Skip to content

Commit 5ac818d

Browse files
committed
Flesh out docs a bit more
1 parent fe90f64 commit 5ac818d

File tree

1 file changed

+25
-0
lines changed

1 file changed

+25
-0
lines changed

docs/mig.md

Lines changed: 25 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,22 @@
11
# vGPU/MIG configuration
22

3+
This page details how to configure Multi Instance GPU (MIG) in Slurm.
4+
5+
## Pre-requisites
6+
7+
- Image built with cuda support. This should automatically recompile slurm against NVML.
8+
9+
## Inventory
10+
11+
Add relevant hosts to the ``vgpu`` group, for example in ```environments/<environment>/inventory/groups``:
12+
13+
```
14+
[vgpu:children]
15+
cuda
16+
```
17+
18+
## Configuration
19+
320
Use variables from the [stackhpc.linux.vgpu](https://github.com/stackhpc/ansible-collection-linux/tree/main/roles/vgpu) role.
421

522
For example in: `environments/<environment>/inventory/group_vars/all/vgpu`:
@@ -15,3 +32,11 @@ vgpu_definitions:
1532

1633
The appliance will use the driver installed via the ``cuda`` role. Use ``lspci`` to determine the PCI
1734
addresses.
35+
36+
## compute_init
37+
38+
Use the ``enable_vgpu`` metadata option to enable creation of mig devices on rebuild.
39+
40+
## gres configuration
41+
42+
TODO

0 commit comments

Comments
 (0)