Month in Review: March 2023
In March, we not only focused on the UI but also made significant changes to the backend, added a few changes for developers, and released a new version of the Buffer API.
Minor UI Updates
- We tuned the code editor a bit, so it highlights matching brackets and automatically closes brackets and double quotes.
- Every page that contains a search bar at the top includes information about the number of entities on a specific page (e.g., the number of configurations, buckets, or tables).
- When searching through options in the select component, the search query is highlighted in the search result.
- Searching in Storage has been revamped: Search panels on bucket and table details are now directly above the table of searched content, and the results are also shown in this table (previously, the results were on a separate page).
- When you try to delete a configuration, we check all flows to see if they use the configuration. If so, you see a warning and additional information.
- Editing of configuration descriptions has been redesigned.
- Configuration options for the Hive data source component (extractor) have been extended, and you should be able to configure everything via the UI.
- It is now possible to set the "Write always" flag via the UI. This is useful if you have a table that you are updating during the execution of a job and you want to output that table even if the job fails.
Other Important Updates
Storage Backend Registration
The SuperAdmin role is no longer needed to register the backend to our Storage API. This will remove a limitation where only our support was able to register the backend for you. However, to make this non-destructive, our support is still needed to assign the backend to a specific project.
Buffer API Beta
A new version of the Buffer API has been released, which has fixes for several bugs we found during beta testing.
Read-Only User Role Available on AWS EU and AWS US Stacks
Automatic Dropping of Badly-Structured Tables
When a job creates a new table and the load to that table immediately fails, the table is dropped and a warning is logged in the job. Thus, "broken" tables won't be left behind anymore, and, you won't need to go to Storage to drop a table manually when its structure is wrong.
Component developers will appreciate the new environment variable
KBC_PROJECT_FEATURE_GATES, which contains a list of features assigned to a specific project. Read more in our documentation.