Zero Downtime Migration Release Notes

These release notes provide downloading instructions for the latest product software and documentation, and describe new features, fixed bugs, known issues, and troubleshooting information for Zero Downtime Migration Release 21c (21.2).

What's New in This Release

Zero Downtime Migration Release 21c (21.2) includes the following new features.

  • Physical Migration Supports Direct Data Transfer

    You can migrate databases using direct data transfer during a physical migration to avoid backing up the source database to an intermediate store such as Object Storage or NFS. Recovery Manager (RMAN) allows Zero Downtime Migration to support active database duplication and restore from service.

    Direct Data Transfer Support

  • Reuse an Existing RMAN Backup as a Migration Source

    You can use an existing level 0 backup to skip the full backup phase of a migration job. Zero Downtime Migration takes level 0 and level 1 backups on the fly for both online and offline physical migration jobs. Now Zero Downtime Migration lets you re-use existing source database backup in place of performing a full back up.

    Run a Migration Job Using an Existing RMAN Backup

  • CPAT Lists Remedies and Can Create Remedy Scripts

    The Cloud Premigration Advisor Tool (CPAT) is integrated with Zero Downtime Migration, and can be used with logical migration jobs. CPAT can generate migration remedial scripts for failing checks that you can run if you choose.

    Using the Cloud Premigration Advisor Tool

  • Run CPAT Updates in Zero Downtime Migration Home

    Keep the Cloud Premigration Advisor Tool (CPAT) up to date to get the latest tool version.

    Updating the Cloud Premigration Advisor Tool

  • Migrate From Amazon Web Services RDS to Oracle Autonomous Database

    You can now migrate an Oracle Database from Amazon Web Services (AWS) RDS to Oracle Autonomous Database (ADB) using the Zero Downtime Migration offline logical migration method. To transfer the data from AWS, you can use an Amazon Simple Storage Service (Amazon S3) Bucket or a database link (DBLINK).

    Migrating from Amazon Web Services RDS to Oracle Autonomous Database

  • Migrate to Oracle Autonomous Database on Exadata Cloud@Customer

    Zero Downtime Migration supports migrations to Oracle Autonomous Database on Exadata Cloud@Customer from any on-premises Oracle Database, including existing Exadata Cloud@Customer systems, using the offline logical migration method and NFS as a data transfer medium.

    Migrating to Oracle Autonomous Database on Exadata Cloud@Customer

  • Platform Support Includes IBM AIX and Oracle Solaris

    In addition to Linux-x86-64, Zero Downtime Migration now supports IBM AIX and Oracle Solaris source environments for logical offline migration mode.

    Supported Platforms

  • Estimate Data Pump Dump Size with STATISTICS Method

    In logical migrations, Zero Downtime Migration estimates the Data Pump dump size using the BLOCKS method by default as part of the pre-check and as part of the actual migration part of phase ZDM_DATAPUMP_ESTIMATE_SRC. A new response file parameter, DATAPUMPSETTINGS_DATAPUMPPARAMETERS_ESTIMATEBYSTATISTICS, lets you choose the STATISTICS method to provide you with an estimated dump size.

    DATAPUMPSETTINGS_DATAPUMPPARAMETERS_ESTIMATEBYSTATISTICS

  • Automatically Create or Remap Tablespaces

    For logical migrations, Zero Downtime Migration can automatically discover the source database tablespaces associated with user schemas that are being migrated, and automatically create them in the target database before the Data Pump import phase.

    An important benefit is that you can create BIGFILE tablespaces on the target which significantly reduces the number of data files for your database. You can also enable AUTOEXTEND and specify the NEXT EXTENT size, and you can exclude specified tablespaces from being created automatically.

    In addition, with automatic remap enabled, Zero Downtime Migration discovers any source tablespaces that must be remapped, instead of pre-creating the tablespace on the target.

    Automatic Tablespace Creation and Automatic Tablespace Remap

  • Migration Customization Enhancements

    You can edit a user action script locally and update the script in the user action metadata. If you do not have a local copy of the script, you can now locate the cached user action script in the user action metadata using the ZDMCLI command query useraction, and the full base path to the script is displayed in the output.

    Modifying User Action Scripts

    You can now easily get data from one or more user actions and use that data to run another user action. For example, you can perform an action on the source database and use that data to perform an action on the target database.

    Using User Action Output in a User Action

Bugs Fixed

Zero Downtime Migration Release 21.2 incudes the bug fixes listed in the following table.

Table - Bugs Fixed In Zero Downtime Migration Release 21.2

Bug Number Description
32749606 ZDM PHYSICAL MIGRATION INTERMITTENTLY FAILS WITH PRGZ-3553 CHARACTER SET MISMATCH
33013588 ZDM: CONVERT NONCDB TO CDB FAILED - PRGO-1930 & PRCZ-2103
32808775 ZDM PHYSICAL MIGRATION WITH NONCDB TO PDB CONVERSION FOR SRC AS SINGLE INSTANCE DB ISSUES
33123365 DB_DOMAIN OVERRIDEN BY SOURCE SPFILE CONTENT
32767162 ZDM: MIGRATION FROM NON CDB TO CDB IS FAILING, WHEN TARGET IS VMDB AND BACKUP MEDIUM IS NFS
31507335 ZDM: OFFLINE LOGICAL MIGRATION - ABORT SHOULD CLEAN THE DBLINK CREATED
32100209 ZDM : EXPORT PHASE FAILS WHEN THE EXPORT DIRECTORY IS IN ASM
32960299 ZDM SVC PREFERRED INSTANCES ATTRIBUTE IS NOT GETTING MAPPED CORRECTLY
32447725 ZDM : ABORTED ZDM JOB NOT ABORTING EXPORT/IMPORT JOBS
32833720 LOGICAL MIGRATION WORKFLOW FAILING WITH STANDARD EDITION SOURCE DATABASES
33048353 ZDM HIT ERROR: ASMCMD-9459 USER TRIED TO MOVE OR COPY A PASSWORD FILE WITHIN THE SAME DISK GROUP
33032852 ZDM: ZDM_UTILS REPORTED COREDUMP ERRORS DUE TO CREDENTIAL DOES NOT EXIST
32464487 ZDM: OSS FEDERATED USER ACCOUNT FAILING ZDM_VALIDATE_SRC
32689878 UNABLE TO START ZDM SERVICE NODE AFTER UPGRADE FROM 19.7 TO 21.1 -> CRS_ERROR:TCC-0004: THE CONTAINER WAS NOT ABLE TO START.
32886747 LOGICAL MIGRATION: ZDM DOESN'T DETECT LONG RUNNING TRANSACTION DURING SWITCHOVER PHASE

Downloading the Zero Downtime Migration Installation Software

For a fresh installation of the latest Zero Downtime Migration software version, go to https://www.oracle.com/database/technologies/rac/zdm-downloads.html.

Downloading the Zero Downtime Migration Documentation

You can browse and download Zero Downtime Migration documentation at https://docs.oracle.com/en/database/oracle/zero-downtime-migration/

General Information

At the time of this release, there are some details and considerations about Zero Downtime Migration behavior that you should take note of.

Cloud Premigration Advisor Tool Support

Cloud Premigration Advisor Tool (CPAT) is supported with Zero Downtime Migration for the following use cases:

  • CPAT is run automatically on the source database environment during logical migration jobs from Oracle Cloud and on-premises Oracle Database sources (default behavior)

  • CPAT is run manually from a remote server against an Amazon Web Services RDS Oracle Database source; in other words, CPAT is not run by ZDMCLI migrate database (see Running CPAT with a Remote Connection)

CPAT is not supported in the following use cases:

  • Physical migration jobs

  • Generating fixup scripts for Amazon Web Services RDS Oracle Database sources

UNDO Tablespaces Added to the Source Database

Zero Downtime Migration adds UNDO tablespaces to the production database to match the target instance count if the production database has fewer instances.

To prevent Zero Downtime Migration from adding UNDO tablespaces to the source database, you can match the target database nodes count to that of the source database until the switchover, then you can add additional nodes to the target database after the switchover.

Cross-Edition Migration

Zero Downtime Migration cannot be used to migrate an Enterprise Edition database to a Standard Edition database. In the converse case, Standard Edition databases can be migrated to Enterprise Edition databases, but only using the logical migration work flow.

EXT3 File System Support

There is a known issue which prevents Zero Downtime Migration from being installed in EXT3 file systems. The root cause is MySQL bug 102384. This is not a limitation of Zero Downtime Migration, but a constraint from MySQL. When that bug is resolved, Zero Downtime Migration is expected to work on EXT3 file systems.

Known Issues

At the time of this release, the following are known issues with Zero Downtime Migration that could occur in rare circumstances. For each issue, a workaround is provided.

Logical Migration Using DBLINK Fails with PRGZ-1177

Issue: "PRGZ-1177 : Database link "dblink_name" is invalid and unusable" error causes failure in a logical migration using a database link in a PDB or multitenant database in version 12.1.0.x.

Solution: See 12c PDB or Multitenant Only: ORA-02085: Database Link "LINK_NAME_HERE" Connects To "TARGET_DB" (Doc ID 2344831.1)

PRGZ-1306 Thrown During Physical Migration from Single Instance Database with No Grid Infrastructure Configured

Issue: When attempting to migrate a single instance database with no Grid Infrastructure configured using the -sourcesid parameter, the migration job fails with error PRGZ-1306.

Solution: This is an issue that is being tracked with Bug 33308423. As a workaround, re-run the migration command substituting the -sourcesid option with the -sourcedb option. Otherwise do not perform the physical migration on a single instance database with no GI configured until the bug is fixed and applied to the Zero Downtime Migration server home.

ORA-01031 on Full Export from an Oracle 12.1 Source

Issue: When performing a full database export with Export Data Pump from an Oracle Database 12c (12.1) source database, the following errors occur:

05-AUG-21 10:36:12.483: ORA-31693: Table data object "SYS"."TABLE" failed to load/unload and is being skipped due to error: ORA-01031: insufficient privileges

Solution: See My Oracle Support document EXPDP - ORA-31693 ORA-01031 (Insufficient Privileges) On Some Tables When Exporting from 12cR1 (Doc ID 1676411.1)

Restore Fails When Source Uses WALLET_ROOT

Issue: Zero Downtime Migration does not currently handle the migration of the TDE wallet from the source database to the target when the source database is using the wallet_root initialization parameter. Without the wallets available on the target database, the restore phase fails with an error similar to the following: 

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of restore command at 06/15/2021 07:35:11
ORA-19870: error while restoring backup piece
/rman_PRD1/ZDM/IQPCZDM/c-3999816841-20210614-00
ORA-19913: unable to decrypt backup

Solution: Manually copy the wallet to the target and resume the job.

ORA-17503 on Dump Import from Amazon S3 to Autonomous Database Dedicated Infrastructure

Issue: When attempting to import a Data Pump dump file from an Amazon S3 bucket to an Oracle Autonomous Database Dedicated Infrastructure (ADB-D) target, an error occurs in ADB-D.

The ADB-D alert log shows an error similar to the following:

ORA-39000: bad dump file specification
ORA-31640: unable to open dump file
"https://s3-region.amazonaws.com/dump_file_path.dmp
for read
ORA-17503: target Failed to open file
https://s3-region.amazonaws.com/dump_file_path.dmp
ORA-17500: ODM err:ODM Client Error

Solution: Use the database link data transfer medium in Zero Downtime Migration to import Data Pump dumps to an ADB-D target. 

DATA_TRANSFER_MEDIUM=DBLINK
SOURCEDATABASE_ENVIRONMENT_NAME=AMAZON
SOURCEDATABASE_ENVIRONMENT_DBTYPE=RDS_ORACLE

PRCZ-4026 Thrown During Migration to Oracle Database 19.10 Target

Issue: When attempting to migrate to an Oracle Database 19.10 home at target, the migration job fails at phase ZDM_FINALIZE_TGT with error PRCZ-4026, because of Oracle Clusterware (OCW) Bug 31070231.

PRCZ-4026 : Resource ora.db_unique_name.db is already running on nodes node.

Solution: Apply the Backport Label Request (BLR) for Bug#32646135 to the target 19.10 dbhome to avoid the reported issue. Once the BLR is applied, you can resume the failed migration job to completion.

Precaution: For physical migrations, you can avoid this issue by ensuring that your target database home is not on Oracle Database 19.10.

ORA-39006 Thrown During Logical Migration to Autonomous Database Dedicated Infrastructure Over Database Link

Issue: When attempting to migrate a database to an Autonomous Database Dedicated Infrastructure target over a database link, the migration job fails with error ORA-39006.

ORA-39006: internal error

Solution: This is a Data Pump issue that is being tracked with Bug 31830685. Do not perform logical migrations over a database link to Autonomous Database Dedicated Infrastructure targets until the bug is fixed and the fix is applied to the Autonomous target database.

Zero Downtime Migration Service Fails To Start After Upgrade

Issue: The following scenario occurs:

  1. Perform migration jobs with Zero Downtime Migration 19.7

  2. Response files used in those jobs are removed

  3. Upgrade to Zero Downtime Migration 21.1

  4. Attempt to run a migration

The following errors are seen.

CRS_ERROR:TCC-0004: The container was not able to start.

CRS_ERROR:One or more listeners failed to start. Full details will be found in the appropriate container log fileContext [/rhp] startup failed due to previous errors sync_start failed with exit code 1.

A similar error is found in the log files located in zdm_installation_location/base/crsdata/hostname/rhp/logs/.

Caused by: oracle.gridhome.container.GHException: Internal error:PRGO-3003 : Zero downtime migration (ZDM) template file /home/jdoe/zdm_mydb.rsp does not exist.

Solution: To recover, manually recreate the response files listed in the log, and place them in the location specified in the log.

Environments With Oracle 11.2.0.4 Must Apply Perl Patch

Issue: Before using Zero Downtime Migration, you must apply a PERL patch if your source database environment meets either of the following conditions.

  • Clusterware environment with Oracle Grid Infrastructure 11.2.0.4
  • Single instance environment with Oracle Database 11.2.0.4

Solution: Download and apply Perl patch version 5.28.2 or later. Ensure that both the source and target Oracle Database 11g home include the patch for BUG 30508206 - UPDATE PERL IN 11.2.0.4 DATABASE ORACLE HOME TO V5.28.2.

Oracle GoldenGate Hub Certificate Known Issues

Issue: Oracle Zero Downtime Migration leverages Oracle GoldenGate for its logical online migration work flow; an Oracle GoldenGate hub is set up on OCI compute for this purpose.

The Oracle GoldenGate hub NginX Reverse Proxy uses a self-signed certificate which will cause the following error:

SunCertPathBuilderException: unable to find valid certification path to requested target when ZDM Server makes a REST API call.

Solution: See My Oracle Support document Zero Downtime Migration - GoldenGate Hub Certificate Known Issues (Doc ID 2768483.1)

Data Transfer Medium COPY Issues

Issue: Migrating data using logical migration with DATA_TRANSFER_MEDIUM=COPY set in the Zero Downtime Migration response file fails.

Solution: When you specify DATA_TRANSFER_MEDIUM=COPY you must also specify the following DUMPTRANSFERDETAILS_SOURCE_* parameters. 

DUMPTRANSFERDETAILS_TRANSFERTARGET_DUMPDIRPATH=<Target path to transferthe dumps to >
DUMPTRANSFERDETAILS_TRANSFERTARGET_HOST=<Target Db server or Target sidetransfer node >
DUMPTRANSFERDETAILS_TRANSFERTARGET_USER=<user having write access to specified path>
DUMPTRANSFERDETAILS_TRANSFERTARGET_USERKEY=<user authentication keypath on zdm node>

Source Discovery Does Not Find 'cut' in Default Location

Issue: Discovery at the source database server fails to find cut in the standard location.

The source database deployment's standard cut location is /bin/cut. If cut is not in the location, Zero Downtime Migration cannot discover the source database information correctly, and the migration fails in its initial phases.

Solution: To resolve the issue, ensure that cut is installed in the standard /bin/cut path or create a symbolic link to the installed location, for example:

ln -sf <installed_location_of_the_cut> /bin/cut

PRGG-1043 : No heartbeat table entries were found for Oracle GoldenGate Replicat process

Issue: An online logical migration job can report error PRGG-1043: No heartbeat table entries were found for Oracle GoldenGate Replicat process process_name due to one of the following causes:

  1. Initialization parameter job_queue_processes was set to zero in the source or target database.

    Solution: Run the following statements on the database: 

    show parameter job_queue_processes;
alter system set job_queue_processes=100 scope=both;
exec dbms_scheduler.set_scheduler_attribute('SCHEDULER_DISABLED','FALSE');

  2. Scheduled job GG_UPDATE_HEARTBEATS is not active in the source database.

  3. The server hosting Oracle GoldenGate deployments has a different time zone than the source database.

    Solution: First, do one of the following solutions:

    • Modify the time zone for the server hosting Oracle GoldenGate deployments, OR

    • Use the web UI for the Oracle GoldenGate deployment to add Extract parameter TRANLOGOPTIONS SOURCE_OS_TIMEZONE and restart Extract.

      For example, if the source database time zone is UTC-5, then set parameter TRANLOGOPTIONS SOURCE_OS_TIMEZONE -5. For more information, see TRANLOGOPTIONS in Reference for Oracle GoldenGate.

    Then, ensure that the DST_PRIMARY_TT_VERSION property in the source database is up to date.

Troubleshooting

If you run into issues, check here in case a solution is published. For each issue, a workaround is provided.

Connectivity Issues

General Connectivity Issues

Issue: If connectivity issues occur between the Zero Downtime Migration service host and the source or target environments, or between source and target environments, check the following areas.

Solution: Verify that the SSH configuration file (/root/.ssh/config) has the appropriate entries: 

Host *
  ServerAliveInterval 10
  ServerAliveCountMax 2

Host ocidb1
  HostName 192.0.2.1
  IdentityFile ~/.ssh/ocidb1.ppk
  User opc
  ProxyCommand /usr/bin/nc -X connect -x www-proxy.example.com:80 %h %p

Note that the proxy setup might not be required when you are not using a proxy server for connectivity. For example, when the source database server is on Oracle Cloud Infrastructure Classic, you can remove or comment the line starting with ProxyCommand.

If the source is an Oracle RAC database, then make sure you copy the ~/.ssh/config file to all of the source Oracle RAC servers. The SSH configuration file refers to the first Oracle RAC server host name, public IP address, and private key attributes.

Communications Link Failure

Issue: If the MySQL server crashes you will see errors such as this one for the ZDM operations: 

$ ./zdmcli query job -jobid 6
Exception [EclipseLink-4002] (Eclipse Persistence Services -
2.7.7.qualifier): org.eclipse.persistence.exceptions.DatabaseException
Internal Exception: com.mysql.cj.jdbc.exceptions.CommunicationsException:
Communications link failure
The last packet sent successfully to the server was 0 milliseconds ago. The
driver has not received any packets from the server.
Error Code: 0
Query: ReadAllQuery(referenceClass=JobSchedulerImpl sql="SELECT
JOB_IDENTIFIER, M_ACELIST, ARGUMENTS, ATTRIBUTES, CLIENT_NAME,
COMMAND_PROVIDED, COMPARTMENT, CONTAINER_TYPE, CREATEDATE, CREATOR,
CURRENT_STATUS, DB_OCID, DBNAME, DEPLOYMENT_OCID, DISABLE_JOB_EXECUTION,
ELAPSED_TIME, END_TIME, EXECUTE_PHASES, EXECUTION_TIME, IS_EVAL, IS_PAUSED,
JOB_TYPE, METHOD_NAME, METRICS_LOCATION, OPERATION, PARAMETERS,
PARENT_JOB_ID, PAUSE_AFTER_PHASE, RESULT, PHASE, JOB_SCHEDULER_PHASES,
REGION, REST_USER_NAME, RESULT_LOCATION, SCHEDULED_TIME, SITE, SOURCEDB,
SOURCENODE, SOURCESID, SPARE1, SPARE2, SPARE3, SPARE_A, SPARE_B, SPARE_C,
START_TIME, STOP_AFTER_PHASE, TARGETNODE, JOB_THREAD_ID, UPD_DATE, USER_NAME,
ENTITY_VERSION, CUSTOMER FROM JOBSCHEDULER WHERE (PARENT_JOB_ID = ?)")

Solution: If such Communications errors are seen, restart the Zero Downtime Migration service so that the MySQL server is restarted, after which the pending jobs will resume automatically.

Stop the Zero Downtime Migration service:

zdmuser> $ZDM_HOME/bin/zdmservice stop

Start the Zero Downtime Migration service:

zdmuser> $ZDM_HOME/bin/zdmservice start

Evaluation Fails in Phase ZDM_GET_TGT_INFO

Issue: During the evaluation (-eval) phase of the migration process, the evaluation fails in the ZDM_GET_TGT_INFO phase with the following error for the Oracle RAC instance migration. 

Executing phase ZDM_GET_TGT_INFO
Retrieving information from target node "trac11" ...
PRGZ-3130 : failed to establish connection to target listener from nodes [srac11, srac12]
PRCC-1021 : One or more of the submitted commands did not execute successfully.
PRCC-1025 : Command submitted on node srac11 timed out after 15 seconds.
PRCC-1025 : Command submitted on node srac12 timed out after 15 seconds.

Solution:

  1. Get the SCAN name of source database and add it to the /etc/hosts file on both target database servers, with the public IP address of the source database server and the source database SCAN name. For example: 
    192.0.2.3 source-scan
  2. Get the SCAN name of the target database and add it to the /etc/hosts file on both source database servers, with the public IP address of the target database server and target database SCAN name. For example:
    192.0.2.1  target-scan

Note:

This issue, where the SCAN IP address is not added to /etc/hosts file, might occur because in some cases the SCAN IP address is assigned as a private IP address, so it might not be resolvable.

Object Storage Is Not Accessible

Issue: When Object Storage is accessed from the source or target database server, it may fail with the following error.
About to connect() to swiftobjectstorage.xx-region-1.oraclecloud.com port 443 (#0)
Trying 192.0.2.1... No route to host
Trying 192.0.2.2... No route to host
Trying 192.0.2.3... No route to host
couldn't connect to host
Closing connection #0
curl: (7) couldn't connect to host

Solution: On the Zero Downtime Migration service host, in the response file template ($ZDM_HOME/rhp/zdm/template/zdm_template.rsp), set the Object Storage Service proxy host and port parameters listed below, if a proxy is required to connect to Object Storage from the source database server. For example: 

SRC_OSS_PROXY_HOST=www-proxy-source.example.com
SRC_OSS_PROXY_PORT=80

In the response file template ($ZDM_HOME/rhp/zdm/template/zdm_template.rsp), set the Object Storage Service proxy host and port parameters listed below, if a proxy is required to connect to Object Storage from the target database server. For example: 

TGT_OSS_PROXY_HOST=www-proxy-target.example.com
TGT_OSS_PROXY_PORT=80

SSH Error "EdDSA provider not supported"

Issue: The following error messages appear in $ZDM_BASE/crsdata/zdm service hostname/rhp/zdmserver.log.0. 

[sshd-SshClient[3051eb49]-nio2-thread-1] [ 2020-04-04 00:26:24.142 GMT ]
 [JSChChannel$LogOutputStream.flush:1520]  2020-04-04: WARNING: org.apache.sshd.client.session.C:
 globalRequest(ClientConnectionService[ClientSessionImpl[opc@samidb-db/140.238.254.80:22]])[hostkeys-00@openssh.com,
 want-reply=false] failed (SshException) to process: EdDSA provider not supported

[sshd-SshClient[3051eb49]-nio2-thread-1] [ 2020-04-04 00:26:24.142 GMT ]
 [JSChChannel$LogOutputStream.flush:1520]  2020-04-04: FINE   : org.apache.sshd.client.session.C:
 globalRequest(ClientConnectionService[ClientSessionImpl[opc@samidb-db/140.238.254.80:22]])[hostkeys-00@openssh.com,
 want-reply=false] failure details
org.apache.sshd.common.SshException: EdDSA provider not supported
    at org.apache.sshd.common.util.buffer.Buffer.getRawPublicKey(Buffer.java:446)
    at org.apache.sshd.common.util.buffer.Buffer.getPublicKey(Buffer.java:420)
    at org.apache.sshd.common.global.AbstractOpenSshHostKeysHandler.process(AbstractOpenSshHostKeysHandler.java:71)
    at org.apache.sshd.common.global.AbstractOpenSshHostKeysHandler.process(AbstractOpenSshHostKeysHandler.java:38)
    at org.apache.sshd.common.session.helpers.AbstractConnectionService.globalRequest(AbstractConnectionService.java:723)
    at org.apache.sshd.common.session.helpers.AbstractConnectionService.process(AbstractConnectionService.java:363)
    at org.apache.sshd.common.session.helpers.AbstractSession.doHandleMessage(AbstractSession.java:400)
    at org.apache.sshd.common.session.helpers.AbstractSession.handleMessage(AbstractSession.java:333)
    at org.apache.sshd.common.session.helpers.AbstractSession.decode(AbstractSession.java:1097)
    at org.apache.sshd.common.session.helpers.AbstractSession.messageReceived(AbstractSession.java:294)
    at org.apache.sshd.common.session.helpers.AbstractSessionIoHandler.messageReceived(AbstractSessionIoHandler.java:63)
    at org.apache.sshd.common.io.nio2.Nio2Session.handleReadCycleCompletion(Nio2Session.java:357)
    at org.apache.sshd.common.io.nio2.Nio2Session$1.onCompleted(Nio2Session.java:335)
    at org.apache.sshd.common.io.nio2.Nio2Session$1.onCompleted(Nio2Session.java:332)
    at org.apache.sshd.common.io.nio2.Nio2CompletionHandler.lambda$completed$0(Nio2CompletionHandler.java:38)
    at java.security.AccessController.doPrivileged(Native Method)
    at org.apache.sshd.common.io.nio2.Nio2CompletionHandler.completed(Nio2CompletionHandler.java:37)
    at sun.nio.ch.Invoker.invokeUnchecked(Invoker.java:126)
    at sun.nio.ch.Invoker$2.run(Invoker.java:218)
    at sun.nio.ch.AsynchronousChannelGroupImpl$1.run(AsynchronousChannelGroupImpl.java:112)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: java.security.NoSuchAlgorithmException: EdDSA provider not supported
    at org.apache.sshd.common.util.security.SecurityUtils.generateEDDSAPublicKey(SecurityUtils.java:596)
    at org.apache.sshd.common.util.buffer.keys.ED25519BufferPublicKeyParser.getRawPublicKey(ED25519BufferPublicKeyParser.java:45)
    at org.apache.sshd.common.util.buffer.keys.BufferPublicKeyParser$2.getRawPublicKey(BufferPublicKeyParser.java:98)
    at org.apache.sshd.common.util.buffer.Buffer.getRawPublicKey(Buffer.java:444)
    ... 22 more
[sshd-SshClient[3051eb49]-nio2-thread-1] [ 2020-04-04 00:26:24.142 GMT ]
 [JSChChannel$LogOutputStream.flush:1520]  2020-04-04: FINE   : org.apache.sshd.client.session.C:
 sendGlobalResponse(ClientConnectionService[ClientSessionImpl[opc@samidb-db/140.238.254.80:22]])[hostkeys-00@openssh.com]
 result=ReplyFailure, want-reply=false

[sshd-SshClient[3051eb49]-nio2-thread-2] [ 2020-04-04 00:26:24.182 GMT ]
 [JSChChannel$LogOutputStream.flush:1520]  2020-04-04: FINE   : org.apache.sshd.common.io.nio2.N:
 handleReadCycleCompletion(Nio2Session[local=/192.168.0.2:41198, remote=samidb-db/140.238.254.80:22])
 read 52 bytes

Solution: Zero Downtime Migration uses the RSA format.

Full Backup Phase (ZDM_BACKUP_FULL_SRC) Issues

Backup Fails with ORA-19836

Issue: Source database full backup fails with one of the following errors. 

</ERRLINE><ERRLINE>ORA-19836: cannot use passphrase encryption for this backup
</ERRLINE><ERRLINE>RMAN-03009: failure of backup command on C8 channel at 04/29/2019
      20:42:16
</ERRLINE><ERRLINE>ORA-19836: cannot use passphrase encryption for this backup
</ERRLINE><ERRLINE>RMAN-03009: continuing other job steps, job failed will not be
      re-run

Solution 1: This issue can occur if you specify the -sourcedb value in the wrong case. For example, if the value obtained from SQL command SHOW PARAMETER DB_UNIQUE_NAME is zdmsdb, then you need to specify it as zdmsdb in lower case, and not as ZDMSDB in upper case, as shown in the following example. 

zdmuser> $ZDM_HOME/bin/zdmcli migrate database -sourcedb zdmsdb -sourcenode ocicdb1 -srcroot
-targetnode ocidb1 -targethome /u01/app/oracle/product/12.1.0.2/dbhome_1
-backupuser backup_user@example.com -rsp /u01/app/zdmhome/rhp/zdm/template/zdm_template_zdmsdb.rsp
-tgtauth zdmauth -tgtarg1 user:opc
-tgtarg2 identity_file:/home/zdmuser/.ssh/zdm_service_host.ppk
-tgtarg3 sudo_location:/usr/bin/sudo

Solution 2: For Oracle Database 12c Release 1 and later, ensure that $ORACLE_HOME/network/admin/sqlnet.ora points to the correct location of the TDE wallet, as shown here. 

ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME)))

For Oracle Database 11g Release 2 (11.2.0.4) only, ensure that $ORACLE_HOME/network/admin/sqlnet.ora points to the correct location of the TDE wallet as shown below, and replace the variable $ORACLE_UNQNAME with the value obtained with the SQL statement SHOW PARAMETER DB_UNIQUE_NAME. 

ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME)))

For example:

SQL> show parameter db_unique_name
db_unique_name    string      oci112_region

ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/oci112_region)))

Solution 3: Run the following query and make sure that the wallet status is OPEN. 

SQL> select * from v$encryption_wallet
WRL_TYPE
-------------
WRL_PARAMETER
-------------
STATUS
-------------
file
/opt/oracle/dcs/commonstore/wallets/tde/abc_test
OPEN

Backup Fails with ORA-19914 and ORA-28365

Issue: Source database full backup fails with the following errors. 

channel ORA_SBT_TAPE_3: backup set complete, elapsed time: 00:00:15
channel ORA_SBT_TAPE_3: starting compressed full datafile backup set
channel ORA_SBT_TAPE_3: specifying datafile(s) in backup set
input datafile file number=00005 name=+DATA/ODA122/7312FA75F2B202E5E053050011AC5977/DATAFILE/system.382.1003858429
channel ORA_SBT_TAPE_3: starting piece 1 at 25-MAR-19
RMAN-03009: failure of backup command on ORA_SBT_TAPE_3 channel at 03/25/2019 19:09:30
ORA-19914: unable to encrypt backup
ORA-28365: wallet is not open
continuing other job steps, job failed will not be re-run
channel ORA_SBT_TAPE_3: starting compressed full datafile backup set
channel ORA_SBT_TAPE_3: specifying datafile(s) in backup set

Solution: Ensure that the wallet is opened in the database, and in case of CDB, ensure that the wallet is opened in the CDB, all PDBs, and PDB$SEED. See Setting Up the Transparent Data Encryption Wallet in the Zero Downtime Migration documentation for information about setting up TDE.

Either the Bucket Named Object Storage Bucket Name Does Not Exist in the Namespace Namespace or You Are Not Authorized to Access It

See Oracle Support Knowledge Base article "Either the Bucket Named '<Object Storage Bucket Name>' Does not Exist in the Namespace '<Namespace>' or You are not Authorized to Access it (Doc ID 2605518.1)" for the desciption and workarounds for this issue.

https://support.oracle.com/rs?type=doc&id=2605518.1

Restore Phase (ZDT_CLONE_TGT) Issues

Restore Database Fails With AUTOBACKUP does not contain an SPFILE

Issue: During the execution of phase ZDT_CLONE_TGT, restore database fails with the following error. 

channel C1: looking for AUTOBACKUP on day: 20200427
channel C1: AUTOBACKUP found: c-1482198272-20200427-12
channel C1: restoring spfile from AUTOBACKUP c-1482198272-20200427-12
channel C1: the AUTOBACKUP does not contain an SPFILE

The source database is running using init.ora file, but during the restore target phase, the database is trying to restore the server parameter file (SPFILE) from autobackup, therefore it fails.

Solution: Start the source database using an SPFILE and resubmit the migration job.

Restore Database Fails With ORA-01565

Issue: During the execution of phase ZDT_CLONE_TGT, restore database fails with the following error. 

</ERRLINE><ERRLINE>With the Partitioning, Real Application Clusters, Automatic Storage Management, OLAP
</ERRLINE><ERRLINE>and Real Application Testing options
</ERRLINE><ERRLINE>
</ERRLINE><ERRLINE>CREATE PFILE='/tmp/zdm833428275/zdm/PFILE/zdm_tgt_mclone_nrt139.pfile' FROM SPFILE
</ERRLINE><ERRLINE>*
</ERRLINE><ERRLINE>ERROR at line 1:
</ERRLINE><ERRLINE>ORA-01565: error in identifying file '?/dbs/spfile@.ora'
</ERRLINE><ERRLINE>ORA-27037: unable to obtain file status
</ERRLINE><ERRLINE>Linux-x86_64 Error: 2: No such file or directory
</ERRLINE><ERRLINE>Additional information: 3
</ERRLINE><ERRLINE>
</ERRLINE><ERRLINE>
</ERRLINE><ERRLINE>Disconnected from Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
</ERRLINE><ERRLINE>With the Partitioning, Real Application Clusters, Automatic Storage Management, OLAP

Solution: Start the target database using an SPFILE and resume the migration job.

Transparent Data Encryption Related Issues

Transparent Data Encryption General Information

Depending on your source database release, Transparent Data Encryption (TDE) wallet configuration may be required.

  • Oracle Database 12c Release 2 and later

    For Oracle Database 12c Release 2 and later releases, TDE wallet configuration is mandatory and must be enabled on the source database before migration begins.

    If TDE is not enabled, the database migration will fail.

    Upon restore, the database tablespaces are encrypted using the wallet.

  • Oracle Database 12c Release 1 and earlier

    On Oracle Database 12c Release 1 and Oracle Database 11g Release 2 (11.2.0.4), TDE configuration is not required.

For information about the behavior of TDE in an Oracle Cloud environment, see My Oracle Support document Oracle Database Tablespace Encryption Behavior in Oracle Cloud (Doc ID 2359020.1).

Job Fails in Phase ZDM_SETUP_TDE_TGT

Issue: The phase ZDM_SETUP_TDE_TGT fails with one of the following errors. 

Executing phase ZDM_SETUP_TDE_TGT
Setting up Oracle Transparent Data Encryption (TDE) keystore on the target node oci1121 ...
oci1121: <ERR_FILE><Facility>PRGZ</Facility><ID>ZDM_KEYSTORE_NOT_SETUP_ERR</ID><ARGS><ARG>oci112_phx1z3</ARG></ARGS></ERR_FILE>
PRGO-3007 : failed to migrate database "db11204" with zero downtime
PRCZ-4002 : failed to execute command "/u01/app/18.0.0.0/grid/perl/bin/perl" using the privileged execution plugin "zdmauth" on nodes "oci1121"
PRCZ-2103 : Failed to execute command "/u01/app/18.0.0.0/grid/perl/bin/perl" on node "oci1121" as user "root". Detailed error:
<ERR_FILE><Facility>PRGZ</Facility><ID>ZDM_KEYSTORE_NOT_SETUP_ERR</ID><ARGS><ARG>oci112_phx1z3</ARG></ARGS></ERR_FILE>

Error at target server in /tmp/zdm749527725/zdm/log/mZDM_oss_standby_setup_tde_tgt_71939.log
2019-06-13 10:00:20: Keystore location /opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME does not exists for database 'oci112_region'
2019-06-13 10:00:20: Reporting error:
<ERR_FILE><Facility>PRGZ</Facility><ID>ZDM_KEYSTORE_NOT_SETUP_ERR</ID><ARGS><ARG>oci112_region</ARG></ARGS></ERR_FILE>

Solution:

  • Oracle Database 12c Release 1 and later

    On the target database, make sure that $ORACLE_HOME/network/admin/sqlnet.ora points to the correct location of the TDE wallet. For exmaple: 

    ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME)
  • Oracle Database 11g Release 2 (11.2.0.4) only

    On the target database, make sure that $ORACLE_HOME/network/admin/sqlnet.ora points to the correct location of the TDE wallet, and replace the $ORACLE_UNQNAME variable with the value obtained from the SHOW PARAMETER DB_UNIQUE_NAME SQL command.

    For example, run

    SQL> show parameter db_unique_name
db_unique_name         string      oci112_region

    and replace

    ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME)))

    with

    ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/oci112_region)))

Post Migration Automatic Backup Issues

Troubleshooting Post Migration Automatic Backup Failures

Issue: Post migration, on the target database, Automatic Backup might fail.

You can verify the failure using the console in Bare Metal, VM and Exadata > DB Systems > DB System Details > Database Details > Backups.

Solution: Get the RMAN configuration settings from one of the following places.

  • Zero Downtime Migration documentation in Target Database Prerequisites, if captured
  • The log files at /opt/oracle/dcs/log/hostname/rman/bkup/db_unique_name/
  • /tmp/zdmXXX/zdm/zdm_TDBNAME_rman.dat

For example, using the second option, you can get the RMAN configuration settings from /opt/oracle/dcs/log/ocidb1/rman/bkup/ocidb1_abc127/rman_configure*.log, then reset any changed RMAN configuration settings for the target database to ensure that automatic backup works without any issues.

If this workaround does not help, then debug further by getting the RMAN job ID by running the DBCLI command, list-jobs, and describe the job details for more error details by running the DBCLI command describe-job -i JOB ID from the database server as the root user.

For example, during the test, the following highlighted settings were modified to make Automatic Backup work.

rman target /
Recovery Manager: Release 12.2.0.1.0 - Production on Mon Jul 8 11:00:18 2019
Copyright (c) 1982, 2017, Oracle and/or its affiliates.  All rights reserved.
connected to target database: ORCL (DBID=1540292788)
RMAN> show all;
using target database control file instead of recovery catalog
RMAN configuration parameters for database with db_unique_name OCIDB1_ABC127 are:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 30 DAYS;
CONFIGURE BACKUP OPTIMIZATION OFF;
CONFIGURE DEFAULT DEVICE TYPE TO DISK; # default
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE SBT_TAPE TO '%F'; # default
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '%F'; # default
CONFIGURE DEVICE TYPE 'SBT_TAPE' PARALLELISM 4 BACKUP TYPE TO COMPRESSED BACKUPSET;
CONFIGURE DEVICE TYPE DISK PARALLELISM 1 BACKUP TYPE TO BACKUPSET; # default
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE SBT_TAPE TO 1; # default
CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE SBT_TAPE TO 1; # default
CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1; # default
CONFIGURE CHANNEL DEVICE TYPE DISK MAXPIECESIZE 2 G;
CONFIGURE CHANNEL DEVICE TYPE 'SBT_TAPE' MAXPIECESIZE 2 G FORMAT '%d_%I_%U_%T_%t' PARMS
 'SBT_LIBRARY=/opt/oracle/dcs/commonstore/pkgrepos/oss/odbcs/libopc.so ENV=(OPC_PFILE=/opt/oracle/dcs/commonstore/objectstore/opc_pfile/1245080042/opc_OCIDB1_ABC127.ora)';
CONFIGURE MAXSETSIZE TO UNLIMITED; # default
CONFIGURE ENCRYPTION FOR DATABASE ON;
CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
CONFIGURE COMPRESSION ALGORITHM 'MEDIUM' AS OF RELEASE 'DEFAULT' OPTIMIZE FOR LOAD TRUE;
CONFIGURE RMAN OUTPUT TO KEEP FOR 7 DAYS; # default
CONFIGURE ARCHIVELOG DELETION POLICY TO BACKED UP 1 TIMES TO 'SBT_TAPE';
CONFIGURE SNAPSHOT CONTROLFILE NAME TO '+RECO/ OCIDB1_ABC127/controlfile/snapcf_ocidb1_abc127.f';
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK clear;
RMAN>

Post Migration Automatic Backup Fails With DCS-10045

Issue: Post migration, Automatic Backup fails with the following error for non-TDE enabled migrated Oracle Database releases 11.2.0.4 and 12.1.0.2. 

DCS-10045: Validation error encountered: Backup password is mandatory to take OSS backup for non-tde enabled database...

You can verify this error by getting the RMAN job ID by running DBCLI command list-jobs, and describe the job details to get the error details by running DBCLI command describe-job -i JOB ID from the database server as the root user.

Solution:

  1. Find the TDE wallet location.

    The Oracle Cloud Infrastructure provisioned database instance will have following entry in sqlnet.ora. 

    ENCRYPTION_WALLET_LOCATION=(SOURCE=(METHOD=FILE)(METHOD_DATA=(DIRECTORY=/opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME)))
  2. Remove the cwallet.sso file from the wallet location.

    For example, /opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME.

  3. For Oracle Database 11g Release 2, do the folowing steps.
    1. Connect to database using SQL*Plus as sysdba and verify the current wallet location.
      SQL> select * from v$encryption_wallet;
WRL_TYPE    WRL_PARAMETER                                            STATUS
file        /opt/oracle/dcs/commonstore/wallets/tde/ocise112_region  OPEN
    2. Close the wallet in the database.
      SQL> alter system set wallet close;
    3. Open the wallet using the wallet password.
      SQL> alter system SET WALLET open IDENTIFIED BY "walletpassword"
    4. Set the master encryption key.
      SQL> alter system set encryption key identified by "walletpassword"
    5. Recreate the autologin SSO file.
      /home/oracle>orapki wallet create -wallet /opt/oracle/dcs/commonstore/wallets/tde/$ORACLE_UNQNAME -auto_login
Oracle PKI Tool : Version 11.2.0.4.0 - Production
Copyright (c) 2004, 2013, Oracle and/or its affiliates. All rights reserved.
Enter wallet password:            #
    6. Retry Automatic Backup.
  4. For Oracle Database 12c, do the folowing steps.
    1. Connect to database using SQL*Plus as sysdba and verify the current wallet location and status.
      SQL> SELECT wrl_parameter, status, wallet_type FROM v$encryption_wallet;
WRL_PARAMETER                                            STATUS              WALLET_TYPE
/opt/oracle/dcs/commonstore/wallets/tde/ocise112_region  OPEN_NO_MASTER_KEY  OPEN

      If the STATUS column contains a value of OPEN_NO_MASTER_KEY, you must create and activate the master encryption key.

    2. Close the wallet in the database.
      SQL> alter system set wallet close;
    3. Open the wallet-using password.
      SQL> ADMINISTER KEY MANAGEMENT SET KEYSTORE open IDENTIFIED BY "walletpassword" CONTAINER=all;
    4. Set the master encryption key.
      SQL> ADMINISTER KEY MANAGEMENT SET KEY IDENTIFIED BY "walletpassword" with backup;

      Log in to each PDB and run

      SQL> ALTER SESSION SET CONTAINER = PDB_NAME;
SQL> ADMINISTER KEY MANAGEMENT SET KEY IDENTIFIED BY "walletpassword" with backup;
    5. Create the auto login keystore.
      SQL> ADMINISTER KEY MANAGEMENT CREATE AUTO_LOGIN KEYSTORE FROM KEYSTORE 'path to wallet directory' IDENTIFIED BY "walletpassword";
    6. Retry Automatic Backup.

Post Migration Automatic Backup Fails With DCS-10096

Issue: Post migration, Automatic Backup fails with the following error. 

DCS-10096:RMAN configuration 'Retention policy' must be configured as 'configure retentio n
      policy to recovery window of 30 days'

You can verify this error by getting the RMAN job ID by running DBCLI command list-jobs, and describe the job details for more error details by running DBCLI command describe-job -i JOB ID from the database server as the root user.

Solution: Log in into RMAN prompt and configure the retention policy. 

[oracle@racoci1 ~]$ rman target /
Recovery Manager: Release 12.2.0.1.0 - Production on Wed Jul 17 11:04:35 2019
Copyright (c) 1982, 2017, Oracle and/or its affiliates.  All rights reserved.
connected to target database: SIODA (DBID=2489657199)
RMAN> CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 30 DAYS;

old RMAN configuration parameters:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 7 DAYS;

new RMAN configuration parameters:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 30 DAYS;

new RMAN configuration parameters are successfully stored

Retry Automatic Backup.

Miscellaneous Issues

INS-42505 Warning Shown During Installation

Issue: The following warning is shown during installation.
/stage/user/ZDM_KIT_relnumber>./zdminstall.sh setup
oraclehome=/stage/user/grid oraclebase=/stage/user/base
ziploc=/stage/user/ZDM_KIT_relnumber/rhp_home.zip -zdm
---------------------------------------
Unzipping shiphome to gridhome
---------------------------------------
Unzipping shiphome...
Shiphome unzipped successfully..
---------------------------------------
##### Starting GridHome Software Only Installation #####
---------------------------------------
Launching Oracle Grid Infrastructure Setup Wizard...

[WARNING] [INS-42505] The installer has detected that the Oracle Grid
Infrastructure home software at (/stage/user/grid) is not complete.
   CAUSE: Following files are missing:
...

Solution: This warning message can be ignored. It does not affect the installation or cause any issues for migration.

Evaluation Fails in Phase ZDM_GET_SRC_INFO

Issue: During the evaluation (-eval) phase of the migration process, the evaluation fails in the ZDM_GET_SRC_INFO phase with the following error for the source single instance deployed without Grid infrastructure. 

Executing phase ZDM_GET_SRC_INFO
retrieving information about database "zdmsidb" ...
PRCF-2056 : The copy operation failed on node: "zdmsidb".
Details: {1}
PRCZ-4002 : failed to execute command "/bin/cp" using the privileged
execution plugin "zdmauth" on nodes "zdmsidb"
scp: /etc/oratab: No such file or directory

Solution: Make an ORACLE_HOME value entry in file /etc/oratab with value db_name:$ORACLE_HOME:N, as shown in this example. 

zdmsidb:/u01/app/oracle/product/12.2.0.1/dbhome_1:N

Unable to Rerun MIGRATE DATABASE Command

Issue: Zero Downtime Migration blocks attempts to rerun the MIGRATE DATABASE command for a specified database if that database is already part of an ongoing migration job.

Workaround: If you want to resubmit a database migration, you can stop the ongoing migration job in either EXECUTING or PAUSED state using the ZDMCLI ABORT JOB command as follows.
-bash-4.2$ ./zdmcli abort job -jobid 70
server.example.com: Audit ID: 189

Migration Job Fails at ZDM_GET_SRC_INFO

Issue: A migration job fails with the following error. 

[opc@zdm-server rhp]$ cat /home/opc/zdm_base/chkbase/scheduled/job-34-2021-01-23-14:10:32.log
zdm-server: 2021-01-23T14:10:32.155Z : Processing response file ...
zdm-server: 2021-01-23T14:10:32.262Z : Starting zero downtime migrate operation ...
PRCZ-4002 : failed to execute command "/bin/cp" using the privileged execution plugin "zdmauth" on nodes "PROD.compute-usconnectoneb95657.oraclecloud.internal"

Solution: You must set up SSH connectivity without a passphrase for the oracle user.

Migration Evaluation Failure with Java Exception Invalid Key Format

Issue: The following conditions are seen:

  • Zero Downtime Migration migration -eval command fails with the following error. 

    Result file path contents:
"/u01/app/zdmbase/chkbase/scheduled/job-19-2019-12-02-03:46:19.log"
zdm-server.ocitoolingsn.ocitooling.oraclevcn.com: Processing response
file ...
null

  • The file $ZDM_BASE/<zdm service host>/rhp/rhpserver.log.0 contains the following entry. 

    Verify below error message observed in file $ZDM_BASE/<zdm service
host>/rhp/rhpserver.log.0
rhpserver.log.7:[pool-58-thread-1] [ 2019-12-02 02:08:15.178 GMT ]
[JSChChannel.getKeyPair:1603]  Exception :
java.security.spec.InvalidKeySpecException:
java.security.InvalidKeyException: invalid key format

  • The Zero Downtime Migration installed user (For example: zdmuser) private key (id_rsa) file has the following entries.

    -----BEGIN OPENSSH PRIVATE KEY----------
MIIEogIBAAKCAQEAuPcjftR6vC98fAbU4FhYVKPqc0CSgibtMSouo1DtQ06ROPN0
XpIEL4r8nGp+c5GSDONyhf0hiltBzg0fyqyurSw3XfGJq2Q6EQ61aL95Rt9CZh6b
JSUwc69T4rHjvRnK824k4UpfUIqafOXb2mRgGVUkldo4yy+pLoGq1GwbsIYbS4tk
uaYPKZ3A3H9ZA7MtZ5M0sNqnk/4Qy0d8VONWozxOLFC2A8zbbe7GdQw9khVqDb/x
END OPENSSH PRIVATE KEY-----

Solution: Authentication key pairs (private and public key) are not generated using the ssh-keygen utility, so you must generate authentication key pairs using steps in Generating a Private SSH Key Without a Passphrase.

After generating authentication key pairs, the private key file content looks like the following.

-----BEGIN RSA PRIVATE KEY-----
MIIEogIBAAKCAQEAuPcjftR6vC98fAbU4FhYVKPqc0CSgibtMSouo1DtQ06ROPN0
XpIEL4r8nGp+c5GSDONyhf0hiltBzg0fyqyurSw3XfGJq2Q6EQ61aL95Rt9CZh6b
JSUwc69T4rHjvRnK824k4UpfUIqafOXb2mRgGVUkldo4yy+pLoGq1GwbsIYbS4tk
uaYPKZ3A3H9ZA7MtZ5M0sNqnk/4Qy0d8VONWozxOLFC2A8zbbe7GdQw9khVqDb/x
-----END RSA PRIVATE KEY-----

Set up connectivity with the newly generated authentication key pairs and resume the migration job.

Migration Evaluation Fails with Error PRCG-1022

Issue: The following conditions are seen: 

$ZDM_HOME/bin/zdmcli migrate database -sourcedb zdmsdb -sourcenode ocicdb1 
-srcauth zdmauth -srcarg1 user:opc 
-srcarg2 identity_file:/home/zdmuser/.ssh/zdm_service_host.ppk 
-srcarg3 sudo_location:/usr/bin/sudo -targetnode ocidb1 -backupuser backup_user@example.com 
-rsp /u01/app/zdmhome/rhp/zdm/template/zdm_template_zdmsdb.rsp -tgtauth zdmauth 
-tgtarg1 user:opc -tgtarg2 identity_file:/home/zdmuser/.ssh/zdm_service_host.ppk 
-tgtarg3 sudo_location:/usr/bin/sudo -eval

PRCG-1238 : failed to execute the Rapid Home Provisioning action for command  'migrate database'
PRCG-1022 : failed to connect to the Rapid Home Provisioning daemon for cluster anandutest
Failed to retrieve RMIServer stub: javax.naming.ServiceUnavailableException
[Root exception is java.rmi.ConnectException: Connection refused to host:
anandutest; nested exception is: java.net.ConnectException: Connection refused (Connection refused)]

Solution: Start the Zero Downtime Migration service using the $ZDM_HOME/bin/zdmservice START command, then run any ZDMCLI commands.

Unable to Resume a Migration Job

Issue: Zero Downtime Migration writes the source and target log files to the /tmp/zdm-unique id directory in the respective source and target database servers.

If you pause a migration job and and then resume the job after several (sometimes 15-20 days), the /tmp/zdm-unique id directory might be deleted or purged as part of a clean up or server reboot that also cleans up /tmp.

Solution: After pausing a migration job, back up the /tmp/zdm-unique id directory. Before resuming the migration job, check the /tmp directory for /zdm-unique id, and if it is missing, restore the directory and its contents with your backup.

Migration Job Fails at ZDM_SWITCHOVER_SRC

Issue: A migration job fails at ZDM_SWITCHOVER_SRC phase.

Solutions:

  1. Ensure that there is connectivity from PRIMARY database nodes to STANDBY database nodes so the redo log are shipped as expected.

  2. A job will fail at ZDM_SWITCHOVER_SRC if the recovery process (MRP0) is not running at the target. The recovery process reason for failure should be corrected if MRP0 is not running at Oracle Cloud Database Standby Instance, and then the process should be started manually at Oracle Cloud Database Standby Instance before the migration job can be resumed.

Additional Information for Migrating to Exadata Cloud Service

Read the following for general information, considerations, and links to more information about using Zero Downtime Migration to migrate your database to Exadata Cloud Service.

Considerations for Migrating to Exadata Cloud Service

For this release of Zero Downtime Migration be aware of the following considerations.

  • If the source database is release 18c, then the target home should be at release 18.6 or later to avoid issues such as Bug 29445548 Opening Database In Cloud Environment Fails With ORA-600.
  • If a backup was performed when one of the configured instances is down, you will encounter Bug 29863717 - DUPLICATING SOURCE DATABASE FAILED BECAUSE INSTANCE 1 WAS DOWN.
  • The TDE keystore password must be set in the credential wallet. To set the password as part of the Zero Downtime Migration workflow, specify the -tdekeystorewallet tde_wallet_path or -tdekeystorepasswd argument irrespective of whether the wallet uses AUTOLOGIN or PASSWORD. In either case the password is stored in the credential wallet. If the -tdekeystorepasswd argument is not supplied, then Zero Downtime Migration skips the setting tde_ks_passwd key in the credential wallet, and no error is thrown.
  • The target environment must be installed with latest DBaaS Tooling RPM with db_unique_name change support to be installed.
  • Provision a target database from the console without enabling auto-backups. In the Configure database backups section do not select the Enable automatic backups option.

Exadata Cloud Service Database Registration

Post migration, register the Exadata Cloud Service database, and make sure its meets all of the requirements.

Run the following commands on the Exadata Cloud Service database server as the root user.

/root>dbaascli registerdb prereqs --dbname db_name --db_unique_name db_unique_name

/root>dbaascli registerdb begin  --dbname db_name --db_unique_name db_unique_name

For example

/root>dbaascli registerdb prereqs --dbname ZDM122 --db_unique_name ZDM122_phx16n
DBAAS CLI version 18.2.3.2.0
Executing command registerdb prereqs --db_unique_name ZDM122_phx16n
INFO: Logfile Location: /var/opt/oracle/log/ZDM122/registerdb/registerdb_2019-08-14_05:35:31.157978280334.log
INFO: Prereqs completed successfully
/root>

/root>dbaascli registerdb begin --dbname ZDM122 --db_unique_name ZDM122_phx16n
DBAAS CLI version 18.2.3.2.0
Executing command registerdb begin --db_unique_name ZDM122_phx16n
Logfile Location: /var/opt/oracle/log/ZDM122/registerdb/registerdb_2019-08-14_05:45:27.264851309165.log
Running prereqs
DBAAS CLI version 18.2.3.2.0
Executing command registerdb prereqs --db_unique_name ZDM122_phx16n
INFO: Logfile Location: /var/opt/oracle/log/ZDM122/registerdb/registerdb_2019-08-14_05:45:29.000432309894.log
INFO: Prereqs completed successfully
Prereqs completed
Running OCDE .. will take time ..
OCDE Completed successfully.
INFO: Database ZDM122 registered as Cloud database
/root>

Exadata Cloud Service Automatic Backup Issues

Check the backup configuration before you enable automatic backup from the console. You can use the get config command as shown in the first step below. You should see bkup_oss=no before you enable automatic backup.

You might see the error message in the console, "A backup configuration exists for this database. You must remove the existing configuration to use Oracle Cloud Infrastructure's managed backup feature."

To fix this error, remove the existing configuration.

First, make sure the automatic backup is disabled from the UI, then follow these steps to remove the existing backup configuration.

  1. Generate a backup configuration file.
    /var/opt/oracle/bkup_api/bkup_api get config --file=/tmp/db_name.bk --dbname=db_name

    For example:

    /var/opt/oracle/bkup_api/bkup_api get config --file=/tmp/zdmdb.bk --dbname=zdmdb
  2. Open the /tmp/db_name.bk file you created in the previous step.

    For example: Open /tmp/zdmdb.bk

    change bkup_oss=yes from bkup_oss=no

  3. Disable OSS backup by setting bkup_oss=no.
    /var/opt/oracle/bkup_api/bkup_api set config --file=/tmp/db_name.bk --dbname=db_name

    For example:

    /var/opt/oracle/bkup_api/bkup_api set config --file=/tmp/zdmdb.bk --dbname=zdmdb
  4. Check reconfigure status.
    /var/opt/oracle/bkup_api/bkup_api configure_status --dbname=db_name

    For example:

    /var/opt/oracle/bkup_api/bkup_api configure_status --dbname=zdmdb

Now enable automatic backup from console.

Verify the backups from the console. Click Create Backup to create a manual backup, and a backup should be created without any issues. and also Automatic Backup should be successful.

Additional Information for Migrating to Exadata Cloud at Customer

Read the following for general information, considerations, and links to more information about using Zero Downtime Migration to migrate your database to Exadata Cloud at Customer.

Considerations for Migrating to Exadata Cloud at Customer

For this release of Zero Downtime Migration be aware of the following considerations.

  • You must apply the regDB patch for Bug 29715950 - "modify regdb to handle db_unique_name not same as db_name" on all Exadata Cloud at Customer nodes. This is required for the ZDM_MANIFEST_TO_CLOUD phase. Please note that the regDB tool is part of DBaaS Tooling.
  • If the source database is release 18c, then the target home should be at release 18.6 or later to avoid issues such as Bug 29445548 Opening Database In Cloud Environment Fails With ORA-600.
  • PDB conversion related phases are listed in -listphases and can be ignored. Those are no-op phases.
  • If the backup medium is Zero Data Loss Recovery Appliance, then all configured instances should be up at the source when a FULL or INCREMENTAL backup is performed.
  • If a backup was performed when one of the configured instances is down, you will encounter Bug 29863717 - DUPLICATING SOURCE DATABASE FAILED BECAUSE INSTANCE 1 WAS DOWN.
  • The TDE keystore password must be set in the credential wallet. To set the password as part of the Zero Downtime Migration workflow, specify the -tdekeystorewallet tde_wallet_path or -tdekeystorepasswd argument irrespective of whether the wallet uses AUTOLOGIN or PASSWORD. In either case the password is stored in the credential wallet. If the -tdekeystorepasswd argument is not supplied, then Zero Downtime Migration skips the setting tde_ks_passwd key in the credential wallet, and no error is thrown.
  • The target environment must be installed with latest DBaaS Tooling RPM with db_unique_name change support to be installed.

Documentation Accessibility

For information about Oracle's commitment to accessibility, visit the Oracle Accessibility Program website at http://www.oracle.com/pls/topic/lookup?ctx=acc&id=docacc.

Access to Oracle Support

Oracle customers that have purchased support have access to electronic support through My Oracle Support. For information, visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=info or visit http://www.oracle.com/pls/topic/lookup?ctx=acc&id=trs if you are hearing impaired.