yaml
      apiVersion: v1
kind: PersistentVolume
metadata:
  name: baw-filestore-pv
spec:
  storageClassName: baw-filestore-pv
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi
  nfs:
    path: NFS_storage_directory/baw/filestore
    server: NFS_server_IP
  persistentVolumeReclaimPolicy: Recycle
       New in 20.0.3  6. baw-filestore-pvc.yaml
      apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  name: baw-filestore-pvc
spec:
  accessModes:
  - ReadWriteMany
  resources:
    requests:
      storage: 10Gi
  storageClassName: baw-filestore-pv
  volumeName: baw-filestore-pv
status:
  accessModes:
  - ReadWriteMany
  capacity:
    storage: 10Gi