Added cluster_info

lib/ood_core/job/adapter.rb

@@ -33,6 +33,14 @@ def submit(script, after: [], afterok: [], afternotok: [], afterany: [])
         raise NotImplementedError, "subclass did not define #submit"
       end
 
+      # Retrieve the number of active and total cpus, nodes, and gpu nodes
+      # @abstract Subclass is expected to implement {#cluster_stats}
+      # @raise [NotImplementedError] if subclass did not define {#cluster_stats}
+      # @return [Hash] Quantified statistics about the cluster's active/total cpus, nodes, and gpu nodes
+      def cluster_stats
+        raise NotImplementedError, "subclass did not define #cluster_stats"
+      end
+
       # Retrieve info for all jobs from the resource manager
       # @abstract Subclass is expected to implement {#info_all}
       # @raise [NotImplementedError] if subclass did not define {#info_all}

lib/ood_core/job/adapters/slurm.rb

@@ -98,6 +98,24 @@ def initialize(cluster: nil, bin: nil, conf: nil, bin_overrides: {}, submit_host
             @strict_host_checking = strict_host_checking
           end
 
+          # Get a hash describing the number of nodes_active, nodes_total, processors_total, and processors_active
+          # @return [Hash] details about the cluster's active and total nodes, processors, and gpu nodes
+          def get_cluster_stats()
+            sinfo_out = call("sinfo", ["-a", "-h", "-o %A/%D/%C"]).strip.split('/')
+            gres_length = call("sinfo", ["-o %G"]).lines.map(&:strip).map(&:length).max
+            sinfo_out2 = call("sinfo", ["-N", "-h", "-a", "--Format='nodehost,gres:3#{gres_length},statelong'"])
+            gpu_nodes_total = sinfo_out2.lines.uniq.grep(/gpu:/).count
+            gpu_nodes_free = sinfo_out2.lines.uniq.grep(/gpu:/).grep(/idle/).count
+            {
+              "nodes_active": sinfo_out[0].to_i,
+              "nodes_total": sinfo_out[2].to_i,
+              "processors_total": sinfo_out[6].to_i,
+              "processors_active": sinfo_out[3].to_i,
+              "gpu_nodes_total": gpu_nodes_total,
+              "gpu_nodes_active": gpu_nodes_total - gpu_nodes_free
+            }
+          end
+
           # Get a list of hashes detailing each of the jobs on the batch server
           # @example Status info for all jobs
           #   my_batch.get_jobs
@@ -454,6 +472,12 @@ def submit(script, after: [], afterok: [], afternotok: [], afterany: [])
           raise JobAdapterError, e.message
         end
 
+        # Retrieve info about active and total cpus, gpus, and nodes
+        # @return [Hash] information about cluster usage
+        def cluster_stats()
+          @slurm.get_cluster_stats()
+        end
+
         # Retrieve info for all jobs from the resource manager
         # @raise [JobAdapterError] if something goes wrong getting job info
         # @return [Array<Info>] information describing submitted jobs