Post your wrk results

[griimnak]

I wanted to know what are some average numbers you guys get on frameworks / backend sites you've written.
Any language, any framework, using wrk and these settings:

• Tested framework / backend must have a mysql connection, must use a config file, must be in live production real world scenario, not localhost.

(install wrk or git clone

You must be registered for see links

)

wrk -c 20 -d 30 -t 10 --latency http://your-live-site

Here are mine,
python:flask:apache

Running 30s test @ 
  10 threads and 20 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency   134.20ms   66.31ms   1.91s    98.73%
    Req/Sec    15.46      5.16    20.00     56.31%
  Latency Distribution
     50%  128.02ms
     75%  134.09ms
     90%  138.43ms
     99%  284.46ms
  4081 requests in 30.10s, 12.35MB read
  Socket errors: connect 0, read 0, write 0, timeout 10
Requests/sec:    135.60
Transfer/sec:    420.31KB

NOTE: Don't use localhost, it's inaccurate:

Running 30s test @ http://localhost:5000
  10 threads and 20 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     9.97ms    1.07ms  29.54ms   83.39%
    Req/Sec   201.38      8.06   260.00     72.33%
  Latency Distribution
     50%    9.91ms
     75%   10.25ms
     90%   10.88ms
     99%   13.63ms
  60282 requests in 30.08s, 177.35MB read
Requests/sec:   2004.23

 

[MayoMayn]

Running 30s test @ http://localhost:8080/threads/1
  10 threads and 20 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    10.02ms   46.67ms 564.38ms   96.93%
    Req/Sec     0.92k   166.15     1.77k    84.98%
  Latency Distribution
     50%    1.79ms
     75%    2.20ms
     90%    7.66ms
     99%  283.30ms
  267219 requests in 30.02s, 45.11MB read
Requests/sec:   8900.79
Transfer/sec:      1.50MB

Had to test Titan :>

 

[griimnak]

Running 30s test @ http://localhost:8080/threads/1
  10 threads and 20 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    10.02ms   46.67ms 564.38ms   96.93%
    Req/Sec     0.92k   166.15     1.77k    84.98%
  Latency Distribution
     50%    1.79ms
     75%    2.20ms
     90%    7.66ms
     99%  283.30ms
  267219 requests in 30.02s, 45.11MB read
Requests/sec:   8900.79
Transfer/sec:      1.50MB

Had to test Titan :>

lang and framework?

Also i said don't use localhost nub -,-

 

[MayoMayn]

lang and framework?

Also i said don't use localhost nub -,-

Like I'm gonna put Titan live and cluster it for your stupid test :c

Node.js &

You must be registered for see links

 

[griimnak]

Like I'm gonna put Titan live and cluster it for your stupid test :c

Node.js &

You must be registered for see links

okay, i'll update the rules then, clown

• Tested framework / backend must have a mysql connection, must use a config file, must be in live production real world scenario, not localhost.

 

[M8than]

Well theres multiple issues with this. This depends a lot of server specs and web server configs and network latency. You'll never find comparable results.

 

[griimnak]

Well theres multiple issues with this. This depends a lot of server specs and web server configs and network latency. You'll never find comparable results.

figured that

 

[Ecko]

don't use this on my server you idiot

 

[griimnak]

don't use this on my server you idiot

vulnerability exposed

Yeah sorry, didn't know untill @Muffinss told me it's a bad idea to use on shared. Just gonna use that one result to go by

 

[M8than]

don't use this on my server you idiot

To be fair though he can barely generate 200 req/s your server probably didn't even see a spike

 

[Ecko]

I use CloudLinux LVE, all his results will be greatly skewed and not accurate.