Release Notes for Apprenda 8.0.0

The following are release notes for Apprenda Cloud Platform version 8.

Release Date: September 28, 2017

Major Features

  • The Platform now supports Always On Availability Groups for guest application databases. Platform Operators are able to forward deploy guest application databases onto external, off-Platform, database servers while retaining scaling and patching capabilities through the Platform. Platform Operators and DBAs can use this feature to support enterprise-level workflows for High Availability and credential rotation
  • Always On Availability Groups Support for Platform Databases is now availably on the Platform. Platform Operators can provide enterprise-level High Availability and Disaster Recovery scenarios when running an Apprenda Cloud Platform in single or multiple data centers
  • The Platform now supports SQL Server 2016
  • Developer Portal performance utilization enhancements

Minor Features and Improvements

  • Platform Audit Logs can now have an extension service created to process, filter, and deliver off-Platform using the standard Apprenda Platform Extension model
  • Users can now map ports for Docker containers deployed on the Platform easier by using the new Docker Port Mapped Custom Property to define the ports to use

Breaking Changes

  • The minimum supported version of SQL Server is now SQL Server 2012 SP3. SQL Server 2012 without SP3 installed is no longer supported. You must upgrade to SQL Server 2012 SP3 or later before upgrading to Platform version 8.0.0

Bug Fixes

Apprenda Installer & Upgrade

  • The Platform Installer was not using the correct connection information during upgrade on Platforms where SQL password rotation was performed. This has now been fixed, and upgrades can be performed normally after password rotation
  • The Platform can now be installed when Swedish collation is used for the Platform Core database
  • The Platform was not properly removing old versions of the Cluster Manager service in Platforms that had upgraded from 7.0.0 to 7.1.0 only. The problem would then persist in upgrades to subsequent Platform versions. If you have upgraded your Platform from 7.0.0 to 7.1.0, contact your support representative for more information on how to resolve this issue until you upgrade to Platform version 8.0.0.

System Operations Center (SOC)

  • Logs from the Load Manager were not being assigned the correct version information and as a result were not caught by Log Overrides created for the Load Manager
  • Bootstrap Policies for Kubernetes applications could not be created through the Platform Operations Rest API because the api request was not setting all the correct values and would result in a HTTP 500 error
  • The Platform would attempt to remove Audit logs from the Platform when auditing was not enabled on the Platform. This resulted in generating unnecessary logs on the Platform
  • The initial Development Team user created at Platform installation was unable to be removed as a Platform Operator
  • Improved the display for the repository location in the Core Components box on the Infrastructure>Servers page
  • Fixed clearfix style issue that prevented multi-select Custom Property options for servers from being made visible in the server Custom Property editing window

Security

  • Increased security around information in cookies for Platform UIs and how cookies were displayed in the event of a error
  • An authenticated user could modify the securables of an application in the Account Portal by obtaining the application and version IDs

Developer Portal and Tools

  • Removed information about patching single files in the Additional Controls section of the Developer Portal. As of Platform version 6.6.0, you are only able to patch whole archives or only persistence files for an application
  • The Resource Utilization Tracking graphs on an application's Monitor>Components page did not display correctly after the first time a workload was expanded
  • An application's start button was not click-able after there was an error restarting the application
  • Improved UI by fixing typos, making information and error message formatting on an application's Patch and Scale page more readable, and removing styling that made plain text look like a link on some tooltips
  • Removed the options to add or remove non-database files using the patch builder in the Additional Controls
  • The Save and Cancel buttons remained disabled after making changes to Resource Policies on an application's Scale>Policies page
  • The Platform was throwing an error in some cases when a request was made to an application with no claims

Guest Applications Creation and Management

  • Current instance count settings for .NET UIs were not being honored when a Published application was patched and the target stage was also Published. Instance counts for those components would reset to the Platform default (1) or the instance count set in the Deployment Manifest for the patching archive, instead of using the instance count defined for the application being patched
  • The way the Platform was undeploying UI workloads when scaling down instances could cause HTTP 502 errors
  • WCF Services were failing to start because the virtual account running the service did not have read or execute permissions to access the launcher server files. The Platform will now grant then permissions to a virtual account when staring up a WCF Service
  • When patching an application, any custom tokens defined in the Deployment Manifest were not correctly token switched in the patched version when the version was promoted. This bug was introduced in 6.6.0
  • The Platform was not honoring user set CPU limits for the first Java workloads when deployed to a node running systemd 219 v 30+ and Resource throttling was enabled

Misc.

  • Logstash service management was migrated to systemd on RHEL/CentOS 7, enabling the service to restart automatically in the event of a crash