Snowflake Openflow version history¶

Restores support for private Openflow BYOC Deployments by removing all dependencies on URLs outside of Snowflake and AWS, addressing an issue introduced in 1.6.0

Fixed a rare issue that prevented a Runtime from being activated after it had been suspended

Fixed an issue causing connector installations to fail on new Runtimes when bulletins are present

Improved visibility of Runtime operations with new metrics for Connectors

New components to interact with SAP Business Data Cloud and mapping of CSNs into Snowflake Semantic Views

CDC MySQL - Improved reliability by clearing out the table map prior to CDC reconnects

CDC Databases - Fixed a potential deadlock issue with MergeSnowflakeJournalTable when “poll query result” is cleared during operation

CDC MariaDB - Added support for MariaDB in the MySQL components

CDC PostgreSQL - Added support for primary keys of type numeric

CDC Databases - Adjusted backpressure thresholds on some connections when processing a lot of data

Salesforce - Gracefully handle scenarios where we insert duplicate rows in the staging table

CDC Databases - Exposed parameter to enable private connectivity in the PutSnowpipeStreaming processor for data ingest

CDC PostgreSQL - Adjusted yield duration on CaptureChangePostgreSQL to not overuse replication connections

SQL Server - Performance improvement in Snapshot query

Dataverse - Schemas are no longer filtered when no column filtering value is provided

Telemetry - “Bytes Received” is now available for many Snowflake processors after fixing the file size for provenance events

Azure components - Fix ConsumeAzureEventHub by excluding netty-codec-http3 dependency

Google Cloud - Added support for Workload Identity Federation

Azure Blob Storage - Added support for uploading file larger than 200GB

Google Ads - Set the new Authentication Strategy property of the GCP Credentials Controller Service

Multi Database SQL Server - Fix source.table.fqn value handling

Salesforce - Add logging for successful sync operations to ease monitoring via the events table

Improves security by removing unused inbound ports on Load Balancers configured for “Custom Ingress.” You can further limit access with your own Security Group for these Openflow Deployments.

Improves security and eased configuration of Runtimes with an optional Deployment-level IAM Role to securely access AWS resources like RDS, MSK, Kinesis, S3, etc. You can now attach IAM Policies to Openflow’s “NodeInstanceRole” that are granted to all Runtimes in that Openflow Deployment.

Upgrades EKS Cluster from 1.32 to 1.34 for long-term maintenance and security patching

Resolves an issue where restarting some EC2 nodes frequently caused the Openflow Deployment to freeze

Security patches and upgrades to third party libraries

A missing event table will no longer cause failure when creating an Openflow Snowflake Deployment.

Fixed certificate based issues accessing Runtimes and deploying Connectors into Deployments older than 60 days

Security patches and upgrades to third party libraries

Adding support for the new Cleaning Up Runtime state

Adding support for terms accepted trial not started or active trial

If a Runtime fails to create, it will automatically generate a diagnostics bundle and clean up any partially created resources in the cluster.

If a Runtime fails to create, it will automatically generate a diagnostics bundle and clean up any partially created resources in the cluster.

Security patches and upgrades to third party libraries

Improved reliability for high volume deployments by relocating state tracking for replication position and journal versioning in CDC Connectors

Easier debugging with an updated Bulletin Board that can expand stack traces

Fixed bug when rendering Documentation for extensions that lack tags

Viewing component state now supports showing 5,000 local entries and 5,000 cluster entries, up from 500 each

Reduced costs by removing the validation query in the Snowflake Connection Service

Google Drive and Google Sheet - Set the new Authentication Strategy property of the GCP Credentials Controller Service

Added support for Web Identity authentication to AWS MSK IAM Connection Service in Kafka components

Added support for Web Identity authentication to AWS Credentials Controller Service for all AWS components

Added Flow Registry Client support for Bit Bucket Data Center edition

Fixed Worker ID generation in ConsumeKinesis and added provenance data

Added support for nested paths in HashiCorpVaultParameterProvider

Dataverse - Added retry-after mechanism

Added Snowflake Secrets Parameter Provider

CDC Database Connectors - Improved reliability and performance on state management

JIRA - Enriched issues with email addresses

HubSpot - Fixed handling of 414 error code responses while fetching objects

Dataverse - Add Column JSON Filtering parameter

PostgreSQL - Added FIFO FlowFile prioritizer on queues in Postgres Snapshot Load

MySQL - Expose parameter for the starting position of the replication

Dataverse - Added updated_ad and deleted columns

Salesforce - Switched the CSV Reader to RFC 4180

Salesforce - Fixed configuration to capture soft deletes

Security patches and upgrades to third party libraries

Security patches and upgrades to third party libraries

Enable Openflow Oracle Connector in Snowflake Deployments

PrPr release of Horizon Oracle Metadata Connector

Security patches and upgrades to third party libraries

Security patches and upgrades to third party libraries

Improved reliability of high volume CDC Connectors

CDC Database Connectors - New components for multi-databases support are now included in the runtime image

JIRA - Added support for Forge App authentication method

New OAuth2 controller service to get Snowflake issued JWTs for Workload Identity Federation

PostgreSQL - Added First In First Out (FIFO) connection prioritizer in PostgreSQL Snapshot load

CDC Database Connectors - Disabled load balancing in the incremental flow to ensure single node processing of the data

Horizon - Made parameters reachable from the root process group of the connectors using a root parameter context

Dataverse - Added parameter for the new JSON Column Filtering property

Fixed bug in Summary table formatting Process Group task time

Kinesis - Support Output Strategy property in ConsumeKinesis processor

Kinesis - Added the new Kinesis components leveraging the latest AWS client library

SQL Server - Added support for multiple databases

MySQL - Added the possibility to specify the binlog starting position for reading the CDC stream

PostgreSQL - Added support for negative scale in numeric types

SQL Server - Improved the ordering of the ORDER BY clause

Snowpipe Streaming - Improved Input Buffer Handling in PutSnowpipeStreaming2

PostgreSQL - Improved performances in FetchTableSnapshot on large tables with composite primary key

MySQL - Fixed incorrectly replicated DATEs pre 1582-10-15 (Julian calendar)

Horizon - new connector to fetch metadata from Oracle

Oracle - support for multiple logical databases

MySQL, PostgreSQL, SQL Server - no longer writing the unused avro.schema FlowFile attribute

Jira - support for fetching Worklogs

Security patches and upgrades to third party libraries

Security patches and upgrades to third party libraries

Fixed issue where a user can’t log into Openflow if their most recently selected active role was revoked.

Disable creating a Snowflake Deployment if the user’s role is not granted CREATE COMPUTE POOL privilege.

Improved display for Upgrade Failed, Inactive, and Activate Failed states

Always show the current “Run As” role even if it’s not in the current user’s set of account roles

Fixed issue with “Run As” role validation in Create Runtime dialog

Bulletin icons now reflect the severity of the message

Parameters can now be edited by double clicking on the row in the Parameter Context

Included state of system diagnostics API call in the loading skeleton and spinner in the Cluster Listing

Improved awareness of errors through the global banner when extension types fail to load

Updated styling for unset, blank, empty styles throughout the Runtime UI

Fixed incorrect handling of Drop Table actions in UpdateSnowflakeTable processor

Oracle - Improved performance by moving metadata generation in FetchSnapshot processor

Oracle - Fixed handling of column filters DDL

Dataverse - Added optional configuration to filter columns being fetched

Cortex - Improved error message when there is an issue calling Cortex in PromptSnowflakeCortex processor

MySQL - Fixed the filtering out of the “user” table

Salesforce Data Cloud - Added support for detecting deletions of Data Shares and linked Objects in the shares

MySQL - Fixed skipping compressed transaction DDLs and DMLs spanning over the transaction

JIRA - Enrich Jira Worklogs processor

Confluence - Support for Confluence Data Center edition

Added Offset Tracking Resolution to PutSnowpipeStreaming2 processor

Sharepoint - Fixed pagination handling when listing more than 200 items

Salesforce - Added optional lookup key in UpsertSFDCObjects processor allowing user to specify a field other than ID for retrieving the record to upsert

Excel - Added missing SPCS related configuration options

HubSpot - Added support for new object types: Notes, Orders and Carts

Salesforce - Added missing configuration for authentication strategy for usage of the connector in Openflow Snowflake Deployments

PostgreSQL - Migrated the connector to standard identifiers for better management of case sensitivity on object naming

Oracle - Removed the addition of Snowflake Specific Columns to leverage FetchSnapshot processor instead and improve performances

Sharepoint - New Simple Ingest connector that does not fetch the ACLs associated to the data

Salesforce - Added support for specifying the object fields that should be included/excluded when retrieving the data

Released Control Plane Core version 0.84.1.

Fixed certificate issues that blocked access to runtimes and connector deployments in deployments older than 60 days.

Fixed input validation issues when filtering in role selection menus.

Fixed an issue where links to runtimes were shown to users without access privileges.

Fixed an issue where users with only USAGE privilege on a runtime couldn’t create connectors in that runtime.

Fixed an issue for users accessing Openflow over PrivateLink with a network policy enforcing the VPCE ID.

Added support for suspending and activating runtimes in Snowflake deployments.

Snowflake deployments now display their current version immediately after creation is initiated.

Warns users before navigating to a deployment or runtime where VPN connectivity may be required when using custom ingress.

Keeps the selection panel open for multi-select components after a selection is made.

Enforces user permissions for viewing the runtime canvas and hides links if permissions are missing.

Improves setup experience by considering total counts of runtimes and deployments, not just those in the ACTIVE state.

Makes the Snowflake role optional when creating a runtime in a BYOC deployment.

Improves text overflow handling for connector cards.

Fixed an issue that prevented access to runtimes over PrivateLink.

Fixed an issue where a new runtime couldn’t be accessed if its name matched that of a previously deleted runtime.

Added Snowflake Managed Authentication Strategy to SnowflakeConnectionService and PutSnowpipeStreaming.

Improved snapshot query performance by correcting ORDER BY column sorting.

Fixed missing Process Group identifier information in Processor and Controller Service log records.

Added support for workarounds when using self-managed certificates in AWS.

Fixed issues that caused BYOC deployment upgrades to get stuck with invalid image references and job cleanup.

Restored support for adding customer-managed IAM policies to Openflow’s IAM roles.

Updated the Dataverse connector to set empty collation for the Dataverse journal table.

Added better support for case sensitivity on Snowflake objects in MergeSnowflakeJournalTable.

Improved HubSpot pagination handling when retrieving more than 10,000 records.

Unstructured Processing - PerformSnowflakeCortexOCR now uses the AI_PARSE_DOCUMENT function instead of PARSE_DOCUMENT.

Added better support for case sensitivity on Snowflake objects in PutSnowpipeStreaming.

PostgreSQL - Fixed unsigned handling of type OIDs in the CaptureChangePostgreSQL processor.

New Runtime UI 0.53.0.

Fixed a regression that prevented tabbed dialogs from remembering the previously active tab.

Fixed balto icon regressions and selected radio button display issues.

Fixed an issue where Parameter Context update requests weren’t deleted when users canceled the request.

Fixed an issue that caused double scroll bars to appear in the asset upload dialog.

Fixed an issue where the selected asset count could get out of sync.

Improved efficiency of private IP addresses used by EKS cluster nodes, reducing the total number required for scaling out to many Runtime nodes.

Fixed issue with Runtime logs that incorrectly redacted some component IDs.

Confluence connector - Better failure handling and retries when facing API rate limits.

Support for deploying Oracle Runtime Extensions to Runtimes in BYOC Deployments for PrPr customers who have accepted the Terms of Service.

Fixed an issue where Snowflake deployment moved into an active state prematurely during an upgrade.

Fixed a rare issue where Snowflake deployment deletions could get stuck and need manual intervention.

Introduced new deployment upgrade dialog that shows the version mapping.

Support for deploying Oracle Runtime Extensions to Runtimes in BYOC Deployments for PrPr customers who have accepted the Terms of Service.

CDC database connectors: Removed Record Reader from MergeSnowflakeJournalTable processor.

All connectors log the Query ID whenever a connector executes a query in Snowflake.

PrPr release of Oracle Extension for Openflow Runtimes.

Improved the Openflow Connectors upgrade user experience.

PostgreSQL connector now includes a new parameter so you can set the replication slot name.

The PostgreSQL, MySQL, and SQL Server connectors now support column names that include special characters.

Added compression to rows added using the Insert Rows method through PutSnowpipeStreaming2.

MySQL: Added support for compressed bin log events.

Added new processors, UpdateSnowflakeSchema and UpdateSnowflakeStream, to better manage object lifecycles and support case sensitivity.

HubSpot: Added support for new “Notes,” “Orders,” and “Carts” object types.

Slack: Fixed Null Pointer Exception when trying to verify the configuration of ConsumeSlackConservations processor.

Using latest Apache NiFi 2.6.0 release.

Improved the flow upgrade user experience by improving Flow Differences Filters to handle renameProperty, removeProperty, and createControllerService.

New Runtime UI 0.52.0.

Fixed bug allowing default values for dynamic properties.

Improved the performance of the searchable select used in the Property combo editor.

Openflow now supports VPCs with DHCP Option Sets, making it easier to connect to private data sources.

You can now secure Openflow deployments with PrivateLink, while still allowing browser-based authentication to runtimes without PrivateLink.

Fixed an issue during upgrades where IAM inline policies failed by exceeding maximum character limits.

Improved error messages for Snowflake deployment failures to show the root causes.

Fixed a case where BYOC deployment ends up in Not Healthy state but can’t be deleted from Openflow Control Plane.

Removed unnecessary title on Runtime and Deployment state columns.

Improved cookie session handling to allow users to remain logged in, even when Runtime is open in an inactive browser tab.

Addition of the 2 new Oracle CDC connectors.

Confluence connector - The introduction of a new controller service to handle API rate limits will show the connector as a process group with local changes. This can be ignored and will be resolved when upgrading the connector to the next version, when available.

Introduced the UpdateSnowflakeTable processor, which is like UpdateSnowflakeDatabase, but designed for tables and improved case sensitivity.

SQL Server connector: Exposed the new SQL Server query interval property as a parameter.

The new controller service for API rate limits in the Jira connector causes the connector to appear as a process group with local changes. You can safely ignore this; it will be fixed in a future connector upgrade.

Allow users to optionally configure whether end users authenticate over PrivateLink.

The Estimated time to completion shown when creating Snowflake Deployments and Runtimes is now more accurate.

Initial release offering privilege isolation for Openflow runtime authentication and authorization to Snowflake deployments.

Added support for DATETIME columns with PutBigQuery processor.

You can now specify the HTTP protocol version in StandardWebClientServiceProvider.

Better logging and increased timeouts for FetchSharepointFile processor.

Added the option to set the replication slot name in CaptureChangePostgreSQL processor.

You can now use -infinity and +infinity with Postgres TIMESTAMPTZ values.

New controller service StandardAtlassianRequestRateManager to deal with API rate limits for the Jira connector.

Fixed exceptions thrown from ListMicrosoftDataverseTables when table schema isn’t returned by API.

Support deploying the new Openflow ingress controller for PuPr release of Snowflake deployments.

New Runtime UI 0.51.0.

You can now delete individual entries in the component state if the component allows it.

Improved tooltips for Property and Parameter values, especially when values are long or reference external resources.

Fixed an issue from AWS Data Plane Agent 0.39.0 that blocked the first install of an Openflow deployment into a new AWS region.

Fixed issue preventing runtime deletion in Snowflake deployments when a network policy is present.

Fixed an issue that prevented runtime deletion in Snowflake deployments when a network policy was present.

Fixed an issue that prevented new versions of runtime extensions from being used when runtimes were created or upgraded.

CaptureChangeSQLServer: A new setting, Table Changes Query Interval, is introduced to reduce the resource pressure on the source database. Now, the processor queries the source database every 10 seconds (10 sec) by default. To restore the original behavior, change the setting to 0 sec.

Resolved an issue where deployments were left partially upgraded after AWS Data Plane Agent 0.39.0 was used.

Unstructured connectors: Improved reporting on ChunkText failures.

Microsoft Dataverse: Fixed handling of schemas that include the Edm.Date type.

Fixed attribute prefix handling in the XML Reader.

Fixed MongoDB controller service for certain authentication methods when information is provided through the URI.

Added Azure DevOps Flow Registry Client for Git integration with Azure DevOps to version flows.

Added the ability to change the version of a ghosted component if a bundle with the same coordinates and a different version exists.

Added support for AWS Data Plane Agent deployments that have DHCP Option Sets configured on the account.

Upgraded all EKS nodes from Amazon Linux 2 to Amazon Linux 2023.

Added support for AWS accounts that require encrypted EBS volumes by default, even if an unencrypted EBS volume is requested. Customers can enable this by adding IAM Policies to the *-eks-role IAM Role that grant access to their KMS keys.

Error messages are now clearer and more informative when runtime-related failures occur.

Fixed a rare case where an older deployment version disallowed creating a runtime with the same name as a previously deleted runtime.

Deployment listing and details now include the deployment version number.

Control Plane logout page now offers a link back to Snowsight

Searchable select control (used in Create Runtime and Manage Access) now offers improved behavior when text overflows available space.

Fixed a bug that temporarily showed duplicate roles when revoking privileges through the Manage Access dialog.

Added support for AWS Data Plane Agent deployments that have DHCP Option Sets configured on the account.

Allowed customers to delete a runtime and create a new one with the same name shortly thereafter.

Confluence Connector: Refresh frequency is now set to 1 minute and is no longer exposed as a parameter.

Resolved an incompatibility between the Github Registry Client and the latest Jackson release.

Fixed attribute prefix handling in XML Reader

Added StandardProtobufReader controller service for Protobuf record processing

ListTableName won’t fail the entire FlowFile if partial input is incorrect.

Introduced Runtime UI 0.50.0

Added a new logout page that provides users options for logging back in or navigating to the Control Plane.

Enhanced the searchable select control to display options more clearly when text exceeds available space.

Fixed casing and icon issues when inputting attributes during extension verification.

Fixed header styling applied to additionalDetails markdown files.

Support for AWS Tags with dots in the Tag key.

MySQL CDC: Always create a new table (and fail if the table already exists) when replication mode is set to full.

The GitLab Flow Registry Client now supports versioning flows larger than 2MB.

Fixed issue in the MongoDB Controller Service preventing users to authenticate using X509.

Fixed irrelevant error logs about schema hash in UpdateSnowflakeDatabase processor.

Confluence: Fixed a bug that prevented users from being added to authorized users even though they had permissions to the space from the group level.

Fixed NoSuchElementException thrown in ChunkText processor and better failure handling with dedicated relationship.

HubSpot: Fixed bug preventing the List processors to properly go through all the pages.

New Runtime UI 0.48.0.

Upgraded to latest version of Codemirror and updated usage throughout the application.

MS SQL CDC Connector: Added support for incremental only mode.

HubSpot connector: Fixed table creation on invalid object type.

Added StandardProtobufReader Controller Service for Protobuf record processing

Fixes health checks for Load Balancer Target Groups, so everything shows green in the AWS Console.

Supports a finer-grained privilege model for deployments and runtimes including MONITOR and OPERATE privileges.

Supports a finer-grained privilege model for deployments and runtimes including MONITOR and OPERATE privileges.

Supports a finer-grained privilege model for deployments and runtimes including MONITOR and OPERATE privileges.

MS SQL Server: Fixes handling of datetime when used as a primary key.

PostgreSQL connector: Supports TOASTed values.

Uses Google Ads API v21 (Note, v18 is no longer supported).

New Runtime UI 0.47.0.

Slack connectors: Fixes handling of attachments by appending the File ID to the filename for the files stored in the stage.

Adds Google Cloud support to PutSnowpipeStreaming2.

Adds support for Incremental Only mode in PostgreSQL CDC connector.

Fixes error when trying to verify configuration in List Azure processors.

Supports unquoted parameter references with spaces in their names within an expression language.

Resolves an issue that sometimes caused runtime deletion to fail in Snowflake deployments.

Resolves an issue facilitating runtime autoscaling in Snowflake deployments.

Improves readability in Provenance Event dialog.

New AWS BYO-VPC deployments now adds the “Private Security Group” to the EKS cluster, making it easier to configure connections to data sources.

Resolves an issue for new Deployments with a private security group configuration that couldn’t pull images from Snowflake over PrivateLink.

Runtime and Deployment action menus now have separators to help group actions.

Account roles show in a searchable selection with virtual scrolling.

Runtime flows no longer disappear after suspend and reactivate due to a conflicting auto scaling operation.

Adds FlowFile attributes support for Database and Schema properties in PutSnowflakeInternalStageFile.

New GetConfluenceSpaces processor.

PostgreSQL CDC now properly handles DATE, TIME, TIMESTAMP primary keys.

New Runtime UI 0.45.0: Minor improvements to the Component State dialog to improve readability of state entries.

Fixes issue destroying BYOC deployments that was introduced with 0.29.0.

Fixes issue from 0.29.0 release where BYOC deployments in AWS Regions with longer names may fail due to IAM Policy length limitations.

Upgrades the AMI of EKS nodes when the deployment is upgraded.

Removes unnecessary IPv6 Security Group rules for ingress and egress.

Improves ConsumeKafka by introducing an Inject Offset Output strategy to add a field kafkaOffset to the records.

Adds the preview tag for Salesforce, Confluence and HubSpot components.

Better configuration validation in UpdateSnowflakeDatabase to avoid using empty parameters.

Adds GetConfluencePageContent and GetConfluencePageIds processors for Confluence.

Fixes UpdateSnowflakeDatabase to properly redirect to the failure relationship when schema is not specified or does not exist.

Improves error handling of non-authorized calls in HubSpot processors.

New Runtime UI 0.44.0: Improves ConsumeKinesisStream by introducing a schema difference handling strategy to specify how records using the same schema should be grouped.

Fixes issue in rendering the canvas that surfaced on initial page load.

Adds Pipe Info Counter and Channel Error Message to PutSnowpipeStreaming2.

MySQL connector: Supports enabling the connector in Incremental mode only.

HubSpot connector: Improves handling of non-supported object types and fixed processing ordering of the events.

New Runtime UI 0.43.0: The Runtime UI now supports labeling extensions in Preview. The badge is shown in the create dialog, on the canvas, in the operate palette, in the edit dialog, and in listings for extensions not on the canvas.

Private deployments: All images and binaries are provided by Snowflake instead of various internet sources.

Custom Ingress for “Bring Your Own VPC” deployments: Supports enterprise customers who use VPNs to access their cloud infrastructure and self-managed TLS certificates.

Adds end-to-end support for PrivateLink. Previously, data and management communications were available over PrivateLink. Now, the deployment can install over PrivateLink, too.

Adds improvements necessary to support BYOC private deployments.

Improves handling of outbound grants when transferring ownership of a runtime or deployment.

Trial accounts are now permitted to use Openflow with relevant parameter enabled.

Fixes an issue that disrupted use of Control Plane for customers with a large number of Snowflake roles.

In runtime and deployment listings, more actions in the menus are disabled rather than hidden.

Removes a link to accept terms. This change prevents problems when the user doesn’t have an active Snowsight session.

When a new version is detected, prompts the user to reload the CP UI.

Disallows changing ownership of runtimes in Snowflake deployments.

Fixes bug that required a Snowflake role, even when the field was hidden.

Includes improvements necessary to support BYOC private deployments.

Fixes an issue that disrupted Connector deployment for customers with a large number of Snowflake roles.

Fixes an issue with certificate refresh upon renewal which prevented users from logging into older runtimes.

Jira: Improved readability of the flow. The scheduling is now exposed via a parameter.

Adds File Fragment Size and Count to PutSnowpipeStreaming2.

Introduces new Confluence processors for the upcoming connector GetConfluenceGroupUsers, GetConfluencePagePermissions, GetConfluenceSpacePermissions, ListConfluenceGroups.

Adds support for TOASTed value in PostgreSQL CDC.

Fixes initial rendering of canvas when fonts may load slowly.

Fixes parameter removal in Parameter Contexts owned by a Parameter Provider.

New Runtime UI 0.42.0: Improves formatting in Status History dialog when values are lengthy.

Fixes an issue with scaling that left some nodes in a disconnected state.

Kafka Connectors: Fixes referenced readers when writing to Iceberg formatted tables.

Fixes S3 Location Type in PutSnowpipeStreaming2.

Adds support for users to reset all Counters in a single action.

Fixes an issue that caused upgrade failure for runtimes with more than 1 node present.

Adds support for selecting an active role to use in the application, rather than relying on a default role and secondary role inheritance.

Adds support for considering Snowflake role hierarchy during authorization controls.

Adds support for selecting an active role to use in the application, rather than relying on a default role and secondary role inheritance.

Adds support for considering Snowflake role hierarchy during authorization controls.

Adds support for considering Snowflake role hierarchy during authorization controls.

A new controller service better supports Slack API rate limits.

Fixes SnowflakeSignJWT controller service.

Fixes upgrades to pull and use the latest host scripts. This change enables Openflow to more easily make changes to the agent itself during an upgrade.

Confluence JIRA connector: Improves type mapping for the JIRA issues. Uses the new processor for managing lifecycle of views.

Slack connectors: Changes defaults for run schedule properties to avoid rate limiting errors.

Adds support for generating and downloading runtime diagnostic bundles.

Adds support for generating and downloading runtime diagnostic bundles.

Adds support for generating and downloading runtime diagnostic bundles.

Adds the PutSnowpipeStreaming2 processor using SSv2.

New Runtime UI 0.40.0: Fixes a bug that prevented tooltips from closing on the canvas.

Adds support for PrivateLink redirects for the Launch Openflow button.

Fixes an issue where logout doesn’t log the user out if the user revisits soon after.

PostgresSQL, SQL Server and MySQL Connectors: Change to Journal creation process group to remove the false positive error bulletin for PutSnowpipeStreaming when it was asked to create channels on non yet existing tables/streams.

Users must have proper privileges before they can list or view a runtime.

Fixes a bug that caused runtime and deployment listings not to show and prevented creation of new resources.

Git Registry clients have the option to ignore parameter changes when versioning a new version of a flow.

New HubSpot processor to retrieve the schema of HubSpot objects.

New processor UpdateSnowflakeView to manage lifecycle of Snowflake views.

New controller service RemoveFieldRecordReader to drop fields on read.

Supports PostgreSQL Aurora.

CaptureChangeSQLServer generates a valid query when the primary key consists of multiple columns.

UpdateSnowflakeDatabase now checks only column types when required.

New Runtime UI 0.39.0

Improves colors in canvas for Process Group version control status.

Improves styling for better alignment with Balto colors.

Assets are no longer prevented from being re-uploaded in the Manage Assets dialog.

When using form control to increment a numeric value, output from a dirty Edit Processor form is no longer prevented.

Upgrades no longer get stuck when upgrading due to a missing Data Plane UI 0.7.0 image.

New deployments no longer fail to install due to mid-handling failure code when checking for the presence of AWS ECR repositories.

Google Ads: Limits the numbers of calls to Google Ads API when validating the components to avoid rate limit errors.

Fixes NullPointerException in PutSnowpipeStreaming when empty flow files are being processed and Delivery Guarantee is set to Exactly once.

New terms of service flow: Customers can use Control Plane to create Snowflake-managed deployments without accepting BYOC and Connector terms.

New terms of service flow: Customers can use Control Plane to create Snowflake-managed deployments without accepting BYOC and Connector terms.

Kafka Connectors: Ignore column type mismatch in UpdateSnowflakeDatabase for Kafka connectors is more resilient in case of issue with schema inference.

Google Drive & SharePoint Connectors: Improves the flow to avoid a race condition where group synchronization kicks off but PERMS_GROUPS has not been created yet

Kafka Connectors: Warehouse is no longer needed. The corresponding parameter is removed.

Tables without primary keys are retried instead of failed.

New Alter Strategy in UpdateSnowflakeDatabase processor has the option to ignore column type changes.

Fixes fetching of HubSpot archived records.

New deployments send status updates to Openflow Control Plane indicating when upgrades are present.

The PrPr tag is included on some new connectors.

New deployments now surface when an upgrade is available, with link to documentation. earlier deployments can also use this functionality after a migration to a newer version.

Fixes Create runtime failures where the minimum node count is greater than one.

Active role now displays in the current user menu.

Deployments created with AWS Data Plane Agent 0.20.0 are no longer prevented from adopting future updates to EC2 Agent Host scripts.

Supports tagging all AWS resources created and managed by Openflow. Enables deployments governed by security controls like AWS SCP and cost controls like AWS MAP.

New Openflow BYOC deployments and upgrades of existing deployments are no longer blocked by an “Unsupported block type” error.

JIRA: Multi-projects support flattened views in Snowflake destination.

Process Group metrics are now visible when using the Stateless engine.

The toolbar renders properly when font size is scaled in the browser settings.

The UnpackContent shows the TAR option again.

Google Sheets: Improves failure handling by retrying when ingesting data into Snowflake.

Workday: Uses TRUNCATE instead of REPLACE when possible on the destination table.

Sharepoint / Google Drive: Improves failure handling with proper retry / logging in case of failures.

SQL Server: Prevents stream staleness.

Box: Properly reflects permissions when groups are removed from files permissions in Box.

Google Drive (Simple Ingest) - Fixes handling of files being deleted.

Workday: Fixes clustering configuration to have the first processor run on the primary node only.

Skeleton loaders are now shown in the deployment and runtime listings when permissions are evaluated.

Skeleton loaders are now shown in Create Runtime and Add Connector to Runtime dialogs while options are loaded and permissions are evaluated.

Adds the possibility to specify multiple projects to fetch JIRA issues when using ‘Simple Search’.

Improves handling of all fields in the JIRA connector. Improves mapping into destination table by using an individual column per field.

Adds support for the PuPr of Snowflake Structured Maps/Arrays/Objects.

Google Sheets connector now supports Boolean and numbers to be used in the same column.

MySQL: Properly handles a changes in the column filtering parameter during replication.

MySQL: Fixes potential connection leakage when being disconnected from the binlog.

SQL Server: Fixes column ordering handling in the Journal Log table.

Error reporting now shows in banners instead of toast notifications.

Adds support for different ranges in the Status History dialog by selecting different start timestamps.

Introduces a Process Group column to the Parameter Context table to more efficiently see bound Process Groups.

Upgrades Snowflake JDBC Driver to 3.24.2

Resolves an issue that prevented newer runtimes from installing the latest Microsoft Dataverse Connector.

Removes Microsoft SQL Server replication of logical databases.

Adds support for logging in to Openflow runtimes using role names with dashes.

Adds pre-configured version control support for custom flows.

Gracefully shuts down processors and controller services for stateless process groups.

Add kafka.max.offset attribute to Records produced by ConsumeKafka