Oracle® Database

5.1.1 Oracle Base Sharing Issue with Oracle RAC

When you share an Oracle base of a previous release with Oracle RAC, Database Configuration Assistant may report the following errors:

ORA-48141: error creating directory during ADR initialization 
ORA-48189: OS command to create directory failed

Workaround:

Manually modify the permissions of ORACLE_BASE/diag directory by adding the following:

hostname\ORA_GRID_LISTENERS Full Control
Oraclehomeuser Full control

For example:

cacls.exe ORACLE_BASE\diag /E /T /G ORA_GRID_LISTENERS:F OracleHomeUser:F

This issue is tracked with Oracle bug 14353370.

5.1.2 Oracle Universal Installer Wrongly Reports Insufficient Disk Space During an Oracle RAC Installation

Oracle ACFS mounted on any drive letter may result in the following error during an Oracle RAC installation:

[INS-32021] Insufficient disk space on this volume for the selected Oracle home.
[INS-30052] Check for free disk space across nodes failed.

Workaround:

If the location you have chosen for Oracle home has at least 3614 MB of free disk space, then create a top-level directory of ORACLE_HOME on Oracle ACFS drive and use that directory as the mount point.

This issue is tracked with Oracle bugs 14255871 and 14547206.

5.1.3 Cluster Verification Utility Check Fails After Oracle Grid Infrastructure Upgrade

When you upgrade Oracle Grid Infrastructure for a cluster from release 11.1.0.7 to release 12.1, Oracle Cluster Registry Integrity check may fail as Oracle Cluster Registry appears to be on Oracle File System (which is deprecated).

The following is the error:

PRVF-10037 : Failed to retrieve storage type for xx on node xx 

Workaround:

Perform the following steps to upgrade Oracle Grid Infrastructure:

1. Upgrade Oracle Grid Infrastructure release 11.1.0.7 to release 11.2.0.3.

2. Move Oracle Cluster Registry to Oracle Automatic Storage Management.

3. Upgrade Oracle Grid Infrastructure to release 12.1.

This issue is tracked with Oracle bug 13615858.

5.1.4 Deinstallation Tool Does Not Remove invDetails.properties File

The Deinstallation Tool fails to remove the invDetails.properties file when removing Oracle RAC home.

Workaround:

Manually remove the dbhome\inventory directory.

This issue is tracked with Oracle bug 13940962.

5.1.5 Using Deinstallation Tool for a Single-Instance Oracle Database Returns Errors

The Deinstallation Tool fails to remove Oracle home and displays file in use errors as the services and processes running from Oracle home are not removed completely before Oracle home removal.

Workaround:

Manually remove Oracle home.

This issue is tracked with Oracle bug 10035858.

5.1.6 Oracle Restart Integrity Check Fails During Oracle Grid Infrastructure Upgrade

When upgrading Oracle Grid Infrastructure for a standalone server to release 12.1, Oracle Restart integrity check may fail as the service, OracleOHService, does not start after system restart.

Workaround:

Set the Startup Type of OracleOHService to Automatic in the Windows Services Manager and start OracleOHService.

This issue is tracked with Oracle bug 14592705.

5.1.7 Oracle Home has Incorrect Access Control List

If you install Oracle RAC into a shared Oracle home directory that is a subdirectory of an unshared Oracle base directory, then the Access Control List of Oracle home is incorrect. As a result, the shortcuts may not work.

Workaround:

During Oracle RAC installation, do not mount Oracle ACFS as a subdirectory of Oracle base.

This issue is tracked with Oracle bug 14593681.

5.1.8 Oracle ACFS Version Mismatch Issue

When upgrading Oracle Grid Infrastructure for a standalone server to release 12.1, you may encounter the following error:

ACFS-00515: version mismatch

Workaround:

Perform the following steps before beginning the upgrade of Oracle Grid Infrastructure for a standalone server:

1. Stop Oracle High Availability Services stack:

   crsctl stop has
   

2. Disable the automatic startup of Oracle High Availability Services stack:

   crsctl disable has
   

3. Restart the node.

4. Uninstall Oracle ACFS:

   acfsroot uninstall
   

5. Start Oracle High Availability Services stack:

   crsctl start has
   

6. Begin the upgrade of Oracle Grid Infrastructure for a standalone server.

7. Reenable Oracle ACFS volumes and mount points using SQL, Oracle Automatic Storage Management Configuration Assistant, or command-line utilities.

This issue is tracked with Oracle bug 14615024.

5.1.9 OUI-35078 Error When Upgrading Oracle Grid Infrastructure

When upgrading to Oracle Grid Infrastructure for a cluster release 12.1, you may encounter the following errors:

OUI-35078: Error in updating env on nodes 'node_name' 
PRKN-1019: Failed to create directory "C:\%Temp%\oraremservice\"on node" node_name "
Error: The network name cannot be found.

Workaround:

Ignore the error and continue with the upgrade.

This issue is tracked with Oracle bug 14646033.

5.1.10 Setting Access Control List on Remote Node Fails

When upgrading Oracle Grid Infrastructure release 11.2.0.3 to release 12.1, setting the Access Control List on a remote node fails with the following error:

PRKN-1019 : Failed to create directory
""%TMPDIR%\oraremservice\" on node
"node_name", Error: "0|The network name cannot be found.

Workaround:

Ignore the error and continue with the upgrade.

This issue is tracked with Oracle bugs 14639343 and 14579961.

5.1.11 Cluster Verification Utility (CVU) Must Be Stopped Before Oracle Clusterware Upgrade

When upgrading Oracle Clusterware release 11.2 to release 12.1, the CVU uses Oracle home from the earlier release, while Oracle Universal Installer uses the services of release 12.1.

Workaround:

Perform the following steps:

1. Manually stop the CVU before starting the upgrade:

   srvctl stop cvu
   

   or

   crsctl stop res ora.cvu
   

2. Stop the dbconsole service of the target Grid home from the Services control panel on all the nodes.

This issue is tracked with Oracle bug 14669580.

5.1.12 INS-30052 Error During Database Upgrade

When upgrading Oracle Grid Infrastructure for a cluster from release 11.2.0.3 to release 12.1, Oracle Universal Installer may show the following error:

WARNING: [WARNING] [INS-30052] Check for free disk space across nodes failed.

Workaround:

Select Cancel and click Next. The error does not appear again.

This issue is tracked with Oracle bug 14667513.

5.1.13 ORA-01565 Error When Upgrading Oracle Database Using Oracle Database Upgrade Assistant

When upgrading Oracle Database to release 12.1 and specifying an Oracle home user, Oracle Database Upgrade Assistant fails to start the database and displays the following error:

ORA-01565: error in identifying file 'string'

Workaround:

Perform the following steps:

1. Upgrade Oracle Grid Infrastructure database release 11.2 to Oracle Grid Infrastructure release 12c. Do not select the option to configure a Grid Infrastructure Management Repository database. Also, do not specify an Oracle home user.

2. After the upgrade completes, get a list of all Oracle ASM disks using the following command:

   asmtool -list
   

3. For every Oracle ASM disk, use the oradim command, passing the appropriate DosDeviceName in the form \\.\XXXXXXXX to update the permissions of all Oracle ASM disks. On each cluster node, run the following command for each Oracle ASM disk:

   oradim  -ACL -setperm RawDevice -USER oracle -OBJTYPE RawDevice -OBJPATH \\.\ORCLDISKDATA0 -RECURSE false
   

4. Proceed to upgrade Oracle Database.

This issue is tracked with Oracle bug 14646315.

5.1.14 Oracle Grid Infrastructure Configuration Issue

When upgrading Oracle Grid Infrastructure for a cluster from release 11.2.0.3 to release 12.1, Oracle Grid Infrastructure configuration may fail with the following error message in the log file, Grid_home/cfgtoollogs/crsconfig/:

unable to run (%ORACLE_HOME%\perl\bin\perl.exe %ORACLE_HOME%/crs/install/rootcrs.pl -remotenode -upgrade) on node node_name

Workaround:

Perform the following steps:

1. Run the following command on each failed remote node:

   Grid_home\crs\config\crsconfig.bat -upgrade
   

2. Click Skip to ignore the error message on Oracle Universal Installer screen and proceed to the next configuration tool.

This issue is tracked with Oracle bug 14492080.

After performing the last step, continue with the installation. If you see the error message, PRKN-1014 : Failed to execute remote command "Grid_home\bin\olsnodes.exe" on node node_name, then click Ignore, and continue with the installation.

This issue is tracked with Oracle bug 14646125.

5.1.15 Issues While Upgrading Oracle Grid Infrastructure for a Cluster to Release 12.1

Oracle Database Configuration Assistant may hang while creating Oracle Grid Infrastructure Management Repository database because the old Oracle ASM links are not removed.

Workaround:

Select No on the Grid Infrastructure Management Repository Option screen during the database upgrade.

This issue is tracked with Oracle bug 14511564.

5.1.16 Oracle Database Fails to Restart

If you enable the ACCESS_CONTROL.ENABLED disk group attribute, the file access control feature does not work correctly and Oracle Database fails to restart.

Workaround:

Disable the ACCESS_CONTROL.ENABLED attribute and restart Oracle Database.

This issue is tracked with Oracle bug 14083902.

5.1.17 PLS-00923 Error During Database Upgrade

On Windows Server 2008 R2 or later, when you upgrade Oracle Database Release 10.2 that has Oracle Label Security or Oracle Database Vault installed, the olspreupgrade.sql script may fail with the following error:

Errors for FUNCTION GET_SPACE_REQUIRED:
0/0      PLS-00923: native compilation failed: C compilation or linking
           failed:cl : Command line error D8037 : cannot create temporary il
           file; clean temp directory of old il files
Warning: Function created with compilation errors.

Workaround:

This issue is fixed in Oracle Database Release 10.2 and is tracked with Oracle bug 14304710. Apply the fix to the 10.2 Oracle home and then run the olspreupgrade.sql script.

5.1.18 Disk Space Requirement Issue for Personal Edition

The disk space requirement for Personal Edition is 4.85 GB, while Oracle Universal Installer shows 3.9 GB.

This issue is tracked with Oracle bug 16771442.

5.1.19 Oracle Net Listener Fails to Start

Oracle Net Listener may fail to start on a Windows host that does not have the loopback address, 127.0.0.1, configured.

Workaround:

This loopback address is required. However, at an external network level, the host may be configured to have only IPv6 addresses for the supported configurations.

This issue is tracked with Oracle bug 16867401.

5.1.20 Deinstallation Tool Does Not Remove Operating System Groups

The Deinstallation Tool fails to remove the following operating system groups when removing Oracle Database Gateway:

ORA_DBA
ORA_INSTALL
ORA_CLIENT_LISTENERS
ORA_HOMENAME_DBA
ORA_HOMENAME_SVCSIDS

Workaround:

Manually delete the groups, ORA_HOMENAME_DBA and ORA_HOMENAME_SVCSIDS using the following command:

net localgroup /DELETE group_name

If there are no other Oracle homes registered in the central inventory or if the inventory was deleted during the deinstallation, then delete the ORA_DBA, ORA_INSTALL, and ORA_CLIENT_LISTENERS groups too.

This issue is tracked with Oracle bug 16870153.

5.1.21 Deinstallation Tool Error Message

The Deinstallation Tool displays that the cleanup of some registry entries fails with the following error message, however, in most cases, the registry keys do not exist:

The specified path is invalid.. 

Workaround:

Manually delete the registry keys mentioned in the error message, if present.

This issue is tracked with Oracle bug 16805631.

5.1.22 Oracle Configuration Manager (OCM) Left in Disconnected Mode in a Cloned Oracle Home

When using the deriveCCR command, the OCM does not prompt for information and may remain in the disconnected mode in the following conditions:

• The OCM in the source Oracle home is in disconnected mode when copied to the cloned Oracle home.

• There is no connectivity to the end point.

• Credentials cannot be verified.

Workaround:

Change the OCM to connected mode by running the configCCR command from the cloned Oracle home after the deriveCCR command completes.

This issue is tracked with Oracle bug 11723395.

5.1.23 Issue When the Host Name is in Uppercase Letters

Oracle Universal Installer reports the following error on the Specify Management Options screen when you specify the host name of the system with uppercase letters:

[INS-35026] Unable to connect to Enterprise Manager Cloud Control.

This issue is tracked with Oracle bug 16814249.

5.1.24 Oracle Grid Infrastructure Installation Fails with Shared Grid Naming Service (GNS)

While installing Oracle Grid Infrastructure for a cluster on a shared GNS, the installation may fail if you provide the client data file path location on a remote computer.

Workaround:

Copy the client data file to a local computer and provide the local path to Oracle Universal Installer.

This issue is tracked with Oracle bug 17017165.

5.1.25 Services for an Oracle RAC 10.2 Database Instance Fail to Start

After configuring an Oracle RAC Release 10.1, 10.2, or 11.1 on an Oracle Grid Infrastructure release 12.1, when you restart a node, the database instance resource running on that node may erroneously report an online state when it is actually offline. Any service associated with the instance fails to start.

Workaround:

Run the following commands on the node that was restarted:

srvctl stop instance -d db_unique_name -i instance_name
srvctl start instance -d db_unique_name -i instance_name

This issue is tracked with Oracle bug 10224187.

5.1.26 Oracle Home User Password Issue When Using Non-ASCII Characters for Certain Languages

While running Oracle Universal Installer to install Oracle Database, if you provide a non-ASCII password for Oracle home user, the installation may fail with the following error for Simplified Chinese, Traditional Chinese, Japanese, and Korean languages.

INS-20802 Oracle net configuration assistant failed

Workaround:

Specify an ASCII password for Oracle home user.

This issue is tracked with Oracle bug 17073971.

5.2.1 Using Oracle ACFS for the Database Home and Large Object Files

When Oracle home is on Oracle ACFS, the installation location must be of the format, DRIVE_LETTER:\directory, that is, Oracle ACFS should be mounted on a new drive.

This issue is tracked with Oracle bug 9527303.

When installing or upgrading Oracle Database home on Oracle ACFS, Oracle Database Configuration Assistant and Oracle Database Upgrade Assistant fail to instantiate schemas that have large objects (for example, Database sample schemas).

Workaround:

Use SQL*Plus to manage LOB files when Oracle home is on Oracle ACFS.

This issue is tracked with Oracle bug 13702902.

5.2.2 Oracle Grid Infrastructure Listeners Issue

After installing Oracle Grid Infrastructure, the default listener does not listen on public addresses. As a result, even though the listener is available, the listener does not service the database or Oracle ASM instances.

Workaround 1:

If you are not using GNS with DHCP, then perform the following steps:

1. Add the local IP address and host name information to %Systemroot%\System32\Drivers\Etc\Hosts.

   For example, IP address: 192.0.2.254

   host name: dbhost

2. Restart Oracle Grid Infrastructure listener.

This issue is tracked with Oracle bug 9382523.

Workaround 2:

If you have configured GNS and use DHCP to assign the addresses for the cluster nodes, then configure DHCP so that the IP address assigned to the host name for each cluster node (the public IP address) has a smaller numeric value than the IP addresses assigned for the Virtual IP (VIP) address, GNS VIP address, or SCAN VIP address for a node.

For example, you could use IP addresses in the range of xxx.xxx.xxx.0 – xxx.xxx.xxx.099 for only host IP addresses, and all other IP address assignments must use a value of xxx.xxx.xxx.100 or higher.

This issue is tracked with Oracle bug 9441288.

5.2.3 Oracle ACFS Security Realm Issue

When adding objects to an Oracle ACFS security realm, the command is successful but you may see the following error message:

acfsutil sec realm add: ACFS-10583: Internal API error : [Save DOM failed, 202].

Workaround:

Add the security administrator to the realm that contains the entire file system.

This issue is tracked with Oracle bug 14586440.

5.2.4 Incorrect Parameter Error When Running Shortcuts

In Windows 2008 R1, if you run a shortcut that resolves to an executable on an Oracle ACFS file system, and select the Run as administrator check box in the shortcut properties, you may encounter the following error message:

The parameter is incorrect

Workaround 1:

Ensure that you do not select the Run as administrator check box if the program does not require administrator rights. If required, select the check box and run the program.

Workaround 2:

Mount Oracle ACFS file system on a folder on an NTFS file system.

This issue is tracked with Oracle bug 14635955.

5.2.5 Database Files Fail to Delete on Dropping a Database

Dropping a database may not delete the database files physically, if the database was opened (even though it was closed subsequently) before the dropping operation. This issue can occur with a regular database or a pluggable database.

No errors are reported in the SQL*Plus command but the database files may not be deleted physically. The alert log may report errors regarding the failure to delete the files.

Workaround:

Manually remove the relevant database files.

This issue is tracked with Oracle bug 15865933.

5.2.6 Additional Mount Option Support Issue

The SRVCTL modify filesystem and add filesystem commands allow you to specify optional file system mount options on all platforms, using the argument -fsoptions. You use this option to specify Oracle ACFS and operating system specific mount options. However, Windows does not support additional mount options (such as read only), so, -fsoptions has no effect on the underlying mount.

This issue is tracked with Oracle bug 14579939.

5.2.7 Shortcuts Not Organized on Windows Server 2012

The Start screen shows all the shortcuts together, that is, the shortcuts are not grouped appropriately.

This issue is tracked with Oracle bug 16023635.

5.2.8 ORA-01017 Error with Multitenant Container Databases

In multitenant container databases on Microsoft Windows platforms, user authentication through Enterprise User Security is not supported. If you attempt to log in as an Enterprise Security user to a pluggable database, then the login fails with the following error message:

ORA-01017: Invalid username/password;logon denied.

Workaround:

None

This issue is tracked with Oracle bug 16385258.

5.2.9 Grid Management Repository Database Fails to Restart After Process Failure in Heavy Load Situations

If you are using the Grid Management Repository database, and a process failure occurs on a heavily loaded system, the database may fail to restart.

Workaround:

Use the following commands to re-create the Grid Management Repository database listener, then start the Repository database:

crsctl check res ora.MGMTLSNR
srvctl start mgmtdb

This issue is tracked with Oracle bug 16772567.

5.2.10 Oracle Flex ASM Fails to Start

Oracle Flex ASM service may fail to start while moving the Grid home to a new location.

Workaround:

Open the Control Panel and remove all the services that are named Oracle, running from an earlier Grid home.

This issue is tracked with Oracle bug 16801852.

5.2.11 Oracle Enterprise Manager Database Express URL Issue

When you create an Oracle RAC, Oracle Enterprise Manager Express HTTPS URL may not work as the grid listeners do not have permission to access the wallet directory. The HTTP URL works fine.

Workaround:

Add permission for all the wallet files on XDB wallet folder by running the following command:

cacls.exe xdb_wallet path /T /E /G ORA_GRID_LISTENERS:F

If Oracle home user is set during an Oracle RAC installation, then the installation user may not have the permission to update the ACLs on wallet files. In that case, run the command with the specific Oracle home user.

This issue is tracked with Oracle bug 16394272.

5.2.12 Online Database Relocation Issue

Oracle RAC One Node database online relocation sometimes displays the following error:

A fatal error has been detected by the Java Runtime Environment: 
EXCEPTION_ACCESS_VIOLATION (0xc0000005

The relocation completes regardless of the error.

Workaround:

None

This issue is tracked with Oracle bug 14680430.

5.2.13 Oracle Home User Password Issue When Using Non-ASCII Characters with Command-Line Utilities on Non-English Locale

When you run command-line utilities, such as ORADIM, LSNRCTL, or Oracle home user tool, orahomeuserctl.bat on a non-English locale to provide Oracle home user password, any non-ASCII characters used in the password may enter as garbled text. Any non-ASCII data displayed on the screen may also appear as garbled text. This issue reproduces on computers that are using a non-English locale and only if Oracle Enterprise Manager code page is different from the local Windows ANSI code page.

Workaround:

Perform the following steps:

1. Determine the system Windows ANSI code page on the local computer by viewing the value of the following key in Windows registry:

   HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\NLS\CodePage\ACP
   

2. Open a Command Prompt (console) window and run the chcp command:

   chcp Windows_ANSI_code_page
   

3. Set the font property for this Command Prompt window to a TrueType font:

   1. Right-click the title bar of the window and select Properties.

   2. Click Font and select Lucida Console from the list.

   3. Click OK.

4. Then, run the utility in the same Command Prompt window.

This issue is tracked with Oracle bug 16420596.

5.2.14 Oracle ASM Instance Status Issue with Oracle Grid Infrastructure for a Standard Cluster

When you install Oracle Grid Infrastructure for a standard cluster with Oracle Flex ASM storage and shared GNS option, the status of Oracle ASM resources may show INTERMEDIATE state.

Workaround:

None, the state of Oracle ASM resources automatically changes to ONLINE in some time.

This issue is tracked with Oracle bug 17038111.

5.2.15 Oracle Home Permissions Issue on Oracle RAC

During an Oracle RAC installation, the read and execute permissions for the Windows operating system group, Authenticated Users, may be removed from the ORACLE_HOME directory if all the following conditions are true:

• Oracle home location specified is under Oracle base

• Oracle home is on Oracle ACFS mount, shared across nodes

• Oracle base is on a local file system, non-shared location

• Oracle base and Oracle home directory names use different case on Oracle Universal Installer screen

  For example:

  C:\app\base
  c:\app\base\product\dbhome_1
  

Workaround:

Grant read and execute permissions to the Windows operating system group, Authenticated Users, for Oracle home directory, including all its subfolders and files.

This issue is tracked with Oracle bug 17041387.

5.2.16 PRVF-7535 Error on Windows 8

While installing Oracle Database Client (32-bit) on Windows 8, you may see the following error:

PRVF-07535: Proper architecture is not found on node "{0}" [Expected = "{1}" ; Found = "{2}"]

Workaround:

Ignore the error and proceed with the installation.

This issue is tracked with Oracle bug 17015590.

5.2.17 Management Database Relocation Issue

Management Database relocation fails the first time if the target node does not have Oracle ASM instance.

Workaround:

Relocating the management database the second time is successful.

This issue is tracked with Oracle bug 16987418.

5.2.18 Oracle Home User Domain Causes Issues with Oracle Database Extensions for .NET and Oracle Database Gateways

If Oracle home user is a Windows domain user and it belongs to a different Windows domain than the computer's domain, running a.NET stored procedure fails with the following errors:

ORA-28575: unable to open RPC connection to external procedure agent
ORA-06512: at stringline string

Any attempt to access Oracle Database Gateway also fails with the following error:

SQL> select * from dual@holink2;
select * from dual@holink2

ORA-28545: error diagnosed by Net8 when connecting to an agent
Unable to retrieve text of NETWORK/NCR message 6553

This issue does not occur if you are using Windows built-in account or Windows local user account as Oracle home user.

Workaround:

Use an Oracle home user that belongs to the same Windows domain as the computer's domain, or contact Oracle Support.

This issue is tracked with Oracle bug 16221493.

6.2.1 Upgrading to 12.1.0.2

Install Windows DB Bundle Patch 12.1.0.1.20 or higher before upgrading 12.1.0.1 Standard Edition to 12.1.0.2 Standard Edition 2.

6.2.2 Oracle Universal Installer Wrongly Reports Insufficient Disk Space During an Oracle RAC Installation

Oracle ACFS mounted on any drive letter may result in the following error during an Oracle RAC installation:

[INS-32021] Insufficient disk space on this volume for the selected Oracle home.
[INS-30052] Check for free disk space across nodes failed.

Workaround:

If the location you have chosen for Oracle home has at least 3614 MB of free disk space, then create a top-level directory of ORACLE_HOME on Oracle ACFS drive and use that directory as the mount point.

This issue is tracked with Oracle bugs 14255871 and 14547206.

6.2.3 PLS-00923 Error During Database Upgrade

On Windows Server 2008 R2 or later, when you upgrade Oracle Database Release 10.2 that has Oracle Label Security or Oracle Database Vault installed, the olspreupgrade.sql script may fail with the following error:

Errors for FUNCTION GET_SPACE_REQUIRED:
0/0      PLS-00923: native compilation failed: C compilation or linking
           failed:cl : Command line error D8037 : cannot create temporary il
           file; clean temp directory of old il files
Warning: Function created with compilation errors.

Workaround:

This issue is fixed in Oracle Database Release 10.2 and is tracked with Oracle bug 14304710. Apply the fix to the 10.2 Oracle home and then run the olspreupgrade.sql script.

6.2.4 Deinstallation Tool Error Message

The Deinstallation Tool displays that the cleanup of some registry entries fails with the following error message, however, in most cases, the registry keys do not exist:

The specified path is invalid. 

Workaround:

Manually delete the registry keys mentioned in the error message, if present.

This issue is tracked with Oracle bug 16805631.

6.2.5 Deinstallation Tool Does Not Remove invDetails.properties File

The Deinstallation Tool fails to remove the invDetails.properties file when removing Oracle RAC home.

Workaround:

Manually remove the dbhome\inventory directory.

This issue is tracked with Oracle bug 13940962.

6.2.6 Temp Folder Issues During Deinstallation

During Oracle Database, Oracle Client, Oracle Grid Infrastructure, Oracle Gateways, or Global Service Manager (GSD) deinstallation on Microsoft Windows platforms, the deinstall tool does not automatically remove the deinstall boot strap folder in the temp folder.

Workaround:

Manually remove the deinstall boot strap folder in the temp folder.

This issue is tracked with Oracle bug 18250031.

6.2.7 Oracle Grid Infrastructure Deinstallation Error

When you deinstall Oracle Grid Infrastructure for a standalone server home that is configured after cloning Oracle home, the deinstallation may fail with the following error message:

Root script execution failed on one or more nodes

Workaround:

Before deinstalling, change HA_CONFIG=false to HA_CONFIG=true in Oracle_Restart_Home\crs\config\gridconfig.bat file.

This issue is tracked with Oracle bug 18749481.

6.2.8 Operating System Group Issues During Oracle Database Deinstallation

When installing Oracle Database 12c Release 1 products, the following operating system groups which are created may not be deleted automatically during the software deinstallation:

ORA_ASMADMIN
ORA_ASMDBA
ORA_ASMOPER
ORA_CLIENT_LISTENERS
ORA_GRID_LISTENERS
ORA_INSTALL

Workaround:

Manually delete the operating system groups found in the system after the deinstallation for Oracle home release 12c or higher by using the following command:

net localgroup /DELETE group_name

This issue is tracked with Oracle bug 18972422.

6.2.9 Registry Key Issues During Deinstallation

Deinstallation of a single instance Oracle Database or Oracle Client deinstallation or Global Service Manager (GSD) does not remove the following registry entry:

HKEY_LOCAL_MACHINE\SOFTWARE\ORACLE\DEINSTALL

Workaround:

Manually remove the following registry entry:

HKEY_LOCAL_MACHINE\SOFTWARE\ORACLE\DEINSTALL

This issue is tracked with Oracle bug 19251849.

6.2.10 Listener, SCAN Listener, and SCAN VIP Resources Fail to Start After Database Upgrade

After upgrading Oracle Grid Infrastructure or Oracle RAC from release 10.2.0.4 to 11.2.0.4, the listener, SCAN listener, and SCAN VIP resources fail to start when you restart the node.

Workaround:

Manually install Oracle ARP protocol driver and run the following command to start the oraarp driver automatically when restarting the node:

sc config oraarpdrv start= auto

This issue is tracked with Oracle bug 17084320.

6.2.11 Oracle Home User Password Issue When Using Non-ASCII Characters for Certain Languages

While running Oracle Universal Installer to install Oracle Database, if you provide a non-ASCII password for Oracle home user, the installation may fail with the following error for Simplified Chinese, Traditional Chinese, Japanese, and Korean languages.

INS-20802 Oracle net configuration assistant failed

Workaround:

Specify an ASCII password for Oracle home user.

This issue is tracked with Oracle bug 17073971.

6.2.12 Oracle Database Fails to Start

On Microsoft Windows platforms, if the value of ORACLE_HOME in the Windows registry of the database home is different from the database resource, Oracle database fails to start with the following error message:

LRM-00109: could not open parameter file 

This error shows regardless of the SPFILE specified in the corresponding database resource attribute.

Workaround:

Set the same value for ORACLE_HOME in the Microsoft Windows registry and the database resource.

This issue is tracked with Oracle bug 19333533.

6.2.13 PRVG-5150 Error Message When Installing Oracle Grid Infrastructure

When installing Oracle Grid Infrastructure release 12c, Oracle ASM device check may report a prerequisite warning error message PRVG-5150. The error message suggests read access for the file /etc/multipath.conf which does not exist on Windows platforms.

Workaround:

Ignore the error message and do not select the read access option for the file /etc/multipath.conf.

This issue is tracked with Oracle bug 18824041.

6.2.14 Oracle Universal Installer Issue

When installing Oracle Grid Infrastructure on Microsoft Windows Server 2012 R2 in an IPv6 network environment, Oracle Universal Installer displays the incorrect node name for the local node.

Workaround:

Specify the following public host name when running Oracle Universal Installer:

setup.exe -debug  ORACLE_HOSTNAME=public_hostname

This issue is tracked with Oracle bug 18729203.

6.2.15 Oracle ADVM Volume Incompatibility with Oracle ASM and Windows Fast FAT File System

If you use Oracle ASM Dynamic Volume Manager (Oracle ADVM) volumes on cluster nodes which is configured with highly available Oracle Automatic Storage Management (Oracle ASM), a deadlock may occur between Oracle ASM or Oracle ADVM and Windows Fast FAT file system. The deadlock may suspend the acfsformat.exe, acfsmount.exe, or any other tool or application attempting to open an Oracle ADVM volume.

Workaround:

Configure Windows cluster nodes with legacy Oracle ASM if FAT or FAT32 file systems are in use.

This issue is tracked with Oracle bug 19269068.

6.2.16 CVU Check Fails When Running DBCA

On Windows platforms, when you run Database Configuration Assistant (DBCA) to create a prerequisite clock synchronization check, Cluster Verification Utility (CVU) may fail if Network Time Protocol (NTP) is installed as the time synchronization mechanism on the cluster nodes.

Workaround:

Ensure that Windows service NTP is running on all the cluster nodes. Ignore the error and continue to run DBCA.

This issue is tracked with Oracle bug 19463695.

6.2.17 Delete a Database Option Disabled in DBCA

When upgrading Oracle Grid Infrastructure from release 11.2.0.3 to release 12.1.0.2, Database Configuration Assistant (DBCA) fails to delete the 11.2.0.3 database as the delete a database option is incorrectly disabled.

Workaround:

Download and install patch 14613223 which is available on My Oracle Support.

This issue is tracked with Oracle bug 19537493.

6.2.18 JRE Issue When Installing Oracle Database Gateways

When installing Oracle Database Gateways, Oracle Net Configuration Assistant (NETCA) is launched as a separate Java process which uses the JRE in the newly installed Oracle home. This JRE does not have the accessibility libraries and files needed for JAWS.

Workaround:

Perform the following steps:

1. Cancel the NETCA instance launched by Oracle Database Gateways installation.

2. Copy the following files to the installed Oracle home:

   %ORACLE_HOME%\jdk\jre\bin\JavaAccessBridge-64.dll
   %ORACLE_HOME%\jdk\jre\bin\JAWTAccessBridge-64.dll
   %ORACLE_HOME%\jdk\jre\bin\WindowsAccessBridge-64.dll
   %ORACLE_HOME%\jdk\jre\lib\accessibility.properties
   %ORACLE_HOME%\jdk\jre\lib\ext\access-bridge-64.jar
   %ORACLE_HOME%\jdk\jre\lib\ext\jaccess.jar
   

3. Restart NETCA.

This issue is tracked with Oracle bug 19071592.

6.2.19 Proxy_ADVM Fails to Start After GI Upgrade

When upgrading Oracle Grid Infrastructure from release 12.1.0.1 to release 12.1.0.2 with Oracle Flex ASM, proxy_advm and related Oracle ADVM and Oracle ACFS resources fails to start as proxy_advm cannot upgrade to a new home.

Workaround:

Perform the following steps in a rolling fashion:

1. Before upgrading Oracle GI, run the following command to modify Oracle ASM:

   srvctl modify asm -count all
   

2. After the upgrade completes, stop the apx service and delete it on one node using the following command:

   sc stop/delete apx service name
   

3. Restart the node.

4. Proceed to the next node and repeat steps 2 and 3.

This issue is tracked with Oracle bug 19362833.

6.2.20 Database Errors on Oracle ASM

When creating a database from an Oracle Database home for release 12.1.0.2 on Oracle ASM storage after upgrading Oracle Grid Infrastructure home to release 12.1.0.2, you may encounter the following database errors:

ORA-00200: control file could not be created
ORA-00202: control file: '...'
ORA-17502: ksfdcre:4 Failed to create file +...

OR

ORA-01501: CREATE DATABASE failed
ORA-00200: control file could not be created
ORA-00202: control file: '...'
ORA-17502: ksfdcre:3 Failed to create file +...
ORA-15001: diskgroup "..." does not exist or is not mounted
ORA-15040: diskgroup is incomplete

The Recovery Manager may also fail with the following errors:

RMAN-03002: failure of restore command at ...
ORA-01119: error in creating database file +¿
ORA-17502: ksfdcre:4 Failed to create file +...
ORA-15001: diskgroup "..." does not exist or is not mounted
ORA-15040: diskgroup is incomplete
RMAN-06956: create datafile failed; retry after removing +¿ from OS

Workaround:

Apply the fix for bug 19168335 in the software home for release 12.1.0.1 if you create the database from a software home which uses Oracle ASM as data storage. Ensure that you apply the fix in all the database homes for release 12.1.0.1 prior to upgrading Oracle GI to release 12.1.0.2.

If the new database is created on Oracle ACFS or other data destination storage, then the fix is not required.

This issue is tracked with Oracle bug 19168335.

6.2.21 OUI Error When Upgrading Oracle Grid Infrastructure

When you upgrade Oracle Grid Infrastructure release 11.1.0.7 to 11.2.0.3 and then to release 12.1.0.2, Oracle Universal Installer may display the following errors:

Execution of GI Upgrade script is successful on nodes
Configure Grid Infrastructure on nodes of Batch 1 failed
Overall status of execution of root/configuration scripts : failed

Workaround:

Ensure that you run the gridconfig.bat to configure Oracle Grid Infrastructure on the nodes in the batch. If the log file does not complete successfully with the message CLSRSC-325, rerun the gridconfig.bat manually on that node.

This issue is tracked with Oracle bug 18731658.

6.2.22 Incorrect Status for SCAN Listener and Oracle ASM

When installing Oracle Enterprise Manager Cloud Control in a Grid Naming Service (GNS) configured environment, Oracle Management Server (OMS) may show incorrect status for SCAN Listener and Oracle Automatic Storage Management (Oracle ASM).

Workaround:

Perform the following steps:

1. Select the target which shows the incorrect status. From the target drop-down list, select Target Setup, and then select Monitoring Configuration.

2. Enter the correct VIP name with the domain for the Machine Name property of the target and proceed with the installation.

This issue is tracked with Oracle bug 19466637.

6.2.23 INS-30131 Error When Installing Oracle Database or Oracle Client

If the administrative shares are not enabled when performing a single instance Oracle Database or Oracle Client installation for 12c Release 2 (12.2) on Microsoft Windows 7, Microsoft Windows 8, and Microsoft Windows 10, then the installation fails with an INS-30131 error.

Workaround:

Execute the net share command to ensure that the administrative shares are enabled. If they are disabled, then enable them by following the instructions in the Microsoft Windows documentation. Alternatively, perform the client or server installation by specifying the following options:

• For a client installation:

  -ignorePrereq

  -J"-Doracle.install.client.validate.clientSupportedOSCheck=false"

• For a server installation:

  -ignorePrereq -J"-Doracle.install.db.validate.supportedOSCheck=false"

This issue is tracked with Oracle bug 21452473.

6.3.1 Using Oracle ACFS for the Database Home and Large Object Files

When Oracle home is on Oracle ACFS, the installation location must be of the format, DRIVE_LETTER:\directory, that is, Oracle ACFS should be mounted on a new drive.

This issue is tracked with Oracle bug 9527303.

When installing or upgrading Oracle Database home on Oracle ACFS, Oracle Database Configuration Assistant and Oracle Database Upgrade Assistant fail to instantiate schemas that have large objects (for example, Database sample schemas).

Workaround:

Use SQL*Plus to manage LOB files when Oracle home is on Oracle ACFS.

This issue is tracked with Oracle bug 13702902.

6.3.2 Shortcuts Not Organized on Microsoft Windows Server 2012

The Start screen shows all the shortcuts together, that is, the shortcuts are not grouped appropriately.

This issue is tracked with Oracle bug 16023635.

6.3.3 ACFS Driver Fails to Load During Downgrade

When you downgrade from Oracle Database 12c Release 1 (12.1) to Oracle Database 11g Release 2 (11.2), the acfsroot install must run manually before starting the stack on each node after the downgrade. Otherwise, the ACFS driver fails to load.

Workaround 1:

To manually run the acfsroot install, set ORACLE_HOME to Oracle Grid Infrastructure home path for release 11.2, then run %ORACLE_HOME%\bin\acfsroot.bat install.

Workaround 2:

Run the following commands:

crsctl disable crs
reboot
set ORACLE_HOME=old Grid Home
old Grid Home\acfsroot install
crsctl enable crs
crsctl start crs

This issue is tracked with Oracle bug 19076502.

6.3.4 Oracle Home User Password Issue When Using Non-ASCII Characters with Command-Line Utilities on Non-English Locale

When you run command-line utilities, such as ORADIM, LSNRCTL, or Oracle home user tool, orahomeuserctl.bat on a non-English locale to provide Oracle home user password, any non-ASCII characters used in the password may enter as garbled text. Any non-ASCII data displayed on the screen may also appear as garbled text. This issue reproduces on computers that are using a non-English locale and only if Oracle Enterprise Manager code page is different from the local Microsoft Windows ANSI code page.

Workaround:

Perform the following steps:

1. Determine the system Microsoft Windows ANSI code page on the local computer by viewing the value of the following key in Microsoft Windows registry:

   HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\NLS\CodePage\ACP
   

2. Open a Command Prompt (console) window and run the chcp command:

   chcp Windows_ANSI_code_page
   

3. Set the font property for this Command Prompt window to a TrueType font:

   1. Right-click the title bar of the window and select Properties.

   2. Click Font and select Lucida Console from the list.

   3. Click OK.

4. Then, run the utility in the same Command Prompt window.

This issue is tracked with Oracle bug 16420596.

6.3.5 Oracle Enterprise Manager Database Express URL Issue

When you create an Oracle RAC, Oracle Enterprise Manager Express HTTPS URL may not work as the grid listeners do not have permission to access the wallet directory. The HTTP URL works fine.

Workaround:

Add permission for all the wallet files on XDB wallet folder by running the following command:

cacls.exe xdb_wallet path /T /E /G ORA_GRID_LISTENERS:F

If Oracle home user is set during an Oracle RAC installation, then the installation user may not have the permission to update the ACLs on wallet files. In that case, run the command with the specific Oracle home user.

This issue is tracked with Oracle bug 16394272.

6.3.6 Oracle VSS Writer Service Issue When Using Oracle Home User Password With Non-ASCII Characters

When Oracle Volume Shadow Copy Service (VSS) Writer is created on Non-English locale systems and Oracle home user password uses non-ASCII characters, the Writer service fails to start.

Workaround:

Use Services console or Oracle home user tool orahomeuserctl.bat to set the correct password for Oracle VSS Writer service log on account.

This issue is tracked with Oracle bug 19530427.

6.3.7 ORA-65026 Error When Unplugging a PDB

When unplugging a pluggable database (PDB), if the default Database Configuration Assistant (DBCA) template directory assistants/dbca/templates does not exist under ORACLE_BASE, unplugging the PDB fails with an ORA-65026 error.

Workaround:

Create the default DBCA template directory assistants/dbca/templates under ORACLE_BASE prior to unplugging a PDB.

This issue is tracked with Oracle bug 19457087.

6.3.8 Error When Plugging a PDB in Oracle Enterprise Manager Express

When plugging a pluggable database (PDB) on Microsoft Windows platforms, if you do not specify the name of the PDB, you may encounter an error.

Workaround:

Specify the name of the PDB and then proceed with plugging the PDB.

This issue is tracked with Oracle bug 19521689.

6.3.9 Adding Nodes on Oracle RAC Issue

When you add nodes to Oracle Real Application Clusters (Oracle RAC) home on Oracle Automatic Storage Management Cluster File System (Oracle ACFS), you may see the following error message:

setAclsForUserException_desc

Workaround:

Ignore the error and continue with the process to add nodes successfully.

This issue is tracked with Oracle bug 19512151.

6.3.10 Pro*C Not Supported with VS2013

On Windows platforms, Pro*C is not supported with Visual Studio 2013 (VS2013).

This issue is tracked with Oracle bug 19446251.

6.3.11 PDB Error on Second Node

When you attempt to open a pluggable database (PDB) on the second node after creating it on the first node using the storage SNAPSHOT COPY clause which clones a PDB, you may encounter the following errors:

ORA-00704: bootstrap process failure
ORA-01157: cannot identify/lock data file 48 - see DBWR trace file
ORA-01110: data file 48

Workaround:

Close the PDB which is created on the first node and then open it on the first or second node.

This issue is tracked with Oracle bug 19446095.

6.3.12 Database Fails to Start After Private NIC Failure of One Node in a Two Node Cluster

In a two node cluster, when there is a failure of the private NIC on the node that Oracle Grid Infrastructure Management Repository database is running on node A after node B is evicted, then when the private network is reestablished on node A, the other node B joins the cluster but the database instance on the other node is unable to start.

Workaround:

Perform the following steps:

1. Stop the crs stack on both nodes, then start the crs stack on both nodes.

2. Stop the database instance on the node that has the private network failure (node A), then start the database on both nodes. The database instance fails to start on node A this time. After the instance starts on the other node B, then you will be able to start the instance on node A.

This issue is tracked with Oracle bug 19642240.

6.3.13 Password File Authentication Error

On Microsoft Windows platforms, if the database system is configured with a password file based authentication which is not accessible, when you attempt to login as an administrative user to a closed pluggable database (PDB), you may encounter the following error:

ORA-600 [14005]

Workaround:

Perform the following steps to configure the password file:

1. Create a password file with Password Utility:

   C:\> orapwd FILE=PWDsid.ora ENTRIES=max_users
   

   • FILE specifies the password file name.

   • SID identifies the database instance.

   • ENTRIES sets maximum number of entries in the password file.

2. Set the initialization parameter file REMOTE_LOGIN_PASSWORDFILE to exclusive, shared, or none.

3. Start SQL*Plus:

   C:\> sqlplus /NOLOG
   

4. Connect AS SYSDBA:

   SQL> CONNECT / AS SYSDBA
   

   For an Oracle ASM instance, connect AS SYSASM:

   SQL> CONNECT / AS SYSASM
   

5. Start Oracle Database:

   SQL> STARTUP
   

6. Grant appropriate privileges to each user. Users who must perform database administration, for example, would be granted the SYSDBA privilege:

   SQL> GRANT SYSDBA TO db_administrator;
   

   For an Oracle ASM instance:

   SQL> GRANT SYSASM TO SYS;
   

   To test if your configuration is correct, disable the NTS authentication by temporarily setting the SQLNET.ORA parameter AUTHENTICATION_SERVICES to none. After the test completion, you can change the AUTHENTICATION_SERVICES to NTS.

This issue is tracked with Oracle bug 19076343.