multithreading - 如何解决心跳超过“00:00:01”失败？

Question

我有一个执行 HTTP POST 的 .NetCore C# 项目。该项目是在 Kubernetes 中设置的，我注意到以下日志：

Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:45 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:46 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:47 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:48 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:49 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:50 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:51 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:52 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:53 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:54 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:55 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:43:56 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:44:33 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:44:34 +00:00".
warn: Microsoft.AspNetCore.Server.Kestrel[22]
      Heartbeat took longer than "00:00:01" at "02/22/2020 15:44:35 +00:00".

经过一些初步研究，这似乎是线程池饥饿的常见结果。因此，去年 11 月，我将帖子设为异步，并将 Max threads 和 Available threads 记录如下以用于监控目的：

ThreadPool.GetMaxThreads(out int workerThreads, out int completionPortThreads);
ThreadPool.GetAvailableThreads(out int workerThreadAvailable, out int completionPortThreadsAvailable);
_log.Info(new { message = $"Max threads = {workerThreads} and Available threads = {workerThreadAvailable}" });

在过去的几个月中，日志始终显示：Max threads = 32767 和 Available threads = 32766。这似乎很好，但是，我注意到同样的心跳错误，所以想知道这是否真的是线程池饥饿问题。可能有人知道发生了什么，如果这个错误实际上是由其他原因引起的？对此的任何调查/解决提示将不胜感激！

标签： multithreadingasp.net-corekubernetesthreadpoolkestrel-http-server