InfoScale™ 9.0 Agent for Oracle Installation and Configuration Guide - Linux

Last Published:
Product(s): InfoScale & Storage Foundation (9.0)
Platform: Linux
  1. Introducing the Cluster Server agent for Oracle
    1.  
      About the Cluster Server agent for Oracle
    2.  
      About the agent for Oracle ASM
    3.  
      Supported software for VCS agent for Oracle
    4. How the agent makes Oracle highly available
      1.  
        How the Oracle and Netlsnr agents support intelligent resource monitoring
      2.  
        How the agent makes Oracle ASM instance highly available
      3.  
        How the agent supports Oracle container databases (CDBs) and pluggable databases (PDBs)
    5. About Cluster Server agent functions for Oracle
      1. Oracle agent functions
        1.  
          Startup and shutdown options for the Oracle agent
        2.  
          Monitor options for the Oracle agent in traditional database and container database
        3.  
          Startup and shutdown options for the pluggable database (PDB)
        4.  
          Monitor for the pluggable database
        5.  
          Recommended startup modes for pluggable database (PDB) based on container database (CDB) startup modes
        6.  
          How the agent handles Oracle error codes during detail monitoring
        7.  
          Info entry point for Cluster Server agent for Oracle
        8.  
          Action entry point for Cluster Server agent for Oracle
        9. How the Oracle agent supports health check monitoring
          1.  
            Generating the executable required for Oracle health check monitoring
          2.  
            Prerequisites for linking Oracle libraries using the build_oraapi.sh script
          3.  
            Considerations for running build_oraapi.sh
          4.  
            Building Oracle agent executable using build_oraapi.sh
      2.  
        Netlsnr agent functions
      3. ASMInst agent functions
        1.  
          ASMInst agent Startup options
        2.  
          ASMInst agent Shutdown options
      4.  
        ASMDG agent functions
    6.  
      Typical Oracle configuration in a VCS cluster
    7.  
      About setting up Oracle in a VCS cluster
  2. Installing and configuring Oracle
    1.  
      About installing Oracle in a VCS environment
    2.  
      Before you install Oracle in a VCS environment
    3. About VCS requirements for installing Oracle
      1.  
        Location of the $ORACLE_HOME
      2.  
        Failing over Oracle after a VCS node failure during hot backup
    4. About Oracle installation tasks for VCS
      1.  
        Installation tasks for $ORACLE_HOME on shared disks
      2.  
        Installation tasks for $ORACLE_HOME on local disks
      3.  
        Installation tasks for Oracle using ASM
    5. Installing ASM binaries in a VCS environment
      1.  
        Configuring VCS to support Oracle Restart function
    6. Configuring Oracle ASM on the first node of the cluster
      1.  
        Enabling the clustering daemon for ASM-managed database
      2.  
        Disabling automatic startup of ASM
    7.  
      Configuring and starting up ASM on remaining nodes
    8. Installing Oracle binaries on the first node of the cluster
      1.  
        Disabling the clustering daemon
    9.  
      Configuring the Oracle database
    10.  
      Copying the $ORACLE_BASE/admin/SID directory
    11.  
      Copying the Oracle ASM initialization parameter file
    12.  
      Verifying access to the Oracle database
  3. Installing and removing the agent for Oracle
    1.  
      Before you install or upgrade the agent for Oracle
    2.  
      Installing the VCS agent for Oracle software
    3. Upgrading the VCS agent for Oracle
      1.  
        Upgrading VCS agent for Oracle
    4.  
      Disabling the Cluster Server agent for Oracle
    5.  
      Removing the Cluster Server agent for Oracle
  4. Configuring VCS service groups for Oracle
    1.  
      About configuring a service group for Oracle
    2. Configuring Oracle instances in VCS
      1.  
        Configuring a single Oracle instance in VCS
      2.  
        Configuring multiple Oracle instances (single listener) in VCS
      3.  
        Configuring multiple Oracle instances (multiple listeners) in VCS
      4.  
        Configuring an Oracle instance with shared server support in VCS
    3. Before you configure the VCS service group for Oracle
      1.  
        Importing the type definition files for Cluster Server agent for Oracle
    4. Configuring the VCS service group for Oracle
      1.  
        Configuring the VCS service group for Oracle using Cluster Manager (Java console)
      2. Configuring the VCS service group for Oracle using the command-line
        1.  
          Encrypting Oracle database user and listener passwords
      3. Setting up detail monitoring for VCS agents for Oracle
        1. Setting up detail monitoring for Oracle
          1.  
            Enabling and disabling detail monitoring for Oracle resource
        2.  
          Setting up detail monitoring for Netlsnr
      4. Enabling and disabling intelligent resource monitoring for agents manually
        1.  
          Administering the AMF kernel driver
  5. Administering VCS service groups for Oracle
    1.  
      About administering VCS service groups
    2.  
      Bringing the service group online
    3.  
      Taking the service group offline
    4.  
      Switching the service group
    5.  
      Modifying the service group configuration
  6. Pluggable database (PDB) migration
    1.  
      Migrating Pluggable Databases (PDB) between Container Databases (CDB)
  7. Troubleshooting Cluster Server agent for Oracle
    1.  
      About troubleshooting Cluster Server agent for Oracle
    2.  
      Error messages common to the Oracle and Netlsnr agents
    3.  
      Error messages specific to the Oracle agent
    4.  
      Error messages specific to the Netlsnr agent
    5.  
      Error messages specific to the ASMInst agent
    6.  
      Error messages specific to the ASMDG agent
    7.  
      Troubleshooting issues specific to Oracle in a VCS environment
    8. Verifying the Oracle health check binaries and intentional offline for an instance of Oracle
      1.  
        Verifying the intentional offline behavior of the VCS Oracle agent
    9.  
      Disabling IMF for a PDB resource
  8. Appendix A. Resource type definitions
    1.  
      About the resource type and attribute definitions
    2. Resource type definition for the Oracle agent
      1.  
        Resource type definition for systemD Support
      2.  
        Attribute definitions for the Oracle agent
      3.  
        About the Sid attribute in a policy managed database
      4. Using systemD attributes for Oracle application service
        1.  
          Adding key-value pairs to SystemDAttrList attribute
        2.  
          Customizing the Oracle service unit file with systemD attributes
        3.  
          Configuring HugePages for Oracle on Linux
    3. Resource type definition for the Netlsnr agent
      1.  
        Attribute definition for the Netlsnr agent
    4. Resource type definition for the ASMInst agent
      1.  
        Resource type definition for the ASMInst agent to support systemD
      2.  
        Attribute definition for the ASMInst agent
    5. Resource type definition for the ASMDG agent
      1.  
        Attribute definition for the ASMDG agent
  9. Appendix B. Sample configurations
    1.  
      About the sample configurations for Oracle enterprise agent
    2. Sample single Oracle instance configuration
      1.  
        Sample VCS configuration file for single Oracle instance
    3. Sample multiple Oracle instances (single listener) configuration
      1.  
        Sample VCS configuration file for multiple Oracle instances (single listener)
    4. Sample multiple instance (multiple listeners) configuration
      1.  
        Sample VCS configuration file for multiple Oracle instances (multiple listeners)
    5. Sample Oracle configuration with shared server support
      1.  
        Sample VCS configuration file for Oracle instance configured with shared server support
    6. Sample Oracle ASM configurations
      1. Sample configuration for ASM disks as raw disks
        1.  
          Sample VCS configuration file for ASM-managed Oracle instance
      2. Sample configuration for ASM disks as VxVM volumes
        1.  
          Sample VCS configuration file for ASM disks as VxVM disks
      3. Sample configuration for ASM disks as CVM volumes
        1.  
          Sample VCS configuration file for ASM that uses CVM volumes
    7.  
      Sample configuration of Oracle pluggable database (PDB) resource in main.cf
    8.  
      Sample configuration of migratable Oracle pluggable database (PDB) resource in main.cf
    9.  
      Sample Configuration of Oracle supported by systemD
    10.  
      Sample configuration of ASMInst supported by systemD
  10. Appendix C. Best practices
    1.  
      Best practices for multiple Oracle instance configurations in a VCS environment
  11. Appendix D. Using the SPFILE in a VCS cluster for Oracle
    1.  
      About the Oracle initialization parameter files
    2.  
      Starting an Oracle instance
    3.  
      Using the SPFILE in a VCS cluster
  12. Appendix E. OHASD in a single instance database environment
    1.  
      About OHASD in a single instance database environment
    2.  
      Configuring the Application agent to make OHASD highly available

Migrating Pluggable Databases (PDB) between Container Databases (CDB)

You can use the hapdbmigrate utility to perform a planned end-to-end migration of pluggable databases across containers of the same database version. The destination container database may be on the same node or on another node. The PDB must have its own independent storage.

If the PDB has parent groups, they will be unlinked and frozen by the utility during the migration. After successful migration, they will be relinked and unfrozen by the utility.

The configuration must meet the following requirements:

  • The PDB and CDB databases must be on Cluster File System (CFS) or File system (VxFS).

  • There must be no parent resources for the PDB resource.

  • The PDB must be plugged in to the source CDB. The PDB resources may or may not be offline.

  • The source and destination CDB resources for the migration must be different.

  • The version of the destination CDB must be the same as the source CDB.

  • The PDB to be migrated is mounted on an independent file system (separate mount points for CDB and PDB datafiles).

  • The CDB and PDB resources must be configured in the same service group.

The utility performs the following actions during the migration:

  • Unlinks and freezes the parent groups, if any, depending on the source CDB group where the PDB resource to be migrated is configured.

  • Takes PDB resources offline.

  • Unplugs the PDB from the CDB and creates an XML file, <pdb_res_name>_<dest_cdb_res_name>.xml in the XML directory provided by the user.

  • Drops the PDB from the CDB keeping the datafiles.

  • Takes offline all the PDB child resources. Unlinks the PDB resource from the source CDB resource and deletes the PDB resource. Unlinks and deletes all PDB child resources.

  • Recreates the PDB resource and all its children with original dependencies in the destination CDB service group.

  • Brings online the PDB child resources on all the nodes where the destination CDB service group is online

  • Plugs the PDB in the destination CDB.

  • Brings the PDB resource online after successful plugging. Unfreezes and links the parent group to the source CDB group.

The hapdbmigrate utility performs certain pre-requisite checks before the migration to verify that the cluster is ready for PDB migration. If the utility encounters any issues, you will need to manually fix the issues.

The utility is present in the $VCSHOME/bin directory.

The log files of the migration are located at $VCSLOG/log/hapdbmigrate.log. The logs are rotated after the file exceeds 5 MB and is saved in .gz format hapdbmigrate.log[1..7].gz.

Note:

Ensure that only one instance of the hapdbmigrate utility is running at a time.

To migrate Pluggable Databases (PDB) between Container Databases (CDB)

  1. Back up the VCS configuration file /etc/VRTSvcs/conf/config/main.cf:
    # cp /etc/VRTSvcs/conf/config/main.cf \
    /etc/VRTSvcs/conf/config/main.cf.save
  2. Verify that the high availability daemon (had) is running on all the nodes in the cluster.
    # hastatus -summary
  3. Verify that there are no resources in faulted or unknown state.
    # hares -state|grep FAULTED
    # hares -state|grep UNKNOWN
  4. Verify that the PDBName attribute is present for the PDB resource with the correct value in the main.cf configuration file.
  5. Verify that the source and destination CDB resources are online.
    # hares -state resname
  6. On the destination CDB, verify the following:

    • The destination CDB is not in suspended mode.

      Any instance of the destination CDB is not in mounted state.

      See the Oracle documentation for more information.

    • If any instance of the destination CDB is in restricted state, ensure that the PDB resource you want to migrate has the StartUpOpt attribute set to restricted.

      # haconf -makerw
      # hares -modify pdb1 StartUpOpt \
      RESTRICTED
      # haconf -dump -makero
  7. Verify that existing dependencies do not conflict with the migration process.

    The PDB child resources must not be dependent on the CDB resource or any of its child resources.

    The PDB child resources must not have any parent, which is not a part of the PDB child hierarchy.

  8. Verify that the XML data directory has read and write permissions for the "oracle" user. The XML data directory must be located either on PDB mounts or at a location accessible to both source and destination CDBs.
  9. Run the hapdbmigrate utility as the root user:

    Note:

    If there are parent groups dependent on the source CDB group, specify the -ignoreparentgrp option.

    # $VCS_HOME/bin/hapdbmigrate -pdbres pdb_resname -cdbres cdb_resname \
    -XMLdirectory xml_directory [-ignoreparentgrp] [-prechecks]
    		-pdbres: Name of the PDB resource, which needs to be migrated
    	-cdbres: Name of the CDB resource, where the PDB needs to migrate
    	-XMLdirectory: XML directory location for the unplugged PDB
    -ignoreparentgrp: Utility proceeds even 
    if the PDB group has parent groups 
    -prechecks: Performs prechecks and validation
    -help|h: Prints usage

    The migration log file is located at $VCSLOG/log/hapdbmigrate.log.

  10. Verify that the PDB resource is online on the destination CDB.
    # hares -state pdb_resname
  11. Relink the parent service group of the source CDB group manually to the destination CDB group, if it depends on the migrated PDB.
    # haconf -makerw
    # hagrp -dep parent_sg
    #Parent     	     Child         	   Relationship
    parent_sg       source_CDB         online local firm
    # hagrp -offline parent_sg -any
    # hagrp -unlink parent_sg source_CDB
    # hagrp -link parent_sg dest_CDB online local firm 
    # hagrp -dep parent_sg
    #Parent     	    Child         	    Relationship
    parent_sg      dest_CDB           online local firm
    # haconf -dump -makero

See Sample configuration of migratable Oracle pluggable database (PDB) resource in main.cf.

Resource type definition for the Oracle agent

Resource type definition for the Netlsnr agent