Unifi Data Platform Version 2.9 is now generally available, and we have expanded our data catalog offering significantly. In the previous release, we provided support for cataloging metadata from multiple Tableau servers centrally in the Unifi Data Catalog. Now, users can Search this data.
Similarly, while we supported crawling tables from a database and creating a dataset for every table, functionality didn’t exist for crawling file directories. Starting from this release we now support crawling directories recursively and creating a new dataset for every distinct file pattern. In this release we enhanced governance functionalities provided by catalog by allowing Data Stewards to endorse datasets and provide information about trust level. Details for these features are below:
Metadata Search across Tableau Servers
For the vast number of organizations that have multi-server Tableau installations, a global view of Tableau data may be elusive. Users connected to different servers are quite likely recreating Views that have been created previously. It can be a tremendous challenge to get a true, comprehensive view of Tableau Projects, Workbooks and Views created across all Tableau sites and servers—until now with Unifi. In our latest release, we extended Unifi to support crawling metadata from a Tableau server into the Unifi Data Catalog. This greatly enhanced the ability for Tableau administrators and data stewards to get an aggregated view of all Tableau Projects, Workbooks and Views across multiple Tableau sites and servers. Using the capabilities for searching metadata cataloged in Tableau through Unifi, a data analyst can easily see if something was created previously and where it exists for potential reuse.
When creating a file system data source users can automatically create new datasets by traversing the files of a directory and/or a subdirectory using new “Crawl and Import” functionality. For crawling, users can choose a base directory to crawl from for supported file sources including S3, Azure ADLS, Azure Blob, Google cloud storage and HDFS. Recursive crawl enables crawling files in nested directories. Users can include or exclude specific file types using REGEX pattern. The product allows generating prefixes and tags when a dataset is created. Once the crawl has completed, users will see the newly created datasets in the Unifi Dataset Explorer.
Data Analysts want to ensure that Datasets they are using are of good quality. We now support the ability for administrators to endorse a dataset as “trusted” or mark it as “deprecated.” Hovering over the trusted icon displays any comments entered when the file was marked as trusted and denotes the user marking the file. A clear distinction is made on a dataset that should not be used.
Stay tuned for more exciting new features we will be launching soon.