Release Notes for Apprenda 8.1.0

Release Date: December 15, 2017

Minor Features and Improvements

  • The Platform now supports, Windows Server 2016, .NET 4.7, and TLS 1.2
  • The Developer Portal application list and application dashboard was improved to reduce the number of backend API calls per page load, change the underlying architecture of the REST API implementation, and only load certain information when requested by users. These changes significantly improve the performance, scale, and page load time for the Developer Portal

Breaking Changes

  • Oracle 11g is no longer supported

Bug Fixes

Apprenda Installer & Upgrade

  • The Apprenda Installer was preventing the addition of a Cloud to a Platform if one node in the new Cloud only had the roles of Platform Coordination, Load Manager, and Cache. This has been fixed to allow the addition of a Cloud to your Platform with a node with those roles
  • During the preparation phase of an upgrade, the Apprenda Installer might attempt to deploy Platform WCF Services to nodes that were offline. The preparation steps would fail because the selected host was unavailable. Offline nodes are now filtered out as potential hosts during these steps so the Installer will only try to deploy to online servers

System Operations Center (SOC)

  • Fixed bug that could prevent the addition or removal of Platform Operators groups on the Platform
  • CPU throttling was causing unusually high resource consumption in the Windows Container
  • Logic for CPU throttling was adjusted to make sure small CPU policies are applied on large nodes


  • Improved security around redirecting users to off Platform content based on HTTP Header and URL parameter input
  • Improved cross site scripting protection in Platform UIs

Developer Portal and Tools

  • Deploying a service instance from the Additional Controls section of the Developer Portal was reporting a successful deployment even when the workload would fail to deploy
  • Scaling down UIs asynchronously using the instance count manager on the Scaling page in the Developer Portal could result in HTTP 502 errors for users attempting to view the UI
  • When using an Eternal User Store with the Platform, changing a user’s Role would sometimes cause the user to be unable to login if the Role was assigned to more than one Tenant
  • Group access to Multi-tenant applications was not being scoped to the Tenant that granted access
  • The ACS NewPackage command was not creating logs. That command will now log error messages to %temp%/Log/SGAB*.txt

Guest Applications Creation and Management

  • Introduced in ACP 8.0.0, applications using Windows authentication for DB connections were prevented from being promoted and deployed on the Platform
  • Fixed deadlocking that would occur when multiple users from the same tenant would try to create applications at the same time
  • Fixed an issue on Linux nodes where multiple duplicate processes could be spawned for Java and Linux service components
  • Improved logging for Platform healing if a Windows service fails to start


  • In the event of a out of memory (oom) error on a Linux node, the Logstash service was being repeated killed by the oom killer on the node. This would cause the Platform to place the node into a reserved state when the node failed the Logstash health check. This could result in resource problems on your Platform, as nodes in the reserved state cannot receive new workload deployments.