fabricmanager Enable

1. fabricmanager Enable in CHROOT 2. pythorch cuda devices test
dasandata · Dec 20, 2023 · f04b030 · f04b030
1 parent 076237e
commit f04b030
Show file tree

Hide file tree

Showing 2 changed files with 19 additions and 1 deletion.
diff --git a/Provisioning/OpenHPC Cluster Building (v2.6-Rocky8 Base OS).md b/Provisioning/OpenHPC Cluster Building (v2.6-Rocky8 Base OS).md
@@ -1329,7 +1329,7 @@ mount | grep ${CHROOT}
 
 # Install gcc, make to VNFS
 yum -y install --installroot ${CHROOT} gcc make \
->> dasan_log_ohpc_nvidia-driver-latest-vnfs.txt 2>&1
+>> dasan_log_ohpc_nvidia-driver-latest-vnfs.txt 2>&1s
 tail dasan_log_ohpc_nvidia-driver-latest-vnfs.txt
 
 # Install nvidia-driver node VNFS
@@ -1353,6 +1353,7 @@ chroot  ${CHROOT}  systemctl enable nvidia-persistenced
 ```bash
 yum -y install --installroot ${CHROOT}   nvidia-fabric-manager cuda-drivers-fabricmanager nvidia-fabric-manager-devel   
 
+chroot $CHROOT systemctl enable nvidia-fabricmanager
 ```
 
 ## ## gpustat (python3) install to VNFS
@@ -1508,6 +1509,14 @@ squeue
 ```
 
 
+```bash
+
+conda activate pythorch
+
+python -c "import torch; print(torch.cuda.is_available())"
+
+```
+
 
 ***
 

diff --git a/Provisioning/OpenHPC Cluster Building (v3.0-Rocky9 Base OS).md b/Provisioning/OpenHPC Cluster Building (v3.0-Rocky9 Base OS).md
@@ -1387,6 +1387,7 @@ chroot  ${CHROOT}  systemctl enable nvidia-persistenced
 ```bash
 yum -y install --installroot ${CHROOT}   nvidia-fabric-manager cuda-drivers-fabricmanager nvidia-fabric-manager-devel   
 
+chroot $CHROOT systemctl enable nvidia-fabricmanager
 ```
 
 ## ## gpustat (python3) install to VNFS
@@ -1537,6 +1538,14 @@ squeue
 ```
 
 
+```bash
+
+conda activate pythorch
+
+python -c "import torch; print(torch.cuda.is_available())"
+
+```
+
 
 ***