Base Command Manager / Bright Cluster Manager Release Notes

Release notes for Bright 8.0-4

== cmdaemon ==

- New Features

* Improved validation of Kubernetes labels, the correct format should be "key=value"

- Fixed Issues

* Bonding configuration for SLES to consider connectedMode
* An issue which caused the cloud nodes to try to start openvpn service also when using DirectConnect(AWS) or ExpressRoute(Azure)
* An issue with cloud nodes name resolution on the head node when DirectConnect/ExpressRoute is in use
* Public IP always assigned to cloud director even if DirectConnect/ExpressRoute was in use
* An issue with cm-scale{-cluster} service in the case of slurm job arrays
* Monitoring power actions could prevent cmdaemon from stopping
* An issue with removing slurm queues when slurmserver role is offloaded from the head node
* Restart of CMDaemon on the active head node results in the restart of the managed services
* In some cases, CMDaemon showing out-of-date provisioning hosts as up-to-date
* Fixed bmcsol script, previously it failed when Lmod was used
* Do not allow to set workload manager queue parameter in Options when the parameter is configured also via separate queue parameter (which can lead to failing slurm service)
* grabimage -w doesn't sync the extended attributes to an empty directory
* Added new mounts.pl config file to handle custom mappings for FUSE filesystems for the mounts healthcheck
* Removed message about node installer disk size validation that appeared after Azure director power-on

== node-installer ==

- Fixed Issues

* Bring down the bridge interfaces before pivoting to local root filesystem
* The node-installer doesn't create the ifcfg- files for the slave interfaces when the bond is a member of bridge
* An issue with unmount of /localdisk/selinux when SELinux is enabled for RHEL7

== cluster-tools ==

- Improvements

* Support for Apache Kafka 0.11.0.0
* Support for Cloudera CDH 5.10.2
* Support for Hortonworks HDP 2.5.6
* Support for Apache Cassandra 3.11
* Support for Apache Flink 1.3.1
* Support for Apache Pig 0.17.0
* Support for Spark on BeeGFS
* Support for Cloudera CDH 5.11.1
* Support for Alluxio 1.5.0
* Support for Apache Zeppelin 0.7.2

- Fixed Issues

* In cm-cluster-extension, no region is now pre-selected to avoid accidental selection of multiple regions
* An issue with cm-cluster-extension during Azure deployment in network environments with restricted network connectivity

== Bright View ==

- New Features

* New and more stable cm-kubernetes-setup based with ncurses dialogs

- Improvements

* Support for Hortonworks HDP 2.5.6
* Support for Apache Cassandra 3.11

- Fixed Issues

* Bright View showing incorrect workload information for Slurm

== cm-docker-registry-setup ==

- New Features

* Remove multiple nodes selection in cm-docker-registry-setup to prevent confusion
* Allow docker-registry to be deployed on head nodes
* Allow the user to specify multiple alternative subject names for the SSL certificate used for docker-registry

- Fixed Issues

* Remove reboot step from cm-docker-registry-setup and replace it with an imageupdate.

== cm-kubernetes-setup ==

- New Features

* New and more stable cm-kubernetes-setup based with ncurses dialogs

- Fixed Issues

* An issue with kubernetes scheduler configuration

== cm-scale ==

- New Features

* Extra nodes management
* Mesos tracker for scaling up or down the nodes depending on the current load

== cod ==

- New Features

* COD for OpenStack: added ability to select volume type for the head and compute nodes VMs

- Fixed Issues

* COD for OpenStack: "cluster list" subcommand: fixed inconsistencies with reporting the creation time of cluster vs. the creation of the image used to create that cluster

== monitoring ==

- Fixed Issues

* mounts healthcheck fails for bind mounted filesystems