Known Issues

Fluentd does not support deferred upgrades when upgrading to 6.0.13.0, and beyond, when starting from a prior release and using the Historic Splunk Insight solution.

DLPX-57971

Oracle SnapSync fails with "RMAN-06183: datafile or datafile copy <datafile> (file number <fileno>) larger than MAXSETSIZE" if a datafile resized in the middle of SnapSync.

One of the following three workarounds can be applied:

1. Re-issue SnapSync.

2. Reduce the frequency of datafile resizes, or

3. Ensure datafile resize operations are not being performed while the SnapSync operation is in progress.

These settings can still be configured via the CLI. This issue will be fixed in the next version release.

No workaround.

DLPX-75517

One of the following two workarounds can be applied:

1. Re-attempt the vPDB provision to a linked CDB. The provision should succeed.
2. Manually create (or rename) the TEMP tablespace in the production PDB$SEED database, take a new snapshot of the vPDB, and provision from that snapshot to the vCDB.

DLPX-64307

The following workaround should resolve the issue without Delphix support intervention.

For RAC databases

1. Add the instances to the clusterware configuration, ensuring you add all instances configured via Delphix. For example:

srvctl add instance -db <Database Name> -instance <Instance Name> -node <Node Name>

2. Refresh the environment. At this point, Delphix will add the instances back to the Delphix configuration, and you should be able to perform actions such as stop, start, disable, enable, etc.

From here, we recommend removing the clusterware configuration.

3. Stop the VDB via the Delphix GUI or CLI.
4. Remove the instances from clusterware

srvctl remove instance -db <Database Name> -instance <Instance Name>

5. Remove the database from clusterware

srvctl remove database -db <Database Name>

6. Refresh the environment.
7. Start the VDB via the Delphix GUI or CLI.

For RACOne databases

1. Stop the instance.

sqlplus "/as sysdba"
shutdown immediate

2. Ensure the RACOne configuration has the "instance prefix" configuration set.

[oracle@mwtestx1 ~]$ srvctl config database -d <vdb name> | grep "Instance name prefix"
Instance name prefix: VDB

3. If this is not set, then set it using

srvctl modify database -db <vdb name> -instance <instance prefix>

4. Set the pfile location to allow clusterware to start the instance

srvctl modify database -db <vdb name> -spfile <path to pfile>

The pfile is located in

<Delphix Mount base>/<VDB Name>/datafile/spfile.ora

5. Start the instance using clusterware

srvctl start database -d vdb

When querying the status of the database, it should now show a running instance. As an example:

[oracle@mwtestx1 ~]$  srvctl status database -d VDB
Instance VDB_1 is running on node mwtestx1
Online relocation: INACTIVE

6. Refresh the environment. At this point, Delphix will add the instances back to the Delphix configuration, and you should be able to perform actions such as stop, start, disable, enable, etc.

From here, we recommend removing the clusterware configuration.

7. Stop the VDB via the Delphix GUI or CLI.
8. Remove the database from clusterware

srvctl remove database -db <Database Name>

9. Remove the database from clusterware

srvctl remove database -db <Database Name>

10. Refresh the environment.
11. Start the VDB via the Delphix GUI or CLI.

Follow these steps:

1. Restart the listener without a name parameter so it displays in uppercase.

2. Update VDBs to use uppercase LISTENER.

3. Refresh the environment. lowercase listener will be removed and LISTENER updated appropriately.

DLPX-74860

DLPX-75517

One of the following two workarounds can be applied:

1. Re-attempt the vPDB provision to a linked CDB. The provision should succeed.
2. Manually create (or rename) the TEMP tablespace in the production PDB$SEED database, take a new snapshot of the vPDB and provision from that snapshot to the vCDB.

Switch the default login shell for the Delphix OS user from /bin/sh to /bin/bash.

To prevent the issue:  Ensure that the vPDB is either closed/mounted or open unrestricted when exiting from dx-post-plug-hook.sh.

After it has happened: close and reopen the vPDB.

To prevent the issue:  Ensure that the vPDB is either closed/mounted or open unrestricted when exiting from dx-post-plug-hook.sh.

After it has happened: close and reopen the vPDB.

Verify that the CDB is restarted after merging the keys of the old target CDB into the new target CDB and re-attempt the rewind operation.

If the rewind still fails, provision a new vPDB from the required snapshot instead of the rewind operation.

The latest replica VDB data can still be accessed by doing a Self-Service container Refresh, rather than a point-in-time restore from the template.

DLPX-61079

Certificate import validation may incorrectly reject a root CA certificate

Support must manually import the certificate into the truststore.

DLPX-61405

Masking operation should wait for zfs delete queue to drain

Replication may send more data than expected if masking involves dropping large DBF files.

DLPX-64493

V5 API /roles endpoint missing certain items

View and set these privileges through the GUI

DLPX-66155

Failed DSP engine test leads to multiple blocked client.jar processes on target hosts.

Restarting a Delphix Engine during a network throughput test is not recommended as it may lead to a system hang.

DLPX-66860

ADFS does not like NameIDPolicy sent by SSO app

Create an explicit rule in ADFS that transforms the {{emailAddress}} attribute into a {{nameid}}. The rule type must be "Transform an incoming claim". The incoming claim type must be "Email address" and the outgoing claim type "Name ID". The nameid format must be "Email address".

DLPX-66973

Date format is changed after importing the environment

Either (a) use the GUI import feature and then review the imported date formats for correctness or (b) use EngineSync to export/import jobs, which will not alter the date format.