Cells v2.1 is an important milestone with a lot of attention put into performances, an Authentication module rewritten to get rid of networking configuration issues, and a revamped Cells Console.
Highlights
Main User Interface
- Rewrite of the Uploader UX - Fix inconsistencies in the start/pause/resume interactions
- New activity on metadata updated and on new comments
- Rework Bookmark and Activity panels
- New action button on Workspaces and Cells to directly bookmark or watch a whole workspace.
- Show exact Byte size when hovering human readable size in file info
Admin Dashboard
- Brand new look-n-feel, better menu organization
- Unified pagination, sorting and actions on tables, applied to all dashboards, added quick search on most dashboards
- Better log display, rework search engine to add various criteria, focus on a result and show +/-5 minutes logs
- Indexed logs including nano time for better sorting
- ListWorkspaces API optimization (bulk load graph dependencies)
- [Enterprise] New audit dashboard for exploring storage usage
- [Enterprise] Tech-preview for Admin pages delegation (contact us)
- [Enterprise] Tech-preview for scheduler jobs editor
Storage & Performances
- Huge speed up of move/rename folders
- New advanced options for managing ETags in datasources (use native ETags, compute ETags but store them on Cells side).
- S3-compatible storages : send stats in concurrency to speed up indexation, ability to use an S3-only setup at install time, rework S3 gateway to avoid creating .pydio and use amazon-like empty objects instead for folders.
- Ability to create a folder on filesystem directly within the datasource editor
Authentication
- Massive rewrite of the authentication module that greatly simplifies network flows (no more calls to external URL)
- Fixed case-sensitivity limitation on user/group search
Services startup, APIs and other
- Better error statuses on Rest API (detect lower-level code, special case for 503)
- Special value "unique" for --exclude and --tags parameters to exclude or start services that are declared as unique
- Fixed webDAV double auth request
- Automatically restart forks if necessary
- New optional flag to access nats statistics on a dedicated HTTP endpoint
Upgrade - Readme
Upgrading is done by replacing the binary with the latest version (via the in-app tool or manually). Some important notes:
- Changes are applied to both the data model and the configurations, thus reverting to an earlier version is not an easy task.
- During the first restart, database schema changes will apply and may take some time depending on index size. A second restart may be required If a datasource seems stuck or unavailable.
- AWS S3 datasources have been improved by getting rid of the ".pydio" hidden files and use AWS 0-bytes keys instead. This format change is applied automatically during the first indexation, and this can take some time as well.
We strongly encourage you to run the upgrade on a staging environment first, or at least to warn users that there will be a short downtime (if you need to restart the service a couple of times).
Change log
You can find a summary of the change log here.