Skip to main content

February 2023

2.8.0.* (February 28, 2023)​

  • Prophecy Python libs version: 1.4.2
  • Prophecy Scala libs version: 7.0.7

Features​

A team is not required during sign-up​

Users were required to enter a team name when signing up for Prophecy. Prophecy removed the team requirement. A user can sign up by providing their Name, Email and Company name. They can be added to a team by an administrator after signing up.

New_Sign_up

Added support for Identity provider initiated SSO​

Users would now be able to add Prophecy as an app in Okta, and directly get signed in when App is opened from Okta.

Sorting columns on Metadata tables​

Metadata tables provide various options to sort on its columns. By default, the columns will be sorted on Modified At or Created At time in decreasing order.

Sorting

Cluster consistency​

If a user is editing a Pipeline in a tab where the same Pipeline is already opened in other tabs, Prophecy will ask to detach from the Pipeline in the other tabs. It will allow users to attach the cluster to the Pipeline in the active tab.

A user will see a warning before the cluster is detached from the previous tabs. Attach-cluster

If running Jobs already exist in a tab, the user will be prompted to cancel the running Jobs before attaching the cluster to the current tab. Attach-cluster-warning

In Job editor view we've added the link to the corresponding Databricks Workflow.

Please note: this will only appear if the Job is enabled and released.

Databricks-link

Create Pipelines and Jobs directly from Project Browser​

Previously, a user had to navigate out of Pipeline editor to add a new Job or Pipeline. Now they can do it via the same screen and shown below. Please note, the Pipelines and Jobs can only be created in same Project and Git branch using this option.

Create-pipeline

Alert user on Databricks Token Expiry​

If their Databricks token is expired or is invalid, the user will now see a prompt as below to update their Databricks token. They will be prompted on the Jobs editor or Pipeline editor when attaching to the Fabric.

token-expiry

Improvements​

Allow Run on main branch​

Editing your Pipeline on Main branch is not allowed. Interactive run and Partial run on Gems is allowed when viewing a Pipeline on Main branch. Interims will also work as expected.

Run_main_branch

Cluster restart after setting dependencies​

Improved the cluster restart behaviour when adding a new Pipeline dependency. When a new Pipeline dependency is added, cluster will be detached and prompt for a restart as shown below. A small subset of users were experiencing cluster restart loops when using certain dependencies and this has been addressed

cluster_restart

2.7.0.* (February 16, 2023)​

  • Prophecy Python libs version: 1.4.0
  • Prophecy Scala libs version: 7.0.6

Behaviour Changes​

Spark and Scala Versions are now Required in Livy Fabrics​

For creating a Livy Fabric, previously the Spark and Scala versions were optional and had default values 2.3.0 and 2.11. With this release these fields are made mandatory, and users will need to update their Fabric configurations with respective Spark and Scala version values. Please refer here for more details.

New Features​

Kafka and Watermarking support (Beta)​

With this release, we have added support for Kafka as a streaming Source and Target. Also support for Watermarking is added in all streaming Gems.

Data Fabric ACL​

Support for Data Fabric authorization is added with this release. This feature allows administrators within a team to assign proper access to its team members. For example, only a team’s administrator can perform tasks such as creation and deletion of Data Fabrics.

Ticketing Support​

Users can raise Support tickets, Feature requests, and bugs using the following two options:

  • Within the Prophecy UI by clicking on the Support icon in the Bottom-left corner. freshdesk

  • Submit new tickets and view status of existing tickets from the Customer portal. Please contact us to enable this in your account.

UX improvements​

Auto Connect Gems and Auto suggest on Drag and drop​

While adding new Gems to a Pipeline, the new Gem will be Auto-Connected to a previous Gem when dragged nearby.

Also, when dragging a Gem onto the canvas, user will be auto-suggested a Gem position. Please see the below examples. User can hold Shift key to disable this.

auto-connect

auto-suggest

Zoom-in And Zoom-out​

Users can now use Two-fingers-Pinch to zoom-in or zoom-out on the Pipeline canvas.

Gem Sorting​

Gems are now sorted Alphabetically for easy lookup.

Gem-Sort

Dangling interims​

While editing/creating Pipeline, preview data "interims" now available after the tail node (last Gem on the graph).

Dangling-interim

Improvements​

Project creation on protected main branches​

Now users will be able to create projects when using Repositories with protected main Branches. This is applicable for both Single-fork or Multi-fork Repositories.

Disconnection Error handling​

When a user has an open Pipeline, the Prophecy connection can be lost due to multiple reasons. We have implemented auto-reconnect for a smoother experience. Now the user is given an actionable message in the UI only when needed:

  • If connection is lost due to inactivity (Probably user switched to another tab), then they will see a reconnecting toast as shown below: IDE_disconnect1
  • When your device network is switched IDE_disconnect2
  • When some other unknown disconnection happens (For example: the network provider switched at router level) IDE_disconnect2

Copy pasting Gem in Pipeline editor​

When a user is pasting Gems/group of Gems in Pipeline editor, the Gems will now be pasted at the cursor location. They will also remain selected after paste so that user can easily move them around on canvas as needed.

2.6.0.* (February 3 2023)​

  • Prophecy Python libs version: 6.3.20
  • Prophecy Scala libs version: 1.3.22

New Features​

GCP Support​

Prophecy now has support for integrating through Databricks Partner Connect on GCP. For example, when Databricks is deployed and running on GCP and integration between Prophecy and Databricks is needed, the user can integrate with Prophecy through Databricks Partner Connect.

Interims for Unity Catalog Workspaces​

Pipelines running on Unity Catalog cluster now have the ability to show Vanilla interims on all Actions/Tail nodes of the Pipeline.

Team-level execution metrics​

An administrator will have the option to create the following tables at the time of team creation which are used for storing Execution Metrics. The administrator will grant access to the tables accordingly to their team members.

  • Pipeline Metrics Table
  • Component (Dataset) Metrics Table
  • Interim Table

Seamless Git Integration with stored credentials​

Added the ability to store Git credentials for a user across projects. Now the user can re-use Git credentials without re-authenticating.

Streaming Pipeline Support (Beta)​

With this release, Prophecy now supports Streaming Pipelines as a beta feature. Users can now read data from streaming sources, apply transformations and save data to streaming Targets.

Updates​

Union By Name​

Added Union by name in Set Operations Gem. This can be used now to get union by column names when positions are different in two datasets.

Improvements​

Python code generation taking long time​

For larger schemas, Prophecy's Python optimiser was taking significant time. As a result, compilation was delayed. The new optimization speeds up wide table processing on the UI layer by adding pre-optimization phases. This will try to eagerly evaluate and substitute for preconfigured syntax. The compilation time has been reduced orders of magnitude from seconds to milliseconds.

Blocked log4j dependency​

The Scala maven plugin has a log4j dependency that is inaccessible for some customers. Updated the log4j dependency.