Base Command Manager / Bright Cluster Manager Release Notes
Release notes for Bright 9.0-10
== General ==
- Improvements
* Update cuda-driver to 450.80.02
* Update nvidia-container-toolkit to 3.4.0 and libnvidia-container to 1.3.0
* Added mlnx-ofed51 packages
* Update cuda11.0 to version 11.0 update 1
- Fixed Issues
* Cloud nodes aren't marked as DOWN immediately after they are powered off
* In some cases, an issue with the cmsh image lock / unlock command
== cmdaemon ==
- New Features
* Improved job monitoring for UGE and LSF
- Improvements
* Fixed possible memory leak when adding or removing a large number of nodes
* Improved handling of very large log messages
* Limit the number of bad certificate SSL log messages in the cmdaemon log file
* Added memory / swap total to partition metrics
* Added node based version information
* New adv. config. flags to disable individual sysinfo collection
- Fixed Issues
* An issue with DefaultTime in Slurm JobQueue configuration
* New adv. config. flag JobInformationDisabled=1 to disable job information parsing
* REST calls do not return full measurable name, :parameter missing
* /etc/localtime cannot be frozen from cmd.conf
* dumpmonitoringdata --min argument returns the sum instead of the minimum values across entities
* Exporting and re-importing cmd configuration can cause loss of monitoring healthchecks
* An issue with usernodelogin option of the node category when it set to onlywhenjob and the job prolog is enabled
* Support Kubernetes user names that include a dot
* Reduce the memory consumption of the PBSPro job end time parser
== node-installer ==
- Fixed Issues
* The node-installer attempts to always use the base partition BMC credentials even if it is overridden at the node/category level
== cluster-tools ==
- New Features
* Added slurm power scripts
== Bright View ==
- Improvements
* Selecting a node in Bright View does not it on the dashboard by default.
- Fixed Issues
* An issue with reloading entities when they changed from cmsh or other bright view session
== cm-kubernetes-setup ==
- Fixed Issues
* Make sure the NVIDIA device plugin addon deployment is fully compatible with the PodSecurityPolicies feature
* Possible race condition in the case when CA generation is scheduled for a retry in cm-kubernetes-setup
== cmsh ==
- New Features
* Easier way to create monitoring multiplexers in cmsh
== head node installer client ==
- Fixed Issues
* An issue with loading of custom kernel modules
== jupyter ==
- Improvements
* Introduced support for Jupyter on Ubuntu 18.04, CentOS 8 and RHEL 8
* Deprecated cm-jupyter-eg-kernel-* packages in favor of builtin Jupyter kernel templates
== ml ==
- New Features
* Introduced cm-cudnn8.0-cuda11.1 package (v8.0.4)
* Updated cm-cmake-* packages to v3.18.3
* Updated cm-tensorflow-* packages to v1.15.4 to address some vulnerability issues
* Updated cm-horovod-* packages to v0.20.2
* Updated cm-horovod-* packages to v0.20.0
* Updated cm-xgboost-* packages to v1.2.0
* Updated cm-mxnet-* packages to v1.7.0
* Updated cm-gpytorch-* packages to v1.2.0
* Updated cm-fastai-* packages to v1.0.63
* Updated cm-cmake-* packages to v3.18.2
* Updated cm-pytorch-* packages to v1.6.0
* Updated cm-opencv3-* packages to v3.4.11
* Updated cm-nccl2-* packages to v2.7.8