Discover the new features, logic changes, bug fixes, breaking changes, and known issues and limitations in Piiano Vault release v1.0.0.
New features​
- New
pvault-migrate
standalone utility enabling the upgrading and downgrading of a Vault installation in a manual and controlled manner. The utility is provided within the Vault container images. - Tokens count is now reported to the log along with other metrics.
- Improved logs during Vault initialization.
- Improvements to JavaScript support for custom transformations:
- Transformations now can receive additional parameters using the
X-Trans-Param
header. - Transformations now receive richer context that includes
User
,Role
,Reason
,Collection
(name),Props
,Prop
,TxParam
, andExtra
. - Transformations can now specify
required_props
to define additional properties that should be available for the transformation. - Transformations can now use HTTP
post
andget
methods with explicitly allowlisted domains specified for each.
- Transformations now can receive additional parameters using the
- The CLI
selftest bench
command supports new flags:--skip-recreate
flag to skip collection recreation.--rate
flag to define a target RPS.- Added missing
vault_id
,customer_identifer
,customer_env
, andregistration_uid
properties to the logs.
Bug fixes​
- Fixed CLI help that was showing some incorrect global flags.
- Cache generation number now stays the same during restart if there wasn't a configuration change.
- Fixed object operations with input consisting of a list of IDs to return details in the same order as the IDs that were passed.
- Fixed CLI list transformation command printing incorrect output.
- Replaced the 500 status code returned when adding a non-nullable property to a non-empty collection with a 400 (bad request) status code.
- Replaced the 500 status code returned for concurrency conflicts with a 409 (conflict) status code.
Breaking changes​
- Encrypted properties now use Authenticated Encryption with Associated Data (AEAD) to help improve data integrity.
- The Rotate tokens operation now returns a map to eliminate dependency on response order.
- Production mode now requires that a cloud-backed KMS is set.
Known issues and limitations​
- Queries can only match untransformed properties (properties with transformations, such as
ssn.mask
, are not matched). - No pagination is provided for the tokens API operations Search tokens and Detokenize tokens.
- No pagination is provided for the List objects API operation when retrieving a list of objects using IDs. When requesting a number of objects which exceeds the page size, an error is returned.