benchmark: calibrate config array-vs-concat

According to nodejs/performance#186 this benchmark was taking 160 secs for a single run. Based on a research in a dedicated machine, the results doesn't have variation based on the configs, so we don't need to bench all variations. Signed-off-by: RafaelGSS <[email protected]> PR-URL: #59587 Reviewed-By: Luigi Pinca <[email protected]> Reviewed-By: James M Snell <[email protected]>

Author: RafaelGSS

benchmark/dgram/array-vs-concat.js

@@ -5,18 +5,18 @@ const common = require('../common.js');
 constdgram=require('dgram');
 constPORT=common.PORT;
 
-// `num` is the number of send requests to queue up each time.
+// `n` is the number of send requests to queue up each time.
 // Keep it reasonably high (>10) otherwise you're benchmarking the speed of
 // event loop cycles more than anything else.
 constbench=common.createBenchmark(main,{
-len: [64,256,512,1024],
-num: [100],
-chunks: [1,2,4,8],
+len: [64,512,1024],
+n: [100],
+chunks: [1,4],
 type: ['concat','multi'],
 dur: [5],
 });
 
-functionmain({ dur, len,num, type, chunks }){
+functionmain({ dur, len,n, type, chunks }){
 constchunk=[];
 for(leti=0;i<chunks;i++){
 chunk.push(Buffer.allocUnsafe(Math.round(len/chunks)));
@@ -28,23 +28,23 @@ function main({dur, len, num, type, chunks }){
 constonsend=type==='concat' ? onsendConcat : onsendMulti;
 
 functiononsendConcat(){
-if(sent++%num===0){
+if(sent++%n===0){
 // The setImmediate() is necessary to have event loop progress on OSes
 // that only perform synchronous I/O on nonblocking UDP sockets.
 setImmediate(()=>{
-for(leti=0;i<num;i++){
+for(leti=0;i<n;i++){
 socket.send(Buffer.concat(chunk),PORT,'127.0.0.1',onsend);
 }
 });
 }
 }
 
 functiononsendMulti(){
-if(sent++%num===0){
+if(sent++%n===0){
 // The setImmediate() is necessary to have event loop progress on OSes
 // that only perform synchronous I/O on nonblocking UDP sockets.
 setImmediate(()=>{
-for(leti=0;i<num;i++){
+for(leti=0;i<n;i++){
 socket.send(chunk,PORT,'127.0.0.1',onsend);
 }
 });