他的回复:
给虚拟机新增扩展关联字段originalOwnerType（对应回复的第3步）这里如果使用https://160.101.2.27:26335/rest/cmdb/v1/classes/CLOUD_VM/attrs这个url，返回"exceptionId": "sys.res.detail.reason.resource_type_not_found"，"exceptionType": "ROA_EXFRAME_EXCEPTION"现在使用https://160.101.2.27:26335/rest/cmdb/v1/classes/CLOUD_VM_NOVA/attrs这个url, PUT，报文为：{ "name": "originalOwnerType", "writeable": true, "type": "string", "length": 128, "displayName": { "en_US": "originalOwnerType", "zh_CN": "来源系统类型" } }返回200。这步完成。但不知CLOUD_VM_NOVA和CLOUD_VM是什么关系，这里的属性注册只能使用CLOUD_VM_NOVA。2.驱动重新注册了，原来驱动DriverZ03在接入管理界面停止并删除了。3.上报弹性云主机，因上面使用的className是CLOUD_VM_NOVA，这里使用url为https://160.101.2.27:26335/rest/cmdb/v1/instances/audit/CLOUD_VM_NOVA（如果url为https://160.101.2.27:26335/rest/cmdb/v1/instances/audit/CLOUD_VM也能上报成功）PUT报文为：{ "reportState": "NORMAL", "errorMessage": "success", "supportResClass": [ "CLOUD_VM_NOVA" ], "taskId": "464f8f7c-d8f8-4c78-9ea6-1913f16727ef", "totalNum": 1, "datas": [ { "status": "active", "azoneId": "4B22CC4D492C3159B26015C6B447BA94", "azoneName": "松山湖", "osType": "CRULERos", "osVersion": "vz01", "originalOwnerType": "DriverZ04", "name": "cvmzanz44", "projectId": "3d5ec7c28a8d488cb85b4db3fa3eed5d", "regionId": "gd-tbw-1", "nativeId": "anzzlm0aVM0044", "hypervisorType": "KVM", "flavorId": "8776AA3F93593999A48DC545CCB03E30", "floatingIp": "192.168.171.62" } ]}返回2004.注册性能模型url 为https://160.101.2.27:26335/rest/performance/v2/classes/CLOUD_VM 和 https://160.101.2.27:26335/rest/performance/v2/classes/CLOUD_VM_NOVA都能注册上，但 CLOUD_VM_NOVA注册上后，下面的第6步看不到监控任务列表，看来在监控任务端，CLOUD_VM这个url注册的指标生效了。{ "className": "CLOUD_VM", "namespace": "CLOUD_VM", "system": { "systemType": "DriverZ04", "enUs": "DriverZ04", "zhCn": "DriverZ04" }, "indicators": [{ "name": "cpuUsage", "dataType": "float", "dataUnit": "%", "displayName": { "enUs": "CPU Usage", "zhCn": "CPU使用率" } }, { "name": "memoryUsage", "dataType": "float", "dataUnit": "%", "displayName": { "enUs": "Memory Usage", "zhCn": "内存使用率" } } ], "extraCondition":{ "name":"originalOwnerType", "value":["DriverZ04"] } }这里的 namespace 都用的 CLOUD_VM跟后面的性能指标上报的报文保持一致。5. 查看虚拟机列表，右边有“来源系统类型”，性能详情链接也激活了（对应回复的第2步）。这里不知为何规格和可用分区界面显示是空的。6.查看监控任务列表，正常数有数字（对应回复的第1步）。暂停10秒以上，重新启动过。5.上报性能数据：https://160.101.2.27:26335//V1.0/3d5ec7c28a8d488cb85b4db3fa3eed5d/metric-dataPOST报文：[ { "metric": { "namespace": "CLOUD_VM", "dimensions": [ { "name": "nativeId", "value": "anzzlm0aVM0044" } ], "metric_name": "cpuUsage" }, "value": 0.34, "collect_time": 1600149356614, "ttl": 172800, "type": "float", "unit": "%" }, { "metric": { "namespace": "CLOUD_VM", "dimensions": [ { "name": "nativeId", "value": "anzzlm0aVM0044" } ], "metric_name": "memoryUsage" }, "value": 0.31, "collect_time": 1600149356614, "ttl": 172800, "type": "float", "unit": "%" }]这里的资源id，使用的是 nativeId，不知道是不是ManageOne关联不到资源？实在不知道什么环节出问题了。我看道大神的帖子： https://bbs.huaweicloud.com/forum/thread-72003-1-1.html《运维南向驱动接入FAQ-持续更新》中有MO运维侧南向接入的实例，但没更新性能上报的例子，能不能把性能上报的例子写一个，或者在这贴一个例子？
