May 2026
The Continuous Compliance 2026.3.0.0 release adds support for Brazil CPF and CNPJ formats in the Checkdigit framework and enhances file format management for Mainframe Copybook formats. You can now edit, duplicate, and export Mainframe Copybook formats on source engines, aligning copybook management more closely with other supported file types.
This release also includes performance and stability fixes, particularly for bulk file update operations and mainframe dataset handling during upgrades.
New features
Read‑only root filesystem support for Containerized Masking
This release enforces a read‑only root filesystem for all Containerized Masking Engine containers (app, database, and proxy), improving security and aligning with Kubernetes best practices. Required writable paths use automatically provisioned ephemeral volumes, with no changes needed to existing persistent volumes or masking jobs. For Helm deployments, this setting is enabled by default; to retain the previous behaviour, set containerSecurityContext.readOnlyRootFilesystem: false before upgrading.
Brazil CPF and CNPJ support added to the Checkdigit framework
The Checkdigit framework now supports Brazil’s new alphanumeric CNPJ format, in addition to the existing numeric CNPJ and CPF formats. This support is provided through new default instances for Brazil CPF and Brazil CNPJ. For details, see the Checkdigit documentation.
Enhanced file format management for Mainframe Copybook formats
This release extends file format management to Mainframe Copybook formats, bringing them in line with other supported file types. You can now edit existing copybook formats, duplicate them under a new name (with the option to copy or exclude masking assignments), and export them for reuse or backup. This reduces the effort required to maintain copybook definitions across environments and simplifies promoting formats between engines.
Fixed issues
| Bug number | Description |
|---|---|
| DLPX-96288 | Fixed performance issues impacting bulk file update operations. |
| DLPX-96763 | Fixed an issue where Continuous Compliance Usage Reporting displayed duplicated values for Oracle RAC data sources. |
| DLPX-96819 | Fixed an issue where missing mainframe dataset format files could cause the masking engine to become inaccessible after upgrade |
| DLPX-96888 | Fixed an issue where renaming a plugin via the PUT API or JAR manifest caused Delphix Sync to import it as a new plugin, breaking algorithm associations. |
Known issues
|
Key |
Summary |
Workaround |
|---|---|---|
| DLPX-80640 | Masking drop Indexes task does not find and remove indexes on SQL Server partitioned tables in both on-the-fly and in-place masking jobs. | Manually script the drop and recreate indexes. |
| DLPX-93701 | On the fly masking job fails when using different source and target connector types (mixed Cloud and On-premise connectors) for JSON, XML, and Parquet files. | None |
| DLPX-95510 | Parquet masking jobs fail in GCP interoperability mode when the GCP bucket name contains underscores. This occurs because the Hadoop S3A client enforces AWS bucket-naming rules before the request reaches GCS. | None |