feat: 增加 AI 仪表盘获取集群数据汇总的接口 GetSummaryClusterInfo

.changeset/fast-emus-breathe.md

@@ -0,0 +1,5 @@
+---
+"@scow/scow-scheduler-adapter-interface": minor
+---
+
+增加 AI 仪表盘获取集群数据汇总的接口 GetSummaryClusterInfo

protos/config.proto

@@ -124,6 +124,54 @@ message GetClusterInfoResponse {
   uint32 pending_job_count = 17;
 }
 
+message SummaryPartitionInfo {
+  enum PartitionStatus {
+    UNKNOWN = 0;
+    NOT_AVAILABLE = 1;
+    AVAILABLE = 2;
+  }
+
+  string partition_name = 1;
+  uint32 node_count = 2;
+  float node_usage = 3;
+  uint32 cpu_core_count =4;
+  float cpu_usage = 5;
+  uint32 gpu_core_count = 6;
+  float gpu_usage = 7;
+  uint32 pending_job_count = 8;
+  PartitionStatus partition_status = 9;
+}
+
+message GetSummaryClusterInfoRequest {
+  // 关联账户 用于筛选已授权集群分区
+  repeated string account_names = 1;
+}
+
+message GetSummaryClusterInfoResponse {
+  string cluster_name = 1;
+  repeated SummaryPartitionInfo partitions = 2;
+  // 3-17 Newly added parameters
+  // only returns in version later than 1.7.0
+  uint32 node_count = 3;
+  uint32 running_node_count = 4;
+  uint32 idle_node_count = 5;
+  uint32 not_available_node_count = 6;
+  uint32 cpu_core_count =7;
+  uint32 running_cpu_count = 8;
+  uint32 idle_cpu_count = 9;
+  uint32 not_available_cpu_count = 10;
+  uint32 gpu_core_count = 11;
+  uint32 running_gpu_count = 12;
+  uint32 idle_gpu_count = 13;
+  uint32 not_available_gpu_count = 14;
+  uint32 running_job_count = 15;
+  uint32 pending_job_count = 16;
+
+  float node_usage = 17;
+  float cpu_usage = 18;
+  float gpu_usage = 19;
+}
+
 message NodeInfo {
 
   enum NodeState {
@@ -182,6 +230,10 @@ service ConfigService {
    * description: get cluster information
    */
   rpc GetClusterInfo(GetClusterInfoRequest) returns (GetClusterInfoResponse);
+   /*
+   * description: get calculated cluster information
+   */
+  rpc GetSummaryClusterInfo(GetSummaryClusterInfoRequest) returns (GetSummaryClusterInfoResponse);
    /*
    * description: get cluster nodes information
    */