Known issues and limitations for Data Privacy (Masking flow)

When upgrading, you may need to run patches if the horizontal pod autoscaling (HPA) feauture is enabled

Applies to: 5.2.0 and later

If you have the horizontal pod autoscaling (HPA) feature enabled and you are upgrading to 5.1.x, you will need to run the following patch commands:

oc patch -n ${cpd_instance_ns} crd dp.dp.cpd.ibm.com --type='json' -p '[{"op": "replace", "path": "/spec/versions/0/schema/openAPIV3Schema/properties/spec/properties/autoScaleConfig/type", "value": "string"}, {"op": "replace", "path": "/spec/versions/0/schema/openAPIV3Schema/properties/spec/properties/autoScaleConfig/default", "value": "false"}]'

oc patch -n ${cpd_instance_ns} dp dp-cr -p '{"spec":{"autoScaleConfig":"false"}}' --type=merge

Where {cpd_instance_ns} is the project where Data Privacy is installed.

CreateContainerError cause problems to run Bulk copy jobs

Applies to: 5.2.0

If you're using a CSI-based storage type with volumes of large number of files, consider disabling SELinux relabeling. SELinux relabeling can cause a CreateContainerError that can prevent you from running bulk jobs. This error is also known to happen when the Red Hat OpenShift Container Platform is upgraded from 4.12 to 4.14 release. To disable SELinux relabeling, see the steps specific for Data Privacy service in the Disabling SELinux relabeling topic.

Details for masked columns display incorrectly

Applies to: 5.2.0
Fixed in: 5.2.1

In the asset preview page, which is known to happen for virtualized join views and watsonx.data connected data, the value for the Masked columns displays an incorrect count. In addition, the masked indicator icon image is either missing or incorrectly displayed from the header of columns with masked data. For details, see the Known issues and limitations for IBM Knowledge Catalog.

Copy related records across tables masking flow type might get stuck at Performing relationship analysis step

Applies to: 5.2.0
Fixed in: 5.2.1

When you're selecting the driver table in Copy related records across tables masking flow type, the page might get stuck at Performing relationship analysis step showing 0% completion owing to the backend connection issue.

Workaround:

1. Scale down and scale up the replicas for dp-subset microservice by using the following commands:

   oc project ${PROJECT_CPD_INST_OPERANDS}
   dp_subset_replicas=oc get deployment/dp-subset -o jsonpath='{.spec.replicas}'
   oc scale deployment/dp-subset --replicas=0
   oc scale deployment/dp-subset --replicas=$dp_subset_replicas
   

2. Confirm that the dp-subset pod is up and running after 3-4 minutes:

   oc get pod -n ${PROJECT_CPD_INST_OPERANDS} | grep subset
   

   and showing the following example output:

   dp-subset-7fbcc9dd4b-rp8zl 3/3 Running 0 4m
   

Data Privacy (Masking flow) is not FIPS-compliant

Applies to: 5.2.0 and later
Fixed in: 5.2.2

Data Privacy (Masking flow) is not Federal Information Processing Standards (FIPS) compliant.

Cannot truncate CHAR data type for Teradata in masking flow

Applies to: 5.2.0

Masking flow job fail as CHAR data type cannot truncate for Teradata relational database. For example, the masking flow job fails with the following error message:

Character data truncation detected for the field <COLUMN_NAME>. The length of the value is 44 and the length of the target column is 26. (error code: INVALID_DATA)

The masked data output contains extra space padding even after the library truncates the data.