# Release notes and document history
<a name="document_history"></a>

The following table describes the major updates and new features for the *Amazon ParallelCluster User Guide*. We also update the documentation frequently to address the feedback that you send us.

| Change | Description | Date | 
| --- |--- |--- |
| [Documentation only release](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#version2-guide) | Amazon ParallelCluster version 2 specific user guide published.<br /> Documentation only release:  Amazon ParallelCluster version 2 has its own separate user guide.   | July 17, 2023 | 
| [Amazon ParallelCluster version 2.11.9 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.9) | Amazon ParallelCluster version 2.11.9 released.<br /> Bug fixes:  Prevent replacement of managed FSx for Lustre file systems and loss of data on cluster updates that include changes to `vpc_security_group_id`.  <br />For details of the changes, see the `CHANGELOG` file for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.9/CHANGELOG.md) package on GitHub. | December 2, 2022 | 
| [Amazon ParallelCluster version 2.11.8 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.8) | Amazon ParallelCluster version 2.11.8 released.<br /> Changes:  Upgrade Intel MPI Library to Version 2021 Update 6 (updated from Version 2021 Update 4). For more information, see [Intel® MPI Library 2021 Update 6](https://software.intel.com/content/www/us/en/develop/articles/intel-mpi-library-release-notes-linux.html#inpage-nav-3-1).   Upgrade EFA installer to 1.19.0   Efa-driver: `efa-1.16.0-1`   Efa-config: `efa-config-1.11-1` (from `efa-config-1.9-1`)   Efa-profile: `efa-profile-1.5-1` (no change)   Libfabric-aws: `libfabric-aws-1.16.0-1` (from `libfabric-1.13.2`)   Rdma-core: `rdma-core-41.0-2` (from `rdma-core-37.0`)   Open MPI: `openmpi40-aws-4.1.4-3` (from `openmpi40-aws-4.1.1-2`)     Upgrade Python runtime, that's used by Lambda functions in the Amazon Batch integration, to python3.9.   Bug fixes:  Prevent cluster tags from being changed during an update because it's not supported.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.8/CHANGELOG.md) package on GitHub. | November 14, 2022 | 
| [Amazon ParallelCluster version 2.11.7 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.7) | Amazon ParallelCluster version 2.11.7 released.<br /> Changes:  Upgrade Slurm to version 20.11.9.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.7/CHANGELOG.md) package on GitHub. | May 13, 2022 | 
| [Amazon ParallelCluster version 2.11.6 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.6) | Amazon ParallelCluster version 2.11.6 released.<br /> Enhancements:  Improve exception management in case of missing networking.   Changes:  OS package updates and security fixes.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.6/CHANGELOG.md) package on GitHub. | April 19, 2022 | 
| [Amazon ParallelCluster version 2.11.5 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.5) | Amazon ParallelCluster version 2.11.5 released.<br /> Enhancements:  Add support for `NEW_CHANGED_DELETED` as value of FSx for Lustre `AutoImportPolicy` option.   Remove support for SGE and Torque schedulers.   Disable `log4j-cve-2021-44228-hotpatch` service on Amazon Linux to avoid incurring potential performance degradation.   Changes:  Upgrade NVIDIA driver to version `470.103.01` (from `470.82.01`).   Upgrade NVIDIA Fabric manager to version `470.103.01` (from `470.82.01`).   Upgrade CUDA library to version `11.4.4` (from `11.4.3`).   [Intel MPI](intelmpi.md) updated to Version 2021 Update 4 (updated from Version 2019 Update 8). For more information, see [Intel® MPI Library 2021 Update 4](https://software.intel.com/content/www/us/en/develop/articles/intel-mpi-library-release-notes-linux.html#inpage-nav-3-1).   Extend head node creation timeout to one hour.   Bug fixes:  Fix DCV connection through browsers.   Fix YAML quoting to prevent custom Tags from being parsed as numbers.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.5/CHANGELOG.md) package on GitHub. | March 1, 2022 | 
| [Amazon ParallelCluster version 2.11.4 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.4) | Amazon ParallelCluster version 2.11.4 released.<br />Changes include:  CentOS 8 support removed. CentOS 8 reaches end of life (EOL) on December 31, 2021.   Upgrade Slurm Workload Manager to version 20.11.8.   Upgrade Cinc Client to `17.2.29`.   [Amazon DCV](dcv.md) updated to Amazon DCV 2021.2-11190. For more information, see [DCV 2021.2-11190— October 11, 2021](https://docs.amazonaws.cn/dcv/latest/adminguide/doc-history-release-notes.html#dcv-2021-2-11190) in the *Amazon DCV Administrator Guide*.   Upgrade NVIDIA driver to version `470.82.01` (from `460.73.01`).   Upgrade CUDA library to version `11.4.3` (from `11.3.0`).   Upgrade NVIDIA Fabric Manager to `470.82.01`.   Disable package update at instance launch time on Amazon Linux 2.   Disable unattended package update on Ubuntu and Amazon Linux 2.   Install Python 3 version of [Amazon CloudFormation helper scripts](https://docs.amazonaws.cn/AWSCloudFormation/latest/UserGuide/cfn-helper-scripts-reference.html) on CentOS 7 and Ubuntu 18.04. (These were already used on Amazon Linux 2 and Ubuntu 20.04.)  <br />Fixes include:  Disable update of [`ec2_iam_role`](cluster-definition.md#ec2-iam-role) parameter.   Fix the `CpuOptions` configuration in the launch template for T2 instances.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.4/CHANGELOG.md), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.11.4/CHANGELOG.md) and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.11.4/CHANGELOG.md) packages on GitHub. | December 20, 2021 | 
| [Amazon ParallelCluster version 2.11.3 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.3) | Amazon ParallelCluster version 2.11.3 released.<br />  Fix [`pcluster createami`](pcluster.createami.md) failure due to Son of Grid Engine sources not being available at `arc.liv.ac.uk`.   Upgrade [Elastic Fabric Adapter](efa.md) installer to 1.14.1 (from 1.13.0)  EFA config: `efa-config-1.9-1` (from `efa-config-1.9`)   EFA profile: `efa-profile-1.5-1` (no change)   EFA Kernel module: `efa-1.14.2` (from `efa-1.13.0`)   RDMA core: `rdma-core-37.0` (from `rdma-core-35.0amzn`)   Libfabric: `libfabric-1.13.2` (from `libfabric-1.13.0amzn1.0`)   Open MPI: `openmpi40-aws-4.1.1-2` (no change)   GPUDirect RDMA is always enabled if supported by the instance type.  The [`enable_efa_gdr`](cluster-definition.md#enable-efa-gdr) and [`enable_efa_gdr`](queue-section.md#queue-enable-efa-gdr) configuration options have no effect.  <br />For details of the changes, see the `CHANGELOG` files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.3/CHANGELOG.md), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.11.3/CHANGELOG.md) and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.11.3/CHANGELOG.md) packages on GitHub. | November 3, 2021 | 
| [Amazon ParallelCluster version 2.11.2 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.2) | Amazon ParallelCluster version 2.11.2 released.<br />Changes include:   Do not install EFA with GPUDirect RDMA (GDR) enabled at bootstrap time if EFA is installed in the base AMI.   Lock version of `nvidia-fabricmanager` package to remain in sync with the NVIDIA driver version installed by Amazon ParallelCluster.   Slurm: Fix issue caused when cluster was stopped and restarted while a node was powering up.   [Elastic Fabric Adapter](efa.md) installer updated to 1.13.0:   EFA config: `efa-config-1.9` (no change)   EFA profile: `efa-profile-1.5-1` (no change)   EFA Kernel module: `efa-1.13.0` (no change)   RDMA core: `rdma-core-35.0amzn` (from `rdma-core-32.1amzn`)   Libfabric: `libfabric-1.13.0amzn1.0` (from `libfabric-1.11.2amzn1.1`)   Open MPI: `openmpi40-aws-4.1.1-2` (no change)     When using a custom AMI with a pre-installed EFA package, no changes are made to EFA at node bootstrap time. The original EFA package deployment is preserved.   <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.2/CHANGELOG.md) and [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.11.2/CHANGELOG.md) packages on GitHub. | August 27, 2021 | 
| [Amazon ParallelCluster version 2.11.1 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.1) | Amazon ParallelCluster version 2.11.1 released.<br />Changes include:   Mount file systems using the `noatime` mount option to stop recording last access time when a file is read. This improves the performance of the remote file system.   [Elastic Fabric Adapter](efa.md) installer updated to 1.12.3:   EFA config: `efa-config-1.9` (from `efa-config-1.8-1`)   EFA profile: `efa-profile-1.5-1` (no change)   EFA Kernel module: `efa-1.13.0` (from `efa-1.12.3`)   RDMA core: `rdma-core-32.1amzn` (no change)   Libfabric: `libfabric-1.11.2amzn1.1` (no change)   Open MPI: `openmpi40-aws-4.1.1-2` (no change)     Retry installations of the `aws-parallelcluster` package on the head node when using Amazon Batch as the scheduler.   Avoid failures when building SGE on an instance type with more than 31 vCPUs.    Pinned to version 1.247347.6 of the Amazon CloudWatch Agent to avoid issues seen in version 1.247348.0.  <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.1/CHANGELOG.md) and [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.11.1/CHANGELOG.md) packages on GitHub. | July 23, 2021 | 
| [Amazon ParallelCluster version 2.11.0 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.11.0) | Amazon ParallelCluster version 2.11.0 released.<br />Changes include:  Added support for Ubuntu 20.04 (`ubuntu2004`) and removed support for Ubuntu 16.04 (`ubuntu1604`) and Amazon Linux (`alinux`). Amazon Linux 2 (`alinux2`) remains fully supported. For more information, see [`base_os`](cluster-definition.md#base-os).   Removed support for Python versions below 3.6.   Default root volume size increased to 35 gibibytes (GiB). For more information, see [`compute_root_volume_size`](cluster-definition.md#compute-root-volume-size) and [`master_root_volume_size`](cluster-definition.md#master-root-volume-size).   [Elastic Fabric Adapter](efa.md) installer updated to 1.12.2:   EFA config: `efa-config-1.8-1` (from `efa-config-1.7`)   EFA profile: `efa-profile-1.5-1` (from `efa-profile-1.4`)   EFA Kernel module: `efa-1.12.3` (from `efa-1.10.2`)   RDMA core: `rdma-core-32.1amzn` (from `rdma-core-31.2amzn`)   Libfabric: `libfabric-1.11.2amzn1.1` (from `libfabric-1.11.1amzn1.0`)   Open MPI: `openmpi40-aws-4.1.1-2` (from `openmpi40-aws-4.1.0`)     Upgraded Slurm to version `20.11.7` (from `20.02.7`).    Install SSM Agent on `centos7` and `centos8`. (SSM Agent is pre-installed in `alinux2`, `ubuntu1804`, and `ubuntu2004`.)   SGE: Always use the shortname as the hostname filter with `qstat`.   Use instance metadata service Version 2 (IMDSv2) instead of instance metadata service Version 1 (IMDSv1) to retrieve instance metadata. For more information, see [Instance metadata and user data](https://docs.amazonaws.cn/AWSEC2/latest/UserGuide/ec2-instance-metadata.html) in the *Amazon EC2 User Guide*.   Upgrade NVIDIA driver to version `460.73.01` (from `450.80.02`).   Upgrade CUDA library to version `11.3.0` (from `11.0`).   Upgrade NVIDIA Fabric Manager to `nvidia-fabricmanager-460`.   Upgrade Python used in Amazon ParallelCluster virtualenvs to `3.7.10` (from `3.6.13`).   Upgrade Cinc Client to `16.13.16`.   Upgrade third-party dependencies of [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/):    `apt-7.4.0` (from `apt-7.3.0`).   `iptables-8.0.0` (from `iptables-7.1.0`).   `line-4.0.1` (from `line-2.9.0`).   `openssh-2.9.1` (from `openssh-2.8.1`).   `pyenv-3.4.2` (from `pyenv-3.1.1`).   `selinux-3.1.1` (from `selinux-2.1.1`).   `ulimit-1.1.1` (from `ulimit-1.0.0`).   `yum-6.1.1` (from `yum-5.1.0`).   `yum-epel-4.1.2` (from `yum-epel-3.3.0`).     <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.11.0/CHANGELOG.md), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.11.0/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.11.0/CHANGELOG.md) packages on GitHub. | July 1, 2021 | 
| [Amazon ParallelCluster version 2.10.4 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.10.4) | Amazon ParallelCluster version 2.10.4 released.<br />Changes include:  Upgraded Slurm to version `20.02.7` (from `20.02.4`).  <br />For more details of the changes, see the CHANGELOG file for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.10.4/CHANGELOG.md) package on GitHub. | May 15, 2021 | 
| [Amazon ParallelCluster version 2.10.3 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.10.3) | Amazon ParallelCluster version 2.10.3 released.<br />Changes include:  Added support for Ubuntu 18.04 and Amazon Linux 2 on Arm-based Amazon Graviton instances in the Amazon China and Amazon GovCloud (US) Amazon Web Services Regions.   [Elastic Fabric Adapter](efa.md) installer updated to 1.11.2:   EFA config: `efa-config-1.7` (no change)   EFA profile: `efa-profile-1.4` (from `efa-profile-1.3`)   EFA Kernel module: `efa-1.10.2` (no change)   RDMA core: `rdma-core-31.2amzn` (no change)   Libfabric: `libfabric-1.11.1amzn1.0` (no change)   Open MPI: `openmpi40-aws-4.1.0` (no change)    <br />For more details of the changes, see the CHANGELOG file for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.10.3/CHANGELOG.md) package on GitHub. | March 18, 2021 | 
| [Amazon ParallelCluster version 2.10.2 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.10.2) | Amazon ParallelCluster version 2.10.2 released.<br />Changes include:  Improve cluster config validation to use the cluster target AMI when invoking the Amazon EC2 [RunInstances](https://docs.amazonaws.cn/AWSEC2/latest/APIReference/API_RunInstances.html) API operation in `--dry-run` mode.   Update the Python version used in the Amazon ParallelCluster virtual environments to 3.6.13.   Fix [`sanity_check`](global.md#sanity-check) for Arm instance types.   Fix `enable_efa` when using `centos8` with the Slurm scheduler or Arm instance types.   Run `apt update` in non-interactive mode (`-y`).   Fix [`encrypted_ephemeral`](cluster-definition.md#encrypted-ephemeral) = true with `alinux2` and `centos8`.  <br />For more details of the changes, see the CHANGELOG file for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.10.2/CHANGELOG.md) package on GitHub. | March 2, 2021 | 
| [Amazon ParallelCluster version 2.10.1 released](https://docs.amazonaws.cn/parallelcluster/latest/ug/document_history.html#aws-parallelcluster-2.10.1) | Amazon ParallelCluster version 2.10.1 released.<br />Changes include:  Added support for the Africa (Cape Town) (`af-south-1`), Europe (Milan) (`me-south-1`), and Middle East (Bahrain) (`me-south-1`) Amazon Web Services Regions. At launch, support is limited in the following ways:   FSx for Lustre and Arm-based Graviton instances aren't supported in any of these Amazon Web Services Regions.   Amazon Batch isn't supported in Africa (Cape Town).   Amazon EBS `io2` and `gp3` volume types aren't supported in the Africa (Cape Town) and Europe (Milan) Amazon Web Services Regions.     Added support for the Amazon EBS `io2` and `gp3` volume types. For more information, see [`[ebs]` section](ebs-section.md) and [`[raid]` section](raid-section.md).   Added support for [Elastic Fabric Adapter](efa.md) on Arm-based Graviton2 instances running `alinux2`, `ubuntu1804`, or `ubuntu2004`. For more information, see [Elastic Fabric Adapter](efa.md).   Install the Arm Performance Libraries 20.2.1 on Arm AMIs (`alinux2`, `centos8`, and `ubuntu1804`). For more information, see [Arm Performance Libraries](arm-performance-libraries.md).   [Intel MPI](intelmpi.md) updated to Version 2019 Update 8 (updated from Version 2019 Update 7). For more information, see [Intel® MPI Library 2019 Update 8](https://software.intel.com/content/www/us/en/develop/articles/intel-mpi-library-release-notes-linux.html#inpage-nav-3-2).   Removed the Amazon CloudFormation `DescribeStacks` API operation call from the Amazon Batch Docker entrypoint to end the job failures caused by throttling by Amazon CloudFormation.   Improved the calls to Amazon EC2 `DescribeInstanceTypes` API operation call when validating a cluster configuration.   Amazon Linux 2 Docker images are pulled from Amazon ECR Public when building the Docker image for the `awsbatch` scheduler.   The default instance type changed from the hardcoded `t2.micro` instance type to the Free Tier instance type for the Amazon Web Services Region (`t2.micro` or `t3.micro`, depending on the Amazon Web Services Region). Amazon Web Services Regions that do not have a Free Tier default to the `t3.micro` instance type.   [Elastic Fabric Adapter](efa.md) installer updated to 1.11.1:   EFA config: `efa-config-1.7` (from `efa-config-1.5`)   EFA profile: `efa-profile-1.3` (from `efa-profile-1.1`)   EFA Kernel module: `efa-1.10.2` (no change)   RDMA core: `rdma-core-31.2amzn` (from `rdma-core-31.amzn0`)   Libfabric: `libfabric-1.11.1amzn1.0` (from `libfabric-1.10.1amzn1.1`)   Open MPI: `openmpi40-aws-4.1.0` (from `openmpi40-aws-4.0.5`)     The [`vpc_settings`](cluster-definition.md#vpc-settings), [`vpc_id`](vpc-section.md#vpc-id), and [`master_subnet_id`](vpc-section.md#master-subnet-id) parameters are now required.   The `nfsd` daemon in the head node is now set to use at least 8 threads. If there are more than 8 cores, it will use as many threads as there are cores. When `ubuntu1604` is used, the setting only changes after the node is rebooted.   [Amazon DCV](dcv.md) updated to Amazon DCV 2020.2-9662. For more information, see [DCV 2020.2-9662— December 04, 2020](https://docs.amazonaws.cn/dcv/latest/adminguide/doc-history-release-notes.html#dcv-2020-2-9662) in the *Amazon DCV Administrator Guide*.   The Intel MPI and HPC packages for Amazon ParallelCluster are pulled from Amazon S3. They are no longer pulled from Intel yum repos.   Changed the default systemd runlevel to `multi-user.target` on all OSs during creation of official Amazon ParallelCluster AMIs. The runlevel is set to `graphical.target` on the head node only when DCV is enabled. This prevents graphical services (such as `x/gdm`) from running when they are not required.   Enabled support for `p4d.24xlarge` instances on the head node.   Increase the maximum number of retry attempts when registering Slurm nodes in Amazon Route 53.  <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.10.1/CHANGELOG.md), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.10.1/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.10.1/CHANGELOG.md) packages on GitHub. | December 22, 2020 | 
| [Amazon ParallelCluster version 2.10.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.10.0) | Amazon ParallelCluster version 2.10.0 released.<br />Changes include:  Added support for CentOS 8 in all Amazon Web Services Regions (outside of the Amazon China and Amazon GovCloud (US) Regions). Removed support for CentOS 6.   Added support for `p4d.24xlarge` instances for compute nodes.   Added support for NVIDIA GPUDirect RDMA on EFA by using the new [`enable_efa_gdr`](cluster-definition.md#enable-efa-gdr) setting.   Added support for Amazon FSx for Lustre features.   Configure your Amazon FSx for Lustre file system to import preferences using the [`auto_import_policy`](fsx-section.md#fsx-auto-import-policy) setting.   Added support for HDD-based Amazon FSx for Lustre file systems using the [`storage_type`](fsx-section.md#fsx-storage-type) and [`drive_cache_type`](fsx-section.md#fsx-drive-cache-type) settings.     Added a Amazon CloudWatch dashboard, including head node metrics and easy access to cluster logs. For more information, see [Amazon CloudWatch dashboard](cloudwatch-dashboard.md).   Added support for using an existing Amazon S3 bucket to store cluster configuration information, using the [`cluster_resource_bucket`](cluster-definition.md#cluster-resource-bucket-section) setting.   Enhanced the [`pcluster createami`](pcluster.createami.md) command.   Added `--post-install` parameter to use a post-installation script when build an AMI.   Added a validation step to fail when using a base AMI created by a different version of Amazon ParallelCluster.   Added a validation step to fail when if the select OS is different from the OS in the base AMI.   Added support for using a Amazon ParallelCluster base AMI.     Enhanced the [`pcluster update`](pcluster.update.md) command.   The [`tags`](cluster-definition.md#tags) setting can now be changed during an update.   Queues can now be resized during an update without stopping the compute fleet     Added `all_or_nothing_batch` configuration parameter for `slurm_resume` script. When `True`, `slurm_resume` will succeed only if all the instances required by all the pending jobs in Slurm will be available. For more information, see [Introducing `all_or_nothing_batch` launches](https://github.com/aws/aws-parallelcluster/wiki/Introducing-all_or_nothing_batch-launches) in the Amazon ParallelCluster Wiki on GitHub.   [Elastic Fabric Adapter](efa.md) installer updated to 1.10.1:   EFA config: `efa-config-1.5` (from `efa-config-1.4`)   EFA profile: `efa-profile-1.1` (from `efa-profile-1.0.0`)   EFA Kernel module: `efa-1.10.2` (from `efa-1.6.0`)   RDMA core: `rdma-core-31.amzn0` (from `rdma-core-28.amzn0`)   Libfabric: `libfabric-1.11.1amzn1.0` (from `libfabric-1.10.1amzn1.1`)   Open MPI: `openmpi40-aws-4.0.5` (from `openmpi40-aws-4.0.3`)     In the Amazon GovCloud (US) Regions, enable support for Amazon DCV and Amazon Batch.   In the Amazon China Regions, enable support for Amazon FSx for Lustre.   Upgrade NVIDIA driver to version 450.80.02 (from 450.51.05).   Install NVIDIA Fabric Manager to enable NVIDIA NVSwitch on supported platforms.   Removed default Amazon Web Services Region of `us-east-1`. The default uses this lookup order.   Amazon Web Services Region specified in `-r` or `--region` argument.   `AWS_DEFAULT_REGION` environment variable.   `aws_region_name` setting in the [`[aws]` section](aws.md) of the Amazon ParallelCluster configuration file (defaults to `~/.parallelcluster/config`).   `region` setting in the `[default]` section of the Amazon CLI configuration file (defaults to `~/aws/config`).    <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.10.0/CHANGELOG.md), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.10.0/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.10.0/CHANGELOG.md) packages on GitHub. | November 18, 2020 | 
| [Amazon ParallelCluster version 2.9.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.9.0) | Amazon ParallelCluster version 2.9.0 released.<br />Changes include:  Added support for multiple queues and multiple instance types in the compute fleet when used with Slurm Workload Manager. When using queues, Auto Scaling groups are no longer used on Slurm. An Amazon Route 53 hosted zone is now created with the cluster and is used for DNS resolution of compute nodes when the Slurm scheduler is used. For more information, see [Multiple queue mode](queue-mode.md).   Added support for [Amazon DCV](dcv.md) on Arm-based Amazon Graviton-based instances.   Added support for disabling hyperthreading on instance types that do not support CPU options in launch templates (for example `*.metal` instance types).   Added support for NFS 4 for file systems shared from the head node.   Removed dependency on [cfn-init](https://docs.amazonaws.cn/AWSCloudFormation/latest/UserGuide/cfn-init.html) when bootstrapping compute nodes to avoid throttling by Amazon CloudFormation when a large number of nodes join the cluster.   [Elastic Fabric Adapter](efa.md) installer updated to 1.9.5:   EFA config: `efa-config-1.4` (from `efa-config-1.3`)   EFA profile: `efa-profile-1.0.0` (new)   Kernel module: `efa-1.6.0` (no change)   RDMA core: `rdma-core-28.amzn0` (no change)   Libfabric: `libfabric-1.10.1amzn1.1` (no change)   Open MPI: `openmpi40-aws-4.0.3` (no change)     Upgraded Slurm to version `20.02.4` (from `19.05.5`).   [Amazon DCV](dcv.md) updated to Amazon DCV 2020.1-9012. For more information, see [DCV 2020.1-9012— August 24, 2020 Release Notes](https://docs.amazonaws.cn/dcv/latest/adminguide/doc-history-release-notes.html#dcv-2020-1-9012) in the *Amazon DCV Administrator Guide*.   When mounting shared NFS drives, use the head node private IP address instead of the hostname.   Added new log streams to CloudWatch Logs: `chef-client`, `clustermgtd`, `computemgtd`, `slurm_resume`, and `slurm_suspend`.   Added support for queue names in pre-install and post-install scripts.   In the Amazon GovCloud (US) Amazon Web Services Regions, use the Amazon DynamoDB on-demand billing option. For more information, see [On-Demand Mode](https://docs.amazonaws.cn/amazondynamodb/latest/developerguide/HowItWorks.ReadWriteCapacityMode.html#HowItWorks.OnDemand) in the *Amazon DynamoDB Developer Guide*.  <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.9.0/CHANGELOG.rst), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.9.0/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.9.0/CHANGELOG.md) packages on GitHub. | September 11, 2020 | 
| [Amazon ParallelCluster version 2.8.1 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.8.1) | Amazon ParallelCluster version 2.8.1 released.<br />Changes include:  Disable screen lock for Amazon DCV sessions to prevent users from being locked out.   Fix [`pcluster configure`](pcluster.configure.md) when including an Arm-based Amazon Graviton-based instance type.  <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.8.1/CHANGELOG.rst), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.8.1/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.8.1/CHANGELOG.md) packages on GitHub. | August 4, 2020 | 
| [Amazon ParallelCluster version 2.8.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.8.0) | Amazon ParallelCluster version 2.8.0 released.<br />Changes include:  Added support for Arm-based Amazon Graviton-based instances (like the `A1` and `C6g`).   Added support for the automatic daily backup features of Amazon FSx for Lustre. For more information, see [`automatic_backup_retention_days`](fsx-section.md#fsx-automatic-backup-retention-days), [`copy_tags_to_backups`](fsx-section.md#fsx-copy-tags-to-backups), [`daily_automatic_backup_start_time`](fsx-section.md#fsx-daily-automatic-backup-start-time), and [`fsx_backup_id`](fsx-section.md#fsx-backup-id).   Removed dependency on Berkshelf from [`pcluster createami`](pcluster.createami.md).   Improved the robustness and user experience of [`pcluster update`](pcluster.update.md). For more information, see [Using `pcluster update`](using-pcluster-update.md).   [Elastic Fabric Adapter](efa.md) installer updated to 1.9.4:   Kernel module: `efa-1.6.0` (updated from `efa-1.5.1`)   RDMA core: `rdma-core-28.amzn0` (updated from `rdma-core-25.0`)   Libfabric: `libfabric-1.10.1amzn1.1` (updated from `libfabric-aws-1.9.0amzn1.1`)   Open MPI: `openmpi40-aws-4.0.3` (no change)     Upgrade NVIDIA driver to Tesla version 440.95.01 on CentOS 6 and version 450.51.05 on all other distributions.   Upgrade CUDA library to version 11.0 on all distributions other than CentOS 6.  <br />For more details of the changes, see the CHANGELOG files for the [aws-parallelcluster](https://github.com/aws/aws-parallelcluster/blob/v2.8.0/CHANGELOG.rst), [aws-parallelcluster-cookbook](https://github.com/aws/aws-parallelcluster-cookbook/blob/v2.8.0/CHANGELOG.md), and [aws-parallelcluster-node](https://github.com/aws/aws-parallelcluster-node/blob/v2.8.0/CHANGELOG.md) packages on GitHub. | July 23, 2020 | 
| [Amazon ParallelCluster version 2.7.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.7.0) | Amazon ParallelCluster version 2.7.0 released.<br />Changes include:  [`base_os`](cluster-definition.md#base-os) is now a required parameter.   [`scheduler`](cluster-definition.md#scheduler) is now a required parameter.   [Amazon DCV](dcv.md) updated to Amazon DCV 2020.0. For more information, see [Amazon DCV releases version 2020.0 with surround sound 7.1 and stylus support](https://aws.amazon.com/about-aws/whats-new/2020/04/nice-dcv-releases-version-2020-0-with-surround-sound-7-1-and-stylus-support/). <br />[Intel MPI](intelmpi.md) updated to Version 2019 Update 7 (updated from Version 2019 Update 6). For more information, see [Intel® MPI Library 2019 Update 7](https://software.intel.com/content/www/us/en/develop/articles/intel-mpi-library-release-notes-linux.html#inpage-nav-3-undefined). <br />[Elastic Fabric Adapter](efa.md) installer updated to 1.8.4:   Kernel module: `efa-1.5.1` (no change)   RDMA core: `rdma-core-25.0` (no change)   Libfabric: `libfabric-aws-1.9.0amzn1.1` (no change)   Open MPI: `openmpi40-aws-4.0.3` (updated from `openmpi40-aws-4.0.2`)     Upgrade CentOS 7 AMI to version 7.8-2003 (updated from 7.7-1908). For more information, see [CentOS-7 (2003) Release Notes](https://wiki.centos.org/Manuals(2f)ReleaseNotes(2f)CentOS7(2e)2003.html).   | May 19, 2020 | 
| [Amazon ParallelCluster version 2.6.1 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.6.1) | Amazon ParallelCluster version 2.6.1 released.<br />Changes include:  Removed `cfn-init-cmd` and `cfn-wire` from logs stored in Amazon CloudWatch Logs. For more information, see [Integration with Amazon CloudWatch Logs](cloudwatch-logs.md).   | April 17, 2020 | 
| [Amazon ParallelCluster version 2.6.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.6.0) | Amazon ParallelCluster version 2.6.0 released.<br />Changes include:  Added support for Amazon Linux 2.   Now Amazon CloudWatch Logs is used to collect cluster and scheduler logs. For more information, see [Integration with Amazon CloudWatch Logs](cloudwatch-logs.md).   Added support for new Amazon FSx for Lustre deployment types `SCRATCH_2` and `PERSISTENT_1`. Support for FSx for Lustre on Ubuntu 18.04 and Ubuntu 16.04. For more information, see [fsx](fsx-section.md).   Added support for Amazon DCV on Ubuntu 18.04. For more information, see [Connect to the head node through Amazon DCV](dcv.md).   | February 27, 2020 | 
| [Amazon ParallelCluster version 2.5.1 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.5.1) | Amazon ParallelCluster version 2.5.1 released. | December 13, 2019 | 
| [Amazon ParallelCluster version 2.5.0 released](https://github.com/aws/aws-parallelcluster/releases/tag/v2.5.0) | Amazon ParallelCluster version 2.5.0 released. | November 18, 2019 | 
| [Amazon ParallelCluster introduces support for Intel MPI](https://docs.amazonaws.cn/parallelcluster/latest/ug/intelmpi.html) | Amazon ParallelCluster version 2.4.1 introduces support for Intel MPI. | July 29, 2019 | 
| [Amazon ParallelCluster introduces support for EFA](https://docs.amazonaws.cn/parallelcluster/latest/ug/efa.html) | Amazon ParallelCluster version 2.4.0 introduces support for Elastic Fabric Adapter (EFA). | June 11, 2019 | 
| [Amazon ParallelCluster documentation released on Amazon documentation site](#document_history) | The Amazon ParallelCluster documentation is now available in 10 languages and in both HTML and PDF formats. | May 24, 2018 |