New Features: 

• Added new Read stage to Pipelines to read one or more Sources.
• Added support to Pipelines state variables to perform actions like track machine state.
• Added new Zip and Gzip stages to Pipelines.
• Added UNS Client to connect to MQTT servers and visualize the namespaces that support JSON and Sparkplug data payloads.
• Added a new SQLite Connector that supports memory or file-based storage.
• Added support for OPC UA Browse to Browse Variable type nodes. This allows UDTs from Siemens and Allen-Bradley ControlLogix to be browsed and read using Branch Reads.
• Changed Flow timing to trigger the Flow at the Flow rate. Prior to this change, the Flow triggered at the Flow rate plus the time it took for the Flow to finish reading all sources.
• Added support for viewing Flow Template Projections and creating Flows from the template settings.
• Added support to PI Event Frame reads to return child Event Frames.
• Added a pendingWrites Connection status to indicate how many writes are pending on a given Connection. Added a queuedWrites Pipeline status to indicate how many writes are waiting to be processed. Both pendingWrites and queuedWrite can be set to only trigger a Flow when all events have been fully processed.
• Enhanced the Modbus Connector to allow users to set the Unit ID at the Input level.
• Added support for disabling flows while they sync to remote hubs.
• Increased the max file ingest size from 32MB to 128MB.
• Removed the 10MB Store and Forward and Input Cache limit.
• Added support for compressed communication between the PI Agent and the Intelligence Hub, improving performance for large queries.
• Added a username and password authentication option to the PI Agent for cases where the Agent and PI Server are not on the same domain.
• Added MQTT connection option to enable and disable clean session.
• Added an option to the JDBC Driver Connector to control the quote/escape character.
• Previously imported OPC UA tags, MQTT topics, etc. are no longer hidden from the browse results in the UI.
• Changed the REST Client Output to only retry requests that fail because of a connection failure rather than 400/500 HTTP errors.
• Remote Clients connected to a Central Configuration Hub can now provide their own clientId GUID.
• Changed the URLs in the UI to use friendly names instead of GUID identifiers.
• Removed version number from the main intelligencehub-runtime.jar file.
• Added the ability to terminate the flow of a Pipeline using a Transform stage. This enables Pipelines to stop execution under certain logical conditions (e.g., Machine is off).
• Added the ability to cherry-pick elements from the compare page and sync them to a different hub.
• Updated the import and sync pages to show a summary result of the import or sync operation.
• Added an internal system source to read the system time.
• Added new Compression stages to Pipelines to support filtering On Change and Only Changes. This functionality is similar to Flow output compression.
• Added a timeout limit to the Pipeline Size Buffer Stage to send data after buffering N events or exceeding time T.
• Added support for viewing, editing, and deleting Pipeline state variables in the UI.
• Added a new Filter stage to Pipelines to enable attribute filtering in payloads and relocation to metadata.
• Added the ability to configure the base URL used by the REST API after login.
• Added output template support for Sparkplug Outputs.
• Increased the OPC UA browse timeout from 60 seconds to 5 minutes.

Fixes:

• Fixed various issues with JavaScript expressions where numeric values were interpreted as either Integer or Double data types. They are now always treated as Doubles.
• Fixed an issue where OPC UA Branch read result order did not always match the browse order.
• Fixed a bug where the runtime could become unresponsive when saving a Condition that referenced a templated Instance or Input.
• Fixed an issue where a Flow was not restarted if referencing a deleted Source or Target.
• Fixed a bug where circular Model references could cause overall runtime performance issues.
• Fixed an issue where a Pipeline Transform stage with long running execution could block expression execution in other parts of the product.
• Fixed an issue where bulk copy of inputs would result in invalid text in the result summary.
• Fixed various data type issues with OPC UA Input CSV import that resulted in missing or incorrect fields.
• Fixed a UI issue where the Connection title disappeared when trying to save an invalid Input that resulted in a 404 response.
• Improved the performance of runtime event logging to better handle overall performance when the runtime is generating many events.
• Fixed a concurrency issue where exceptions could be thrown when processing Dates.
• Fixed a bug introduced in version 3.1 where Instance names were not reflected in SPB output hierarchy.
• Fixed an issue where the GE iFix OPC UA server did not accept the Intelligence Hub’s self-signed certificate.
• Fixed an issue where a Flow could continue to execute after being removed from the project.
• Fixed an issue where the Timed Buffer stage in a Pipeline would stop working after a Pipeline becomes idle.
• Fixed a bug where the AWS IoT SiteWise Connection could not import Models/Assets if there were more than 250 models or 250 assets for a model.
• Fixed a bug where naming a Network Hub as “hub” created an error in the UI.
• Fixed a bug where the REST API would generate a self-signed certificate if the existing certificate in use was deleted.
• Fixed a bug where enabling store and forward on a Connection degraded output performance.

 
Security Patch Updates:

• JDBC Driver Connector
  • CVE-2023-32697: Remote code execution in attacker-controlled URL in JDBC Driver
• Apache Parquet Connector
  • CVE-2023-34455: Out-of-memory vulnerability due to unchecked chunk size in snappy-java
  • CVE-2023-2976: Files or Directories Accessible to External Parties
• Google Cloud PubSub Connector
  • CVE-2023-32731: Privilege escalation or data exfiltration due to missing header validation in grpc-core
• MQTT Broker, Azure Blob Storage, Google Cloud PubSub, OPC UA
  • CVE-2023-34462: OutOfMemory error due to unchecked input

Patch (3.2.1 2023.10.23.1070):

• Reduced error messages logged by Flows and Instances when there are reference errors.
• Fixed an issue on OPC UA connectivity loss where UA subscriptions were not re-created properly, resulting in read errors.
• Improved performance of Store and Forward for all Connections.
• Added the Browse Variable Nodes option to OPC UA outputs. This was added for OPC UA Inputs and may also be needed in some cases for Outputs when writing complex data.
• Fixed a Sparkplug Output issue where bad quality data was published as an Unknown data type, causing Ignition to fail to parse the metric and show the tag as bad quality.
• Changed the Sparkplug Output to not publish the default “HighByte IntelligenceHub” group if it’s not being used.
• Fixed a bug in the SiteWise Output where, on certain Linux operating systems, the Output would leak threads and result in an out of memory error.
• Fixed an issue where missing dynamic outputs were not logging errors on write.

​Patch (3.2.2. 2023.11.9.1115):

• Fixed a bug where dynamic output references that include JavaScript (e.g., ) could cause long timeouts when writing data.