Bug release (3.03.09):
Fixed Oracle objects generation in transparency layer with special characters identifier
Fixed potential archival errors in tables with multi-attribute PK
Fixed “info -test” command where purge errors where masked showing a successful result
Fixed a setup error when selecting only agent and CLI modules, generated an error by trying to set credentials when it was not expected.
Bug release (3.03.08):
Fixed compatibility issues with Oracle Autonomous DB
Fixed error that overwriting custom configured port in setup during upgrade.
“dbcloudbin info -test” command generating a successful check during some uncommon test conditions with error
Bug release (3.03.07):
Fixed private synonyms creation in transparency layer.
Bug release (3.03.06):
Fixed clean command in SQL Server environments.
Fixed behavior in managed table updates. Archived blobs are no longer re-hydrated in the database
Fixed object-store CLI command when settings generates an error. Now it is treated as a warning, since in some circumstances, it may be a legitimate error.
Fixed error message in object-store CLI command when configuring an invalid Azure connection string.
Bug release (3.03.05):
Fixed potential DL in restore -purge in scenarios of critical DB crash.
Installer with “Add schema package” selected was incorrectly trying to set root credentials.
Read stats not setting the table in the event metadata and not converting size units in info field
Multi blob attribute tables do not work transactionally in a full row archive allowing partial archives.
Error in archive/restore operations were not registered as events in event manager.
Wrong error message when Database service field is blank in installer
Included jq package in Marketplace virtual appliance to make it compatible with massive migration scripts.
Marketplace virtual appliance created with full permissions in /var/log directory.
Bug release (3.03.03):
- Fixed Linux default install dir selection.
- Fixed configuration cache refresh in multi-node upgrade scenarios.
- Fixed error where CLI custom configuration file is ignored.
- Fixed marketplace PAYG license validation.
Bug release (3.03.02):
- Fixed some stability issues with J11.
- Fixed setup workflow in marketplace virtual appliance.
- Improved performance in archive/restore/purge jobs due to optimal lock algorithm and backend parallelization.
- New “-workers” option to provide a custom number of parallelization threads for each job.
- Audit jobs to allow validate (and potentially fix) differences between the catalog of links in DB and the content in object-store.
- Supported cloud based DB deployments: Oracle RDS in AWS, Oracle Autonomous DB, Azure SQL Managed Instance.
- Improved health statistics and dashboard content, adding the originating host in every event.
- Check DBCLOUDBIN defined db user password meets database pwd criteria on install.
- Added files for custom settings configuration for agent or CLI (avoiding overwriting in case of upgrade).
- Added HCP S3 based object store driver for Hitachi HCP, setting as deprecated the previous REST based protocol.
- Upgrade mode will request only DBCLOUDBIN credentials in most scenarios.
- Improved security with the option of not storing password for a session in CLI. In this case pwd will be requested in every operation and hidden.
- A session with the name of the managed Schema/DB is automatically generated in setup when adapting an application database.
- Support of Java 11 (11.0.9 or later).
- Name of generated transparency layer schema/DB can be customized using a property in setup execution.
- Generate and feed an ‘activity’ trail to customer area in website
- Provide windows AD integrated authentication for SQLServer
- Add replicated object store functionality
- Allow easy object store configuration in CLI config
- Add in info -test to check the install with a ‘probe’ archive
- Improved security with hash based object validation on read
- Automatically add a ‘root’ session credentials option on install
- Use by default the S3 option for using bucket as address in S3 compatible object stores.
- In layered mode the to-be-purged address was incorrectly recorded upon a DB row update.
- Fixed error stream in CLI commands
- Corrected minor Web issues.
Full transparency mode (layered) implemented. It provides the ability to create a new ‘layer’ DB/schema that contains DBcloudbin objects and links to the archived BLOB’s. This provides full transparency at data model layer not only application access layer, since the application data model is maintained intact.
Introduced “refresh” command in CLI in order to directly refresh the DBcloudbin created data model upon changes in the customer application data model.
Ability to modify agent address in setup configuration wizard.
Improved load-balancing configuration with root webservice readiness probe.
Improved security on encryption services.
Ability to request password by terminal in set-credentials command in order to improve privacy.
Main functionality added:
- Ability to provide full Oracle connection string in setup, in order to provide custom parameters as load balancing DB scenarios
- Improved quota management mechanism.
- Generalized marketplace support to GCP and Azure marketplaces.
- Hitachi HCP plug-in improvements with simple user and password authentication (automatically transforming to accessKey and secretKey)
- Hide secret keys provided for custom object store configuration at setup UI.
Main functionality added:
- Added main functionality for AWS Marketplace support.
- Formalized log activity entries.
- Added metadata to config parameters, in order to improve security (hidden, read-only, privileged update,…).
- Corrected client-side quota management.
- Fixed support of local GCP object store in setup tool
- Correct initialization of config settings supporting online settings update
Main functionality added:
- Support for DBcloudbin upgrade (both database model and SW only scenarios).
- Info command that provides information on versions, running agents and managed table names.
- Archive uninstall files in DB config in order to make servers truly stateless allowing rip&replace scenarios.
Main features included:
- Support SQL server databases.
- Full rewrite of archive, restore, clean and purge commands to support correct table locking and several commands running in parallel.
- Description of profile and connection used in each CLI execution.
- Added consistent behaviour in restore. Now it register objects in purge queue by default. Added -purge option to automatically purge the objects as in previous release.
- Consistent implementation of archive when the previous version of a BLOB was already archived. The previous version is registered in purge queue to avoid being lost.
- Added option newer table and older in purge command in order to fine-grained flexible filtering.
- Problems with purge command. Driver generating an error when purging many rows.
- Filtering up tables with no PK in table selection screen in DBcloudbin installer. Avoiding an error in the setup execution.