轉載：https://blog.csdn.net/qq_32373277/article/details/89453681

用 lua指令碼同時GET,POST壓測

category.lua 指令碼如下：

urimap = {
    "/v3/goods/detail", 
    "/v3/goods/status",
    "/v3/goods/sku_detail",
    "/v3/coupons/lists",
    "/v3/pages/recommend",
    "/v3/carts",
}

methodmap = {
    "POST", 
    "POST",
    "POST",
    "GET
 
",
    "GET",
    "GET",
}

params = {
      [[{"id":2}]], 
      [[{"spu_id":2,"type":1}]],
      [[{"id":2,"sku_id":7}]], -- 雙中括號裡面不轉譯
      "page=1&size=100",  
      "",
      "",
}

math.randomseed(os.time())

init = function()
    local r = {}
    local path = ""   -- 區域性變數（不加local 是全域性變數）
    local
 
 method = "get" -- 預設get

    -- header 頭
    wrk.headers["Hash"]= "85280aa135bbd0108dd6aa424565a"
    wrk.headers["Token"]= ""
   
    for i, v in ipairs(urimap) do -- 鍵從1 開始 非 0
        path = v    -- 路徑
        method = methodmap[i]  -- method

        if method == "POST" then
            wrk.headers[
 
"content-type"]= "application/json" --POST 引數json格式
            wrk.body = params[i]
        end

        if method == "GET" and  params[i] ~= "" then
            path = v .. "?" ..params[i]
        end

        io.write(method, "---", params[i], "----", path, "\n") -- 列印請求方式（1個執行緒會列印一次），引數，路徑（不含域名）
        r[i] =  wrk.format(method, path)    
    end 

    req = table.concat(r)
end

request = function()
      return req
end
    
response = function(status, headers, body)  
    if status ~= 200 then
        print("status:", status)
        print("error:", body)
        wrk.thread:stop()
    else 
       -- print("body:", body)   
    end
end  

done = function(summary, latency, requests)

    local durations=summary.duration / 1000000    -- 執行時間，單位是秒
    local errors=summary.errors.status            -- http status不是200，300開頭的
    local requests=summary.requests               -- 總的請求數
    local valid=requests-errors                   -- 有效請求數=總請求數-error請求數
  
    io.write("Durations:       "..string.format("%.2f",durations).."s".."\n")
    io.write("Requests:        "..summary.requests.."\n")
    io.write("Avg RT:          "..string.format("%.2f",latency.mean / 1000).."ms".."\n")
    io.write("Max RT:          "..(latency.max / 1000).."ms".."\n")
    io.write("Min RT:          "..(latency.min / 1000).."ms".."\n")
    io.write("Error requests:  "..errors.."\n")
    io.write("Valid requests:  "..valid.."\n")
    io.write("QPS:             "..string.format("%.2f",valid / durations).."\n")
    io.write("--------------------------\n")
  
end

指令碼就這麼完事了，開始壓測吧！

wrk -t1 -c1000 -d30s -T10s -s category.lua --latency https://shop-api-crs.chi.com  

-t 開啟多少執行緒
-c 連線數量 
-T 超時時間（timeout），支援時間單位 (s, m, h)
-s Lua指令碼路徑  
--latency 壓測結束完，列印統計資訊  

返回：

Running 15s test @ https://shop-api-crs.chi.com 
  1 threads and 10 connections. 
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency   124.50ms  173.08ms   1.11s    47.38%
    Req/Sec   252.53    111.40   363.00     79.86%
  Latency Distribution
     50%  381.93ms
     75%    0.00us
     90%    0.00us
     99%    0.00us
  3674 requests in 15.08s, 2.51MB read  --（15.08秒內共處理完成了 3674個請求，讀取了2.51MB資料）
Requests/sec:    243.56    -- （平均每秒處理完成請求）
Transfer/sec:    170.29KB  -- （平均每秒讀取資料）
Durations:       15.08s     -- 執行時間，單位是秒
Requests:        3674   -- 總的請求數
Avg RT:          124.50ms  --平均值
Max RT:          1110.714ms  -- 最大值
Min RT:          78.324ms   -- 最小值
Error requests:  0       -- Error請求數
Valid requests:  3674  -- 有效請求數
QPS:             243.56   --QPS

轉載：https://blog.csdn.net/qq_32373277/article/details/89453681

轉載：https://blog.csdn.net/qq_32373277/article/details/89453681