Am I scaling this wrong?

[xJade00]

I have a basic application with a few classes, and I tried to benchmark with a simple

    override fun handle(baseRequest: Request, request: HttpServletRequest, response: HttpServletResponse) {
        response.withStatus(200).sendJson(JsonCreator.from(value = TestReturn("Hello, world")))
        baseRequest.done()
    }

I did 10K requests at 100 requests, and it took a long time.

Using a localhost, is that the problem? Am I doing the whole scaling thing wrong?

[SeunAdelekan]

Hi @Xaanit, I'm curious as well as to why you are getting such results in your benchmark. Running a quick benchmark with ab - 15000 requests with a concurrency level of 50 - on localhost, Kanary performed as shown in the image below.

As shown above, response times were significantly better than those that you got. There are various reasons why you may be getting such results on your system. Can you shed more light on the environment you ran the benchmark on and the benchmarking tool you utilized? In addition, it will be appreciated if you can run the same benchmark with ab (Apache server benchmarking tool - http://httpd.apache.org/docs/2.2/programs/ab.html) and share you results on this issue. I'm interested in finding out what led to your results.

Cheers!

[xJade00]

Hey! I'm using this (which is basically apache's but I could run it faster :P) and I get similar (even better) stats when using ~15K requests with 50 concurrent, mine was 100 requests with 100K concurrent. Maybe try that and see if you get similar? Tomorrow I'll upload exactly what I have to gist code-wise. Thanks @SeunAdelekan

[xJade00]

Sorry! Very busy lately. Here's the promised gist: https://gist.github.com/xaanit/f8914a40c499d051a3839c9472c8d82c @SeunAdelekan