Release Notes
v1.4.0 | Oct 30, 2025
New features
-
Integration with H2OGPTe. Users can now upload files (collections) to a particular H2OGPTe instance via the Drive app UI.
NoteTo enable file upload from Drive to H2OGPTe, set the following secret value to
true.[[Env]]
Name = "ACTIVATE_DRIVE_TO_H2OGPTE"
Secret = "h2oai-drive-secrets"
SecretKey = "activate-drive-to-h2ogpte"This secret is a mandatory secret and should exist in the environment before deploying the Drive Wave app.
Bug fixes
- Fixed an issue where the Export to DAI button disappeared from the action list after some time.
Backend changes
- Optimized internal processes to deliver better performance.
v1.3.1 - v1.3.4 | Sep 19, 2025
New features
- Added key-pair authentication for the Snowflake connector.
Bug fixes
- Fixed a timeout error with the Drive python client using a static token provider.
- Fixed an issue with the progress bar when sending a dataset to Driverless AI.
- Fixed a syntax warning in the Azure URL configuration pattern.
- Fixed an error occuring during dataset import using a connector.
- Fixed an error occuring when there is connector-related secret in secure store without a secret version.
Backend changes
- Added support for H2O Driverless AI v2.2.2, v2.2.3, and v2.2.4.
- Simplified implementation by adopting the Wave SDK's native functionality.
- Ensured DAI instance configuration adheres to DAI profile constraints in the configurable side panel.
v1.3.0 | Aug 01, 2025
New features
-
Integration with H2O Driverless AI Engines. Users can now configure the DAI Engine version and values such as
CPU Units,GPU Units,Memory,Storage, etc., from the Drive app UI. -
Users can now download files directly from the Drive app UI.
NoteNote that this feature is disabled by default. To enable it, set the following secret value to
true.[[Env]]
Name = "ACTIVATE_FILE_DOWNLOAD"
Secret = "h2oai-drive-secrets"
SecretKey = "activate-file_download"This secret is optional and only needs to be set in environments where the feature is enabled.
Bug fixes
- Fixed an issue with a error popping up when selecting a folder on the Drive UI.
Backend changes
- Updated H2O Wave to v1.7.2.
- Upgraded H2O Drive Python client from v2.0.1 to v4.1.0.
- Added support for H2O Driverless AI v2.2.0.
v1.2.1 - v1.2.5 | Jun 12, 2025
Bug fixes
- Fixed an issue when launching a new Driverless AI Engine with a selected dataset.
- Fixed migration job error that occurs when a secret has an empty subject or a differently formatted value.
Security enhancements
- Updated H2O-3 Python client to v3.46.0.7 to align with the latest security standards.
Backend changes
- Updated H2O Wave to v1.6.4
- Added support for uploading datasets to H2O Driverless AI instances with engine versions up to v2.1.0.
v1.2.0 | Mar 20, 2025
- Added H2O Secure Store for storing connector credentials for better credential isolation and security. Individual secrets are created per connector, per user. This replaces the AWS Secrets Manager mechanism of credential storage where all connector credentials were stored in a single secret.
- Secrets can now be easily viewed and edited using the UI.
- Implemented Discovery Service for automatic endpoint detection
- Improved management of demo datasets by updating the demo dataset handling to use local files instead of remote downloads.
- Extended file support by adding support for uploading DocX, RTF, and .mojo file formats through the local file connector.
- Added
AWS Session Tokenfield to the AWS S3 credentials configuration in the AWS S3 connector.
v1.1.2 | Sep 28, 2024
- Updated the default H2O Driverless AI (DAI) version specified in
app.toml.
v1.1.1 | Sep 26, 2024
- Reverted the configuration from using the latest available H2O Driverless AI (DAI) engine version to a hardcoded version to ensure compatibility between the DAI engine and H2O Driverless AI Python client.
v1.1.0 | Sep 25, 2024
- Updated the default H2O Driverless AI (DAI) version to the latest available for new engine creation.
- Resolved an issue with AWS S3 stream reading to enable error-free file downloads.
v1.0.0 | July 24, 2024
- Updated the app to Python v3.12 and fully containerized the environment.
- Fixed inconsistencies between
actand the actual environment configuration, including template generation forapp.toml. - Resolved an issue with bundle pushing to H2O Managed Cloud, which was causing deployment failures.
- Fixed Docker build image issues.
v0.8.5 | Jun 25, 2024
- Enhancements to the deployment pipeline
- Fixed an issue with exporting datasets to H2O Driverless AI.
- Fixed an issue with presigned URLs expiring too soon due to short session lengths.
info
When generating presigned URLs, set H2O Drive client's
session_ttl_secondsaccording to the expected URL TTL.
v0.8.2 - v0.8.4 | Feb & March 2024
- Improved debug logging, handling when no databases are found, and role assignments for the Snowflake connector
- Fixed the runtime version in the app.toml file
v0.8.1 | Dec 21, 2023
- Fixed a bug with the application crashing when no databases are found.
- Added debug logging when connecting to Snowflake.
0.8.0 | Nov 02, 2023
- Added Terraform script to create RDS instances (Postegres, MySql, Oracle)
- Fixed a bug with the side panel timing out
- Added testrail ID markers to test cases so the generated junit xml report can be exported to TestRail
- Added configurable role support to the Snowflake connector
0.7.4 | Apr 19, 2023
Added
- Integrated H2O Driverless AI with H2O Drive
0.7.3 | Apr 19, 2023
Changed
- Updated to Python 3.10
- Reduced volume size of H2O Drive instances to 250GB
v0.7.2 | Feb 25, 2023
Added
- Users can now upload
*.ipynbJupyter Notebook files using the local file upload connector.
Fixed
- Convert http requests to https when using the Amazon S3 connector
- Fixed a UI issue on the 'Get Link' dialog box
- Fixed an issue where double-clicking the 'Create Directory' button caused unexpected behaviour
v0.7.0 & v0.7.1 | Dec 15, 2022
Added
- Users can now upload PDFs from their local machine onto H2O Drive
Fixed
- Removed 'Test Connection' option for all connectors except MSSQL
- Removed unneccessary spaces that occur before/after URLs when configuring connectors
- Improved the connector titles that appear on the side panel for Oracle, Amazon S3, and Amazon Redshift
v0.6.6 | Oct 29, 2022
Fixed
- H2O Drive now prompts the user to complete upload after the progress bar on the UI reaches completion when uploading large files from a local machine
- Fixed an issue with displaying column names properly when importing data into H2O Drive using MSSQL, Oracle, PostgreSQL, and Amazon Redshift connectors
v0.6.5 | Oct 26, 2022
Changed
- Stripped whitespace from inputs
- Removed GO in the SQL query for MSSQL
- Secrets are no longer logged
- Added DB connection verification
- Added URL encoding for password strings
v0.6.4 | Oct 19, 2022
Added
- Support custom Port for MSSQL Server connector
Fixed
- Renamed functions and buttons for Amazon Redshift connector
- Fixed an issue with the 'editing credential profile' flow
v0.6.3 | Aug 19, 2022
Fixed
- Enabled Amazon S3 URLs to include the region name within the AWS S3 Link
v0.6.2 & v0.6.1 | Jul 20, 2022
Changed
- Removed the long "Saving to Drive" step at the end of the import for Amazon S3, Azure Blob Store, Google Cloud Storage, Google Big Query and HTTP Download connectors
- The H2O Drive Endpoint URL can now be copied from the Info dialog box.
- When H2O Drive is launched, it now follows the same light/dark mode as the user's OS setting
- Renamed the light theme to "Light" from "Ember"
Fixed
- Fixed an issue with the H2O logo not loading in Safari
- Fixed the notification that appears when Import Manager is closed right when a download completes
v0.6.0 | Jul 8, 2022
Added
- The following new connectors are availabe in this release:
- Amazon Redshift
- Amazon S3
- Google BigQuery
- PostgreSQL
- Snowflake
- Azure Blob Storage
- Google Cloud Storage
- Via HTTP Link
- Upload Local File
- The HTTP Download import connector can be used to import data using pre-signed URLs
- The Connector Config option can be used to save, manage, and delete credential profiles for different connectors
- The Create Directory option can be used to create directories and organize data imported
- Information about H2O Drive and a link to documentation
- Dark/light theme toggle
- The Import Demo Datasets option allows you to import sample datasets to H2O Drive which you can use to import data to AI Engines such as H2O-3 or Driverless AI and try out different use cases
- A back button was added to navigate back to the home page from the Import and Connector Config wizards
- Adding new credentials for each connector. Credential profiles can be saved and maintained within H2O Drive
- Display progress bars during long running tasks
- Pre-signed URLs can be copied using the copy-to-clipboard button
Changed
- Select Multiple button was changed to Select Mode and Browse Mode.
- New workflow for importing data
- Download button was removed
- Upload button was changed to Upload Local File under the Import dropdown list
Fixed
- Better error handling during import fails
- User input validation
- Fixed an issue with app crashing when a button is double-clicked
- Improved performance
- Switching between Select Mode and Browse Mode is now faster
- Submit and view feedback for this page
- Send feedback about H2O Drive to cloud-feedback@h2o.ai