Release notes

This page contains a list of features, changes, known issues and known limitations of the Nerve releases.

Version 2.1.1

This version was released on August 14, 2020.

Feature list

Bug fixes

• Local workload deployment: An issue that prevented local workload deployment due to file size has been fixed.
• Roles and permissions: Users without admin rights were able to assign admin rights to themselves in certain situations. This has been fixed.
• Fixed an issue where unauthorized access to the Management System was possible in some cases.
• User management: Fixed display and functionality issues.
• Fixed an issue where port forwarding for Docker containers would not survive a reboot.
• Remote connections: In certain situations a fallback to the Guacamole UI could occur. This has been fixed.
• Remote connections: Fixed an issue that would have remote connections displayed as active when a node goes offline.
• Remote connections: In some cases, users could access established remote connections without approval. This has been fixed.
• Fixed an issue where the CPU limit parameters for Docker workloads were not properly used.

Known limitations

• 30 to 50 % of the time a license is activated on the Kontron KBox A-250, the license activation might fail with an error. For online activation, it might fail with Fetched update file is empty! Unable to deploy license!. For offline activation, it might fail with an internal error. The solution is to re-install the Nerve software and attempt the license activation again.
• Nerve Data Services: A number of small display issues in the database preview table can occur. The most severe occurs when more than 9000 entries or more than 40 variables are selected. This renders the browser unresponsive. Selecting the option to wait can make the browser responsive again.
• Updating a node from version 2.1.0 to 2.1.1 may fail with an unexpected end of file error. Retry the update and the update will be eventually successful after a certain number of tries.
• Provisioning a virtual machine workload: When defining the size of the virtual machines raw image, the value must be a factor of 512 Bytes. Entering a value that is not a factor of 512 B will cause an error when the virtual machine workload is deployed.

Known issues

• The Docker default network bridge is automatically assigned to all Docker workloads even if it was not explicitly defined.
• Workload Management: When a new Docker workload version is added while a Docker workload is being deployed to a node, the workflow will be interrupted. The user will be redirected to the deployment log.
• Workload Management: If a node goes offline during the downloading phase of a deployment, the deployment remains in downloading state until the node comes back online.
• Workload Management: Under certain circumstances, a virtual machine workload can be stuck in the Idle state after deployment.
• When performing a factory reset on a node, a new license is required to activate the system.

• If the deployment of a virtual machine workload fails during the download phase, the logical volumes are not removed.

• Nerve Data Services:

  • OPC UA Publisher sends random data via MQTT if multiple MQTT based inputs or outputs are used in the Gateway.
  • The Gateway is restarted twice when a new configuration is applied.
  • After a reboot of the node, data received by the Data Services is sometimes not stored in the database.
  • The node freezes when visualizing huge amounts of data in Grafana.
  • The database becomes unavailable when the system runs out of storage space.
  • Data retention settings are set back to default values after updating the system.
  • The network interface for the OPC UA Publisher cannot be set using the networkInterface configuration object.
  • The OPC UA PubSub SUB IP is not receiving data from all host interfaces.

Scaling and performance limitations

This release has been tested to perform within the following scaling boundaries: