由實體機追查哪一台 VM 的流量最大

# 抓取第一次數據
$before = Get-Counter -Counter @(
    '\Hyper-V Virtual Network Adapter(*)\Bytes Sent/sec',
    '\Hyper-V Virtual Network Adapter(*)\Bytes Received/sec'
)

# 等待 10 秒
Start-Sleep -Seconds 10

# 抓取第二次數據
$after = Get-Counter -Counter @(
    '\Hyper-V Virtual Network Adapter(*)\Bytes Sent/sec',
    '\Hyper-V Virtual Network Adapter(*)\Bytes Received/sec'
)

# 分析並計算平均每秒流量
$results = @()

for ($i = 0; $i -lt $before.CounterSamples.Count; $i++) {
    $instance = $before.CounterSamples[$i].InstanceName

    $sentIndex = $i
    $recvIndex = $i + $before.CounterSamples.Count / 2

    $sentBefore = $before.CounterSamples[$sentIndex].CookedValue
    $sentAfter = $after.CounterSamples[$sentIndex].CookedValue

    $recvBefore = $before.CounterSamples[$recvIndex].CookedValue
    $recvAfter = $after.CounterSamples[$recvIndex].CookedValue

    $sentPerSec = ($sentAfter - $sentBefore) / 10
    $recvPerSec = ($recvAfter - $recvBefore) / 10

    $results += [PSCustomObject]@{
        AdapterName   = $instance
        SentPerSec    = [math]::Round($sentPerSec, 2)
        ReceivedPerSec = [math]::Round($recvPerSec, 2)
        TotalPerSec   = [math]::Round($sentPerSec + $recvPerSec, 2)
    }
}

# 顯示結果，依照總流量排序
$results | Sort-Object TotalPerSec -Descending | Format-Table -AutoSize



用 performance monitor 監查看各VM的流量(找不到)
Get-Counter '\Hyper-V Virtual Network Adapter(*)\Bytes Total/sec'

在 VM 上追查哪個網站流量最大
用 performance monitor 監查看各網站的流量
Web Service: Bytes Sent/sec
 

--建立 table (要記得改 XXX)
CREATE TABLE IF NOT EXISTS public.table_monitor
(
    table_name text COLLATE pg_catalog."default" NOT NULL,
    update_at timestamp without time zone NOT NULL DEFAULT now(),
    update_count bigint NOT NULL DEFAULT 0,
    CONSTRAINT table_monitor_pkey PRIMARY KEY (table_name)
)

TABLESPACE pg_default;

ALTER TABLE IF EXISTS public.table_monitor
    OWNER to XXX;

-- 這一段只要執行一次
CREATE OR REPLACE FUNCTION public.table_monitor_trigger_fnc()
    RETURNS trigger
    LANGUAGE 'plpgsql'
    COST 100
    VOLATILE NOT LEAKPROOF
AS $BODY$
BEGIN
     Update public.table_monitor Set update_count = update_count + 1, update_at = now()  Where table_name = TG_TABLE_NAME; 
RETURN NEW;
END;
$BODY$;

ALTER FUNCTION public.table_monitor_trigger_fnc()
    OWNER TO XXX;

-- 對於被監控的 table, 要執行以下的指令，記得 monitor_table_name 要改為 table 的名字(要符合大小寫)
INSERT INTO public.table_monitor (table_name, update_at, update_count) VALUES ('monitor_table_name', '2000-01-01', 0) ON CONFLICT DO NOTHING;

CREATE OR REPLACE TRIGGER table_monitor_trigger
    AFTER INSERT OR DELETE OR UPDATE 
    ON monitor_table_name
    FOR EACH STATEMENT
    EXECUTE FUNCTION public.table_monitor_trigger_fnc();

監控用的程式碼:

public static void UpdateTableCache()

{

    //標記自已是正在執行的 Thread. 讓舊的 Thread 在執行完畢之後, 應該會自動結束. 以防舊的 Thread 執行超過 10 秒.

    CurrentThreadId = System.Threading.Thread.CurrentThread.ManagedThreadId;



    try

    {

        //確認自已是正在執行的 Thread, 重覆執行. (另一個 Thread 插入執行)

        while (CurrentThreadId == System.Threading.Thread.CurrentThread.ManagedThreadId)

        {

            LastUpdateDate = DateTime.Now;



            foreach (var dbId in CachedDbs)

            {

                var sql = @"select * from table_monitor";



                var dt = Su.PgSql.DtFromSql(sql, dbId);



                foreach (DataRow row in dt.Rows)

                {

                    string changeId = row["update_count"].DBNullToDefault();

                    string CacheKey = TableCacheKey(dbId, row["table_name"].DBNullToDefault());

                    ObjectCache cache = MemoryCache.Default;



                    string OldValue = (string)cache[CacheKey];



                    if (OldValue == null)

                    {

                        cache.Set(CacheKey, changeId, DateTime.MaxValue);

                    }

                    else

                    {

                        if (changeId != OldValue)

                        {

                            cache.Remove(CacheKey);

                            cache.Set(CacheKey, changeId, DateTime.MaxValue);

                        }

                    }

                }

            }



            //每兩秒檢查一次

            System.Threading.Thread.Sleep(2000);

        }

    }

    catch (Exception)

    {

        //依經驗, 只要 DB 能通, 這裡幾乎不會有問題, 所以這裡暫時不處理, 未來有問題時可以考慮寫入文字檔比較好.

    }

}