Skip to content

xpetit/fizzbuzz

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

FizzBuzz

FizzBuzz is a Golang HTTP server exposing a RESTful web API that provides a Fizz buzz service.

It has two endpoints:

  • /api/v2/fizzbuzz
    • Accepts five optional query parameters : three integers int1, int2 and limit, and two strings str1 and str2.
      The default values are: limit=10, int1=2, int2=3, str1=fizz, str2=buzz.
    • Returns a list of strings with numbers from 1 to limit, where: all multiples of int1 are replaced by str1, all multiples of int2 are replaced by str2, all multiples of int1 and int2 are replaced by str1str2.
  • /api/v2/fizzbuzz/stats
    • Accept no parameters
    • Return the parameters corresponding to the most used request, as well as the number of hits for this request

The server is:

  • Ready for production
  • Easy to maintain by other developers

Usage

Requirements:

Use this command to directly update and run the service:

go run github.com/xpetit/fizzbuzz/v5/cmd/fizzbuzzd@latest

To stop it, type CTRL-C.

If you don't trust this program, you can use Docker. Clone this repository and run the following commands inside:

docker build --tag github.com/xpetit/fizzbuzz/v5 .
docker run --rm --publish 8080:8080 github.com/xpetit/fizzbuzz/v5

When the service is running, you can query with it with curl:

curl localhost:8080/api/v2/fizzbuzz/stats
{"most_frequent":{"count":0}}

Using the defaults:

curl localhost:8080/api/v2/fizzbuzz
["1","fizz","buzz","fizz","5","fizzbuzz","7","fizz","buzz","fizz"]

The config object is now populated:

curl localhost:8080/api/v2/fizzbuzz/stats
{"most_frequent":{"count":1,"config":{"limit":10,"int1":2,"int2":3,"str1":"fizz","str2":"buzz"}}}

Custom request:

curl localhost:8080/api/v2/fizzbuzz -Gdlimit=1 -dint1=1 -dint2=1 -dstr1=buzz -dstr2=lightyear
["buzzlightyear"]

Design

In writing this library, several considerations were taken into account:

  1. Should it provide a generalized, extensible implementation that supports any number of "int" and "str"?
  2. Should it provide a FizzBuzz function that not only writes values but returns values?
    1. Should the returned values be a slice of JSON string or a more compact (binary) representation for later reuse?

This was discarded for the following reasons:

  1. If requirements change, it is easy to adapt the Fizz buzz algorithm because it is simple.
  2. A library that streams JSON strings and one that generates them all at once are a very different story. One cannot be the generalized/abstract version of another.

relevant XKCD

Packages

The top-down list of dependencies is as follows:

  • github.com/xpetit/fizzbuzz/v5/cmd/fizzbuzzd: The main program, running the HTTP server.
  • github.com/xpetit/fizzbuzz/v5/handlers: The HTTP handlers.
  • github.com/xpetit/fizzbuzz/v5/stats: The statistics services.
  • github.com/xpetit/fizzbuzz/v5: The Fizz buzz writer WriteTo.

Performance

Benchmarks performed on an AWS EC2 instance (c6i.4xlarge) with the following specs:

CPU Intel Xeon Platinum 8375C CPU @2.90GHz
RAM 32 GB ECC 2666 MHz
Go Version 1.20.5

The programs were compiled with GOAMD64=v4 and ran with GOGC=1000 environment variables.

WriteTo limits memory allocation, here are the results from 0 to 10 million values of Fizz buzz:

BenchmarkWriteTo/[limit:0e+00]-16   271018197       4.431 ns/op   677.05 MB/s    0 B/op   0 allocs/op
BenchmarkWriteTo/[limit:1e+00]-16    11806252       99.79 ns/op    60.13 MB/s   32 B/op   5 allocs/op
BenchmarkWriteTo/[limit:1e+01]-16     5970025       200.2 ns/op   334.67 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+02]-16     1000000        1016 ns/op   685.71 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+03]-16      114363       10508 ns/op   694.45 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+04]-16       10000      106170 ns/op   718.63 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+05]-16        1074     1113399 ns/op   715.19 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+06]-16         100    11305590 ns/op   733.82 MB/s   48 B/op   6 allocs/op
BenchmarkWriteTo/[limit:1e+07]-16           9   117342031 ns/op   735.43 MB/s   49 B/op   6 allocs/op

The service stops writing values as soon as the API consumer no longer requests them.

wrk reports 447k requests/second with a random limit between 1 and 100 (using -db off command-line argument):

Running 10s test @ http://127.0.0.1:8080
  5 threads and 600 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     1.45ms    1.85ms  47.51ms   90.65%
    Req/Sec    90.68k     7.72k  111.89k    68.94%
  4511934 requests in 10.10s, 1.86GB read
Requests/sec: 446918.14
Transfer/sec:    188.49MB

Leaving the database enabled:

Running 10s test @ http://127.0.0.1:8080
  4 threads and 400 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    39.17ms   70.20ms 770.02ms   88.86%
    Req/Sec     7.81k   636.24    10.35k    76.25%
  310796 requests in 10.02s, 131.31MB read
Requests/sec:  31002.22
Transfer/sec:     13.10MB

And with -db :memory: command-line argument (SQLite in-memory DB):

Running 10s test @ http://127.0.0.1:8080
  2 threads and 10 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency   242.48us  189.82us   2.21ms   86.00%
    Req/Sec    22.80k   635.84    24.37k    65.84%
  458245 requests in 10.10s, 193.35MB read
Requests/sec:  45371.08
Transfer/sec:     19.14MB