Beego 和 Newrelic 集成问题

今日不肯埋头，明日何以抬头！每日一句努力自己的话哈哈~哈喽，今天我将给大家带来一篇《Beego 和 Newrelic 集成问题》，主要内容是讲解等等，感兴趣的朋友可以收藏或者有更好的建议在评论提出，我都会认真看的！大家一起进步，一起学习！

我正在使用 beego (https://beego.me) 在 go 中创建一个 web 应用程序。

我需要捕获 newrelic 中的应用程序监控和计量指标，并能够查看 newrelic 中的所有事务。

我已遵循此文档，并使用 beego gorelic 在我的代码中初始化 newrelic 代理。

这是我的 router.go 类 -

import (
    "github.com/astaxie/beego"
    "github.com/yvasiyarov/beego_gorelic"
)

func init() {
    beego_gorelic.initnewrelicagent()
    beego.router("path", &controller{}, "functionname") // code broken intentionally
}

在我的 app.conf 中，我提供了 newrelic 许可证密钥和应用程序名称，如下所示：

newreliclicense = "key"
newrelicappname = "app name"

当我在 dev 模式下运行应用程序时，我能够看到正在打印的指标，如下所示：

send data:{
    "agent": {
        "host": "localhost",
        "version": "0.0.6",
        "pid": 13942
    },
    "components": [
        {
            "name": "name",
            "guid": "com.github.yvasiyarov.gorelic",
            "duration": 60,
            "metrics": {
                "component/runtime/gc/gctime/max[nanoseconds]": 0,
                "component/runtime/gc/gctime/mean[nanoseconds]": 0,
                "component/runtime/gc/gctime/min[nanoseconds]": 0,
                "component/runtime/gc/gctime/percentile95[nanoseconds]": 0,
                "component/runtime/gc/numberofgccalls[calls]": 0,
                "component/runtime/gc/pausetotaltime[nanoseconds]": 0,
                "component/runtime/general/nocgocalls[calls]": 1,
                "component/runtime/general/nogoroutines[goroutines]": 6,
                "component/runtime/memory/inuse/heap[bytes]": 538368,
                "component/runtime/memory/inuse/mcacheinuse[bytes]": 13888,
                "component/runtime/memory/inuse/mspaninuse[bytes]": 13192,
                "component/runtime/memory/inuse/stack[bytes]": 360448,
                "component/runtime/memory/inuse/total[bytes]": 538368,
                "component/runtime/memory/operations/nofrees[frees]": 128,
                "component/runtime/memory/operations/nomallocs[mallocs]": 2903,
                "component/runtime/memory/operations/nopointerlookups[lookups]": 0,
                "component/runtime/memory/sysmem/buckhash[bytes]": 3078,
                "component/runtime/memory/sysmem/heap[bytes]": 66748416,
                "component/runtime/memory/sysmem/mcache[bytes]": 16384,
                "component/runtime/memory/sysmem/mspan[bytes]": 16384,
                "component/runtime/memory/sysmem/stack[bytes]": 360448,
                "component/runtime/memory/sysmem/total[bytes]": 69928960,
                "component/http/responsetime/max[ms]": 0,
                "component/http/responsetime/mean[ms]": 0,
                "component/http/responsetime/min[ms]": 0,
                "component/http/responsetime/percentile75[ms]": 0,
                "component/http/responsetime/percentile90[ms]": 0,
                "component/http/responsetime/percentile95[ms]": 0,
                "component/http/throughput/1minute[rps]": 0,
                "component/http/throughput/ratemean[rps]": 0
            }
        }
    ],
    "url": "https://platform-api.newrelic.com/platform/v1/metrics"
}

这告诉我新的遗迹代理已启动并正在捕获数据。

除此之外，我没有编写任何其他代码来将 newrelic 集成到我的代码中。

这些指标没有通过这些设置和设置报告给 newrelic，而且我无法找到更好的文档来解决这个问题。

我错过了什么？

此外，还有其他方法可以将 newrelic 集成到 beego 应用程序中吗？

附注我没有使用beego filters作为预处理器，而是在控制器中使用prepare方法来预处理请求。然而，这个库似乎在 beego 框架中添加了 2 个过滤器，如下所示：

beego.insertfilter("*", beego.beforerouter, initnewrelictimer, false)
    beego.insertfilter("*", beego.finishrouter, reportmetricstonewrelic, false)

其他日志：

2020/01/09 20:13:29 Init GC metrics collection. Poll interval 10 seconds.
2020/01/09 20:13:29 Init memory allocator metrics collection. Poll interval 60 seconds.
2020/01/09 20:13:29 Init HTTP metrics collection.
2020/01/09 20:13:29 [I] [gorelic.go:71] NewRelic agent started 
2020/01/09 20:13:29 Can not get metrica: Runtime/System/Threads, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Can not get metrica: Runtime/System/FDSize, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Can not get metrica: Runtime/System/Memory/VmPeakSize, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Can not get metrica: Runtime/System/Memory/VmCurrent, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Can not get metrica: Runtime/System/Memory/RssPeak, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Can not get metrica: Runtime/System/Memory/RssCurrent, got error:this metrica was not implemented yet for darwin
2020/01/09 20:13:29 Send data:{
    "agent": {
        "host": "localhost",
        "version": "0.0.6",
        "pid": 14931
    },
    "components": [
        {
            "name": "App Name",
            "guid": "com.github.yvasiyarov.GoRelic",
            "duration": 60,
            "metrics": {
                "Component/Runtime/GC/GCTime/Max[nanoseconds]": 0,
                "Component/Runtime/GC/GCTime/Mean[nanoseconds]": 0,
                "Component/Runtime/GC/GCTime/Min[nanoseconds]": 0,
                "Component/Runtime/GC/GCTime/Percentile95[nanoseconds]": 0,
                "Component/Runtime/GC/NumberOfGCCalls[calls]": 0,
                "Component/Runtime/GC/PauseTotalTime[nanoseconds]": 0,
                "Component/Runtime/General/NOCgoCalls[calls]": 1,
                "Component/Runtime/General/NOGoroutines[goroutines]": 8,
                "Component/Runtime/Memory/InUse/Heap[bytes]": 591616,
                "Component/Runtime/Memory/InUse/MCacheInuse[bytes]": 13888,
                "Component/Runtime/Memory/InUse/MSpanInuse[bytes]": 16592,
                "Component/Runtime/Memory/InUse/Stack[bytes]": 393216,
                "Component/Runtime/Memory/InUse/Total[bytes]": 591616,
                "Component/Runtime/Memory/Operations/NoFrees[frees]": 132,
                "Component/Runtime/Memory/Operations/NoMallocs[mallocs]": 2988,
                "Component/Runtime/Memory/Operations/NoPointerLookups[lookups]": 0,
                "Component/Runtime/Memory/SysMem/BuckHash[bytes]": 3078,
                "Component/Runtime/Memory/SysMem/Heap[bytes]": 66715648,
                "Component/Runtime/Memory/SysMem/MCache[bytes]": 16384,
                "Component/Runtime/Memory/SysMem/MSpan[bytes]": 32768,
                "Component/Runtime/Memory/SysMem/Stack[bytes]": 393216,
                "Component/Runtime/Memory/SysMem/Total[bytes]": 69928960,
                "Component/http/responseTime/max[ms]": 0,
                "Component/http/responseTime/mean[ms]": 0,
                "Component/http/responseTime/min[ms]": 0,
                "Component/http/responseTime/percentile75[ms]": 0,
                "Component/http/responseTime/percentile90[ms]": 0,
                "Component/http/responseTime/percentile95[ms]": 0,
                "Component/http/throughput/1minute[rps]": 0,
                "Component/http/throughput/rateMean[rps]": 0
            }
        }
    ],
    "URL": "https://platform-api.newrelic.com/platform/v1/metrics"
}

这可能是问题所在吗？我该如何解决这个问题？

解决方案

我正在回答我自己的问题，因为我相信我已经找到了一种方法，并且它也可能对其他人有帮助 - 显然这只是更努力地寻找现有插件的问题。我发现这个插件：https://github.com/sergeylanzman/newrelic_beego 正是我所需要的。

如果我们观察代码，作者创建了 3 个方法 -

1. 开始交易
2. 名称交易
3. 结束交易

并将它们作为过滤器插入 beego 环境中。在底层，作者使用了官方的 newrelic go agent (https://github.com/newrelic/go-agent)，并使用 transactions 来报告 apm 数据。

这里是来自插件的示例代码，它注册了 3 个过滤器 -

beego.InsertFilter("*", beego.BeforeRouter, StartTransaction, false)
    beego.InsertFilter("*", beego.AfterExec, NameTransaction, false)
    beego.InsertFilter("*", beego.FinishRouter, EndTransaction, false)
    beego.Info("NewRelic agent start")

它使用 beego 的 context 结构来跟踪跨函数的交易数据（如 context 文档中所述）

随着我们的进展，我将继续对此进行更多探索，并添加对外部调用、数据库调用和分布式跟踪的跟踪，但这解决了我的问题。

附注我已经停止使用之前使用的插件，因为这个包还会向 newrelic 报告 go 运行时统计信息，从而使其他插件过时（个人意见:)）

到这里，我们也就讲完了《Beego 和 Newrelic 集成问题》的内容了。个人认为，基础知识的学习和巩固，是为了更好的将其运用到项目中，欢迎关注golang学习网公众号，带你了解更多关于的知识点！