Base Command Manager / Bright Cluster Manager Release Notes
Release notes for Bright 8.1-23
== General ==
- Improvements
* Updated cm-openssl to 1.1.1k
* Added CUDA 11.0 packages
* Added CUDA 11.1 packages
* Updated cuda-driver to version 440.95.01
* Updated mlnx-ofed50 to version 5.0-2.1.8.0
== cmdaemon ==
- Improvements
* By default, perform a daily malloc trim to reduce memory usage, controlled by adv. config. option MallocTrimInterval
* Disable all monitoring recorders during an upgrade
* Possible memory leak in pexec
* Possible memory leak when adding or removing a large number of nodes
- Fixed Issues
* cmsh and pythoncm clients may stop receiving updates from cmdaemon while setting up AWS cluster extension
* Possible memory corruption when using file customization
* An issue with settings slurm job queue GPU parameters (DefCpuPerGPU and DefMemPerGPU) via Options parameter
* Memory leak when using ssl sessions
* Hadoop access thread is being started also on clusters without Hadoop
* An issue with parsing of commas in openssl subject fields of existing license in request-license
* An issue where the number of CPUs is shown as 0 for Torque jobs in Bright-view
* An issue with showing memory information in Bright-View
== node-installer ==
- Fixed Issues
* An issue where the node-installer may force full provisioning even if a bios_grub partition has already been created during previous full provisioning
== cm-docker-setup ==
- New Features
* Docker registry proxy can be configured from cm-docker-setup
== cmsh ==
- Fixed Issues
* Monitoring commands do not support delimiter
== ml ==
- New Features
* Introduced cm-ml-distdeps for CUDA 11.2
* Introduced cm-ml-pythondeps for CUDA 11.2
* Introduced cm-chainer for CUDA 11.2
* Introduced cm-cudnn8.0-cuda11.1 package (v8.0.4)
* Introduced cm-cudnn8.0 for CUDA 11.2
* Introduced cm-cudnn8.1 for CUDA 10.2 and CUDA 11.2
* Introduced cm-cutensor for CUDA 10.2 and CUDA 11.2
* Introduced cm-fastai2 for CUDA 10.2
* Introduced cm-nccl2 for CUDA 11.2
* Introduced cm-open3d-* packages (v0.10.0)
* Introduced cm-opencv4-* packages
* Introduced cm-openmpi-geib for CUDA 11.2
* Introduced cm-tensorflow2-extra-* for CUDA 10.2
* Introduced cm-xgboost for CUDA 11.2
* Extended support for CUDA 10.2 with cm-chainer-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-dynet-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-fastai-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-gpytorch-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-horovod-*-py37-cuda10.2-gcc packages
* Extended support for CUDA 10.2 with cm-mxnet-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-open3d-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-tensorflow-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-theano-py37-cuda10.2-gcc package
* Extended support for CUDA 10.2 with cm-xgboost-py37-cuda10.2-gcc package
* Updated cm-chainer-* packages to v7.7.0
* Updated cm-cmake-* packages to v3.18.4
* Updated cm-cudnn8.0-* packages to v8.0.5.39
* Updated cm-cudnn8.1-* packages to v8.1.1.33
* Updated cm-dynet-* packages to v2.1.2
* Updated cm-fastai-* packages to v1.0.63
* Updated cm-gpytorch-* packages to v1.3.1
* Updated cm-horovod-* packages to v0.21.3
* Updated cm-mxnet-* packages to v1.7.0
* Updated cm-nccl2-* packages to v2.7.8
* Updated cm-onnx-* packages to v1.8.1
* Updated cm-opencv3-* packages to v3.4.13
* Updated cm-opencv-* packages to v3.4.10
* Updated cm-pytorch-* packages to v1.7.1
* Updated cm-tensorflow2-* packages to v2.4.1
* Updated cm-tensorflow-* packages to v1.15.5 (end-of-life reached)
* Updated cm-tensorrt-* packages to v7.2.2
* Updated cm-theano-* packages to v1.0.5
* Updated cm-xgboost-* packages to v1.3.3
- Changes and Deprecated Features
* Deprecated -py36- variants of Machine Learning packages. Customers are encouraged to switch to -py37- packages to benefit more frequent updates. Uninstallation of old -py36- packages is not strictly necessary, since they can coexist with -py37- packages.
* Renamed cm-tensorrt-cuda10.2-gcc package to cm-tensorrt-cuda10.2
* Switched GCC support for several ML packages from GCC 5 (-gcc) to GCC 8 (-gcc8)
* Deprecated cm-bazel package
* Deprecated cm-fastai-* packages
* Deprecated cm-tensorflow-*, cm-horovod-tensorflow-*, cm-onnx-tensorflow-* and cm-keras-* packages
== monitoring ==
- Improvements
* Exclude az* proc/net/dev network interfaces from monitoring
== slurm ==
- Improvements
* Updated to 20.02.5 and 19.05.8