支持 Vnpu 的监控

[sundysundy]

npu-smi info -t info-vnpu -i 3 -c 0
+-------------------------------------------------------------------------------+
| NPU resource static info as follow: |
| Format:Free/Total NA: Currently, query is not supported. |
| AICORE Memory AICPU VPC VENC VDEC JPEGD JPEGE PNGD |
| GB |
|===============================================================================|
| 0/8 0/21 0/7 0/12 0/3 0/12 0/16 0/8 NA/NA |
+-------------------------------------------------------------------------------+
| Total number of vnpu: 4 |
+-------------------------------------------------------------------------------+
| Vnpu ID | Vgroup ID | Container ID | Status | Template Name |
+-------------------------------------------------------------------------------+
| 132 | 0 | 7de0fc6bb782 | 1 | vir02 |
+-------------------------------------------------------------------------------+
| 133 | 1 | 195b64ed4795 | 1 | vir02 |
+-------------------------------------------------------------------------------+
| 134 | 2 | b4ed6d3d83ef | 1 | vir02 |
+-------------------------------------------------------------------------------+
| 135 | 3 | bbaa419788e0 | 1 | vir02_1c |
+-------------------------------------------------------------------------------+一

当一张卡做好了虚拟NPU 用docker 启动，如何进行相关的监控 如果监控单张卡，会报

[ERROR] 2024/01/15 13:26:18.300656 219 collector/npu_collector.go:1116 get device resource info failed, cardID(2) and deviceID(0) , error code: -8255
[ERROR] 2024/01/15 13:26:18.721731 219 collector/npu_collector.go:1116 get device resource info failed, cardID(3) and deviceID(0) , error code: -8255
[ERROR] 2024/01/15 13:26:19.149178 219 collector/npu_collector.go:1116 get device resource info failed, cardID(4) and deviceID(0) , error code: -8255
[ERROR] 2024/01/15 13:26:19.579369 219 collector/npu_collector.go:1116 get device resource info failed, cardID(5) and deviceID(0) , error code: -8255 相关的错误