你当前正在访问 Microsoft Azure Global Edition 技术文档网站。如果需要访问由世纪互联运营的 Microsoft Azure 中国技术文档网站，请访问 https://docs.azure.cn。

InsightsMetrics 表的查询

项目
11/20/2024

有关在 Azure 门户中使用这些查询的信息，请参阅 Log Analytics 教程。有关 REST API，请参阅查询。

IoT Edge：设备脱机或未按预期速率向上游发送消息

识别在过去 2 天内看到的 IoT Edge 设备，这些设备未在 30 分钟内以预期速率将 D2C 消息发送到 IoT 中心。

// To create an alert for this query, click '+ New alert rule'
let targetReceiver = "upstream";
InsightsMetrics
| where Origin == "iot.azm.ms" and Namespace == "metricsmodule"
| where Name == "edgehub_messages_sent_total"
| extend dimensions=parse_json(Tags)
| extend device = tostring(dimensions.edge_device)
| extend target = trim_start(@"[^/]+/", extractjson("$.to", 
tostring(dimensions), typeof(string)))
| where target contains targetReceiver
| extend source = strcat(device, "::", trim_start(@"[^/]+/", 
tostring(dimensions.from)))
| extend messages = toint(Val)
| extend timeUtc = TimeGenerated
| extend sourceTarget = strcat(source, "::", target)
| project timeUtc, source, sourceTarget, messages, device, _ResourceId
| order by device, sourceTarget, timeUtc
| serialize
| extend nextCount = next(messages, 1)
| extend nextSourceTarget= next(sourceTarget, 1)
| extend diff = iff((messages - nextCount) >= 0, messages - nextCount, 0)
| where sourceTarget == nextSourceTarget and diff >= 0
| project TimeGenerated = timeUtc, source, sourceTarget, messages, diff, 
device, _ResourceId
| make-series sum(diff) default=0 on TimeGenerated from ago(2d) to now() 
step 30m by device, _ResourceId
| mv-expand sum_diff, TimeGenerated
| project TimeGenerated=todatetime(TimeGenerated), device, 
AggregatedValue=toint(sum_diff), _ResourceId

IoT Edge：超出阈值的边缘中心队列大小

在评估期间，设备的 Edge 中心队列大小（sum）超过配置的阈值的次数。

// To create an alert for this query, click '+ New alert' 
let qlenThreshold = 100;
InsightsMetrics
| where Origin == "iot.azm.ms" and Namespace == "metricsmodule"
| where Name == "edgehub_queue_length"
| extend dimensions=parse_json(Tags)
| extend device = tostring(dimensions.edge_device)
| extend ep = tostring(dimensions.endpoint)
| extend qlen = toint(Val)
| project device, qlen, ep, TimeGenerated, _ResourceId
| summarize sum(qlen) by TimeGenerated, device, _ResourceId
| where sum_qlen >= qlenThreshold
| project-away sum_qlen

最大节点磁盘

平均超过 30 分钟的最大节点磁盘使用量。

// To create an alert for this query, click '+ New alert rule'
//InsightMetrics contains all the custom metrics for Container Insights solution
InsightsMetrics // Replace Name with your custom metric
| where Name == "used_percent" and Namespace == "container.azm.ms/disk" 
| summarize val= max(Val) by bin(TimeGenerated, 15m), _ResourceId
| render timechart

每个节点每秒读取 Prometheus 磁盘数

以时间表的形式查看默认 kubernetes 命名空间中的 Prometheus 磁盘读取指标。

// To create an alert for this query, click '+ New alert rule'
// Update TimeGenerated field for custom time range
InsightsMetrics
| where Namespace == 'container.azm.ms/diskio'
| where TimeGenerated > ago(1h)
| where Name == 'reads'
| extend Tags = todynamic(Tags)
| extend HostName = tostring(Tags.hostName), Device = Tags.name
| extend NodeDisk = strcat(Device, "/", HostName)
| order by NodeDisk asc, TimeGenerated asc
| serialize //calculating the PreVal, PrevTimeGenerated to render the chart.
| extend PrevVal = iif(prev(NodeDisk) != NodeDisk, 0.0, prev(Val)), PrevTimeGenerated = iif(prev(NodeDisk) != NodeDisk, datetime(null), prev(TimeGenerated))
| where isnotnull(PrevTimeGenerated) and PrevTimeGenerated != TimeGenerated
//Calculating the rate for disk using PreVal
| extend Rate = iif(PrevVal > Val, Val / (datetime_diff('Second', TimeGenerated, PrevTimeGenerated) * 1), iif(PrevVal == Val, 0.0, (Val - PrevVal) / (datetime_diff('Second', TimeGenerated, PrevTimeGenerated) * 1)))
| where isnotnull(Rate)
| project TimeGenerated, NodeDisk, Rate, _ResourceId
| render timechart

在 InsightsMetrics 中查找

在 InsightsMetrics 中查找以在 InsightsMetrics 表中搜索特定值。/n请注意，此查询需要更新 <SeachValue> 参数，才能生成结果

// This query requires a parameter to run. Enter value in SearchValue to find in table.
let SearchValue =  "<SearchValue>";//Please update term you would like to find in the table.
InsightsMetrics
| where * contains tostring(SearchValue)
| take 1000

正在收集什么数据?

列出收集的性能计数器和对象类型。

InsightsMetrics
| where Origin == "vm.azm.ms"
| summarize by Namespace, Name

虚拟机可用内存

虚拟机可用内存。

InsightsMetrics
| where TimeGenerated > ago(1h)
| where Origin == "vm.azm.ms"
| where Namespace == "Memory"
| where Name == "AvailableMB"
| summarize avg(Val) by bin(TimeGenerated, 5m), Computer
| render timechart

绘制 CPU 利用率趋势(按计算机)图表

计算过去一小时的 CPU 利用率模式，按百分位数绘制图表。

InsightsMetrics
| where TimeGenerated > ago(1h)
| where Origin == "vm.azm.ms"
| where Namespace == "Processor"
| where Name == "UtilizationPercentage"
| summarize avg(Val) by bin(TimeGenerated, 5m), Computer //split up by computer
| render timechart

虚拟机可用磁盘空间

显示最新每个实例的可用磁盘空间的最新报告。

InsightsMetrics
| where TimeGenerated > ago(1h)
| where Origin == "vm.azm.ms"
| where Namespace == "LogicalDisk"
| where Name == "FreeSpaceMB"
| extend t=parse_json(Tags)
| summarize arg_max(TimeGenerated, *) by tostring(t["vm.azm.ms/mountId"]), Computer // arg_max over TimeGenerated returns the latest record
| project Computer, TimeGenerated, t["vm.azm.ms/mountId"], Val

使用检测信号跟踪 VM 可用性

显示在过去一小时内报告的 VM 可用性。

InsightsMetrics
| where TimeGenerated > ago(1h)
| where Origin == "vm.azm.ms"
| where Namespace == "Computer"
| where Name == "Heartbeat"
| summarize heartbeat_count = count() by bin(TimeGenerated, 5m), Computer
| extend alive=iff(heartbeat_count > 2, 1.0, 0.0) //computer considered "down" if it has 2 or fewer heartbeats in 5 min interval
| project TimeGenerated, alive, Computer
| render timechart with (ymin = 0, ymax = 1)

CPU 利用率最高的前 10 台虚拟机

CPU 利用率最高的前 10 台虚拟机。

InsightsMetrics
| where TimeGenerated > ago(1h)
| where Origin == "vm.azm.ms"
| where Namespace == "Processor" and Name == "UtilizationPercentage"
| summarize P90 = percentile(Val, 90) by Computer
| top 10 by P90

最低的后 10 个可用磁盘空间百分比

最低的后 10 个可用磁盘空间百分比(按计算机)。

InsightsMetrics
| where TimeGenerated > ago(24h)
| where Origin == "vm.azm.ms"
| where Namespace == "LogicalDisk" and Name == "FreeSpacePercentage"
| summarize P90 = percentile(Val, 90) by Computer
| top 10 by P90 asc

通过