Update docs

ComputeCanada · Jan 15, 2025 · ee8df9a · ee8df9a
1 parent c9016e1
commit ee8df9a
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 3 deletions.
diff --git a/docs/README.md b/docs/README.md
@@ -548,6 +548,13 @@ the operating system and service software
     and with x86-64 processors (see [NVIDIA/mig-parted issue #30](https://github.com/NVIDIA/mig-parted/issues/30)).
 6. `shard`: total number of [Sharding](https://slurm.schedmd.com/gres.html#Sharding) on the node. Sharding allows sharing the same GPU on multiple jobs. The total number of shards is evenly distributed across all GPUs on the node.
 
+The instance specifications are retrieved from the cloud provider data source, but it is possible to explicitly specify them.
+
+7. `cpus`: number of logical processors on the node - [`CPUs` in slurm.conf](https://slurm.schedmd.com/slurm.conf.html#OPT_CPUs)
+8. `ram`: size of real memory on the node in megabyte - [`RealMemory` in slurm.conf](https://slurm.schedmd.com/slurm.conf.html#OPT_RealMemory)
+9. `gpus`: number of graphical processor on the node - [`Gres=gpu:<gpus>` in slurm.conf](https://slurm.schedmd.com/slurm.conf.html#OPT_Gres_1)
+10. `gpu_type`: type of graphical processor on the node - [`Gres=gpu:<gpu_type>:<gpus>` in slurm.conf](https://slurm.schedmd.com/slurm.conf.html#OPT_Gres_1)
+
 For some cloud providers, it possible to define additional attributes.
 The following sections present the available attributes per provider.
 
@@ -585,7 +592,7 @@ For more information on these attributes, refer to
 - `gpu_type`: name of the GPU model to attach to the instance. Refer to
 [Google Cloud documentation](https://cloud.google.com/compute/docs/gpus) for the list of
 available models per region
-- `gpu_count`: number of GPUs of the `gpu_type` model to attach to the instance
+- `gpus`: number of GPUs of the `gpu_type` model to attach to the instance
 
 #### 4.7.3 Post build modification effect
 

diff --git a/examples/gcp/main.tf b/examples/gcp/main.tf
@@ -24,8 +24,8 @@ module "gcp" {
       type = "n1-standard-2",
       tags = ["node"],
       count = 1,
+      gpus = 1
       gpu_type = "nvidia-tesla-t4",
-      gpu_count = 1
     }
   }
 

diff --git a/gcp/infrastructure.tf b/gcp/infrastructure.tf
@@ -168,7 +168,7 @@ locals {
       specs = merge({
         cpus   = data.external.machine_type[values["prefix"]].result["vcpus"]
         ram    = data.external.machine_type[values["prefix"]].result["ram"]
-        gpus   = try(data.external.machine_type[values["prefix"]].result["gpus"], lookup(values, "gpu_count", 0))
+        gpus   = try(data.external.machine_type[values["prefix"]].result["gpus"], 0)
       }, values.specs)
       volumes = contains(keys(module.design.volume_per_instance), x) ? {
         for pv_key, pv_values in var.volumes: