AWS AWSEC2 documentation change
Summary
Removed Amazon Linux 2 support and references, updated installation steps for NCCL setup, and adjusted documentation links
Security assessment
The changes involve removing deprecated OS support and updating installation procedures without any mention of security vulnerabilities, patches, or security features. This appears to be routine maintenance of supported platforms and installation steps.
Diff
diff --git a/AWSEC2/latest/UserGuide/efa-start-nccl.md b/AWSEC2/latest/UserGuide/efa-start-nccl.md index b764c81b9..2c6f461a4 100644 --- a//AWSEC2/latest/UserGuide/efa-start-nccl.md +++ b//AWSEC2/latest/UserGuide/efa-start-nccl.md @@ -17 +17 @@ The NVIDIA Collective Communications Library (NCCL) is a library of standard col - * Only Amazon Linux 2023, Amazon Linux 2, Ubuntu 24.04, and Ubuntu 22.04 base AMIs are supported. + * Only Amazon Linux 2023, Ubuntu 24.04, and Ubuntu 22.04 base AMIs are supported. @@ -161 +161 @@ You must provision an additional 10 to 20 GiB of storage for the Nvidia CUDA Too -Amazon Linux 2023 and Amazon Linux 2 +Amazon Linux 2023 @@ -201 +201 @@ After the instance has rebooted, reconnect to it. - 1. Enable the EPEL repository and set the distribution to `rhel7`. + 1. Set the distribution variable and set up the CUDA network repository. @@ -203,6 +203,2 @@ After the instance has rebooted, reconnect to it. - $ sudo amazon-linux-extras install epel \ - && distribution='rhel7' - - 2. Set up the CUDA network repository and update the repository cache. - - $ ARCH=$( /bin/arch ) \ + $ distribution=$(. /etc/os-release;echo $ID$VERSION_ID) \ + && ARCH=$( /bin/arch ) \ @@ -212,11 +207,0 @@ After the instance has rebooted, reconnect to it. - 3. (_Kernel version 5.10 only_) Perform these steps only if you are using Amazon Linux 2 with kernel version 5.10. If you are using Amazon Linux 2 with kernel version 4.12, skip these steps. To check your kernel version, run **uname -r**. - - 1. Create the Nvidia driver configuration file named `/etc/dkms/nvidia.conf`. - - $ sudo mkdir -p /etc/dkms \ - && echo "MAKE[0]=\"'make' -j2 module SYSSRC=\${kernel_source_dir} IGNORE_XEN_PRESENCE=1 IGNORE_PREEMPT_RT_PRESENCE=1 IGNORE_CC_MISMATCH=1 CC=/usr/bin/gcc10-gcc\"" | sudo tee /etc/dkms/nvidia.conf - - 2. (`p4d.24xlarge` and `p5.48xlarge` only) Copy the Nvidia driver configuration file. - - $ sudo cp /etc/dkms/nvidia.conf /etc/dkms/nvidia-open.conf - @@ -376 +361 @@ Install GDRCopy to improve the performance of Libfabric. For more information ab -Amazon Linux 2023 and Amazon Linux 2 +Amazon Linux 2023 @@ -607 +592 @@ Install the NCCL tests. The NCCL tests enable you to confirm that NCCL is proper - * Amazon Linux 2023 and Amazon Linux 2 + * Amazon Linux 2023 @@ -682 +667 @@ The following additional information is displayed when using a `p4d.24xlarge` in -Install the machine learning applications on the temporary instance. The installation procedure varies depending on the specific machine learning application. For more information about installing software on your Linux instance, see [Manage software on your Amazon Linux 2 instance](https://docs.aws.amazon.com/linux/al2/ug/managing-software.html). +Install the machine learning applications on the temporary instance. The installation procedure varies depending on the specific machine learning application. For more information about installing software on your Linux instance, see [Manage OS updates](https://docs.aws.amazon.com/linux/al2023/ug/managing-repos-os-updates.html) in the _Amazon Linux 2023 User Guide_.