OpenStack HA support by having the ability to provision multiple controller nodes
Plugins depending on SSH feature of Jenkins (such as Validated Merge plugin) are now supported on PSE.
Update provided CloudBees Jenkins Enterprise to 1.642.4.2
Create a dedicated worker pool for Jenkins build agents.
PSE 1.1.0 will install more quickly when using these AMIs. The following table shows the AMIs for their corresponding AWS regions.
us-east-1 = ami-9cfe3ef1 eu-west-1 = ami-979900e4 ap-southeast-1 = ami-f65f8c95 ap-southeast-2 = ami-b37058d0 eu-central-1 = ami-4a846c25 ap-northeast-2 = ami-153bf07b ap-northeast-1 = ami-cca155ad sa-east-1 = ami-d07ff5bc us-west-1 = ami-12fabf72 us-west-2 = ami-01cb0f61
PSE logs and performance metric data are backed up
CLI operations more easily scriptable; Operations' arguments can be supplied right on the command line.
AWS Enhanced Networking is supported when using the appropriate instance types.
Added the ability to verify argument correctness before executing operations
Worker nodes on OpenStack can now be accessible from outside networks
Master names can include whitespace characters
Update to using Terraform 0.6.14
Castle to apply logging level changes at run-time without having to restart
PSE resources on AWS are now tagged with "cloudbees:pse:cluster"
User can explicitly specify the Docker image to be used for launching CJOC
Unused NFS mounts and mount points are now cleaned up properly.
Removed the ability to specify domain-name at PSE cluster initialization time, as doing so is a manual process that involves customizing/requesting DNS names.
Documented how to retrieve CPU core usage information
New logging will take affect after upgrading without having to restart the storage manager processes
Fixed various issues after the "init" flow has been converted to a standard operation
The operation "elasticsearch-expire-indices" was renamed to "elasticsearch-indices-delete" to be consistent with other existing operations
Fixed an issue where the plugins "cloudbees-workflow-template" and "workflow-cps-checkpoint" could end up disabled.
Fixed an issue where HTTP 404 error was issued when upgrading a PSE cluster where its CJOC already has a license
Fixed an issue where worker node initialization encountered "apt-get" errors
Fixed an issue where "bees-pse prepare" without any arguments failed ungracefully.
Fixed issue where internal Elasticsearch referral URL can be wrong
Fixed issue where ArrayIndexOutOfBoundsException could occur when using AWS backend in storage manager.
Fixed issue where Logstash has wrong ElasticSearch config when running in path_mode