From 09be67c989baa0fb2c327fd0d4620c8563a76162 Mon Sep 17 00:00:00 2001 From: Rodrigo Arias Mallo Date: Tue, 2 Sep 2025 17:21:37 +0200 Subject: [PATCH] Update fox documentation for SLURM and FS Reviewed-by: Aleix Roca Nonell --- content/fox/_index.md | 27 +++++++++++++++++---------- 1 file changed, 17 insertions(+), 10 deletions(-) diff --git a/content/fox/_index.md b/content/fox/_index.md index 554794f..e713c2c 100644 --- a/content/fox/_index.md +++ b/content/fox/_index.md @@ -21,17 +21,28 @@ the detailed specifications: ## Access -To access the machine, request a SLURM session from [hut](/hut) using the `fox` -partition: +To access the machine, request a SLURM session from [apex](/apex) using the `fox` +partition. If you need the machine for performance measurements, use an +exclusive reservation: - hut% salloc -p fox + apex% salloc -p fox --exclusive -Then connect via ssh: +Otherwise, specify the CPUs that you need so other users can also use the node +at the same time: - hut% ssh fox + apex% salloc -p fox -c 8 + +Then use srun to execute an interactive shell: + + apex% srun --pty $SHELL fox% -Follow [these steps](/access) if you don't have access to hut or fox. +Make sure you get all CPUs you expect: + + fox% grep Cpus_allowed_list /proc/self/status + Cpus_allowed_list: 0-191 + +Follow [these steps](/access) if you don't have access to apex or fox. ## CUDA @@ -89,9 +100,5 @@ Then just run `nix develop` from the same directory: The machine has several file systems available. -- `$HOME`: Mounted via NFS across all nodes. It is slow and has low capacity. - Don't abuse. -- `/ceph/home/$USER`: Shared Ceph file system across jungle nodes. Slow but high - capacity. Stores three redundant copies of every file. - `/nvme{0,1}/$USER`: The two local NVME disks, very fast and large capacity. - `/tmp`: tmpfs, fast but not backed by a disk. Will be erased on reboot.