Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/codahale/usl4j

A reasonably complete implementation of the Universal Scalability Law model.
https://github.com/codahale/usl4j

capacity-planning java java-8 scalability universal-scalability-law usl

Last synced: about 3 hours ago
JSON representation

A reasonably complete implementation of the Universal Scalability Law model.

Awesome Lists containing this project

README 

        
# usl4j



[![CircleCI](https://circleci.com/gh/codahale/usl4j.svg?style=svg)](https://circleci.com/gh/codahale/usl4j)



usl4j is Java modeler for [Dr. Neil Gunther][NJG]'s [Universal Scalability Law][USL] as described by

[Baron Schwartz][BS] in his book [Practical Scalability Analysis with the Universal Scalability

Law][PSA]. 



Given a handful of measurements of any two [Little's Law][LL] parameters--throughput, latency, and

concurrency--the [USL][USL] allows you to make predictions about any of those parameters' values

given an arbitrary value for any another parameter. For example, given a set of measurements of

concurrency and throughput, the [USL][USL] will allow you to predict what a system's average latency

will look like at a particular throughput, or how many servers you'll need to process requests and

stay under your SLA's latency requirements.



The model coefficients and predictions should be within 0.02% of those listed in the book.



## Add to your project



```xml



  com.codahale

  usl4j

  0.7.0



```



It depends on [DDogleg Numerics][DDogleg] for least-squares regression.



*Note: module name for Java 9+ is `com.codahale.usl4j`.*



## How to use this



As an example, consider doing load testing and capacity planning for an HTTP server. To model the

behavior of the system using the [USL][USL], you must first gather a set of measurements of the

system. These measurements must be of two of the three parameters of [Little's Law][LL]: mean

response time (in seconds), throughput (in requests per second), and concurrency (i.e. the number of

concurrent clients).



Because response time tends to be a property of load (i.e. it rises as throughput or concurrency

rises), the dependent variable in your tests should be mean response time. This leaves either

throughput or concurrency as your independent variable, but thanks to [Little's Law][LL] it doesn't

matter which one you use. For the purposes of discussion, let's say you measure throughput as a

function of the number of concurrent clients working at a fixed rate (e.g. you used

[`wrk2`][wrk2]).



After your load testing is done, you should have a set of measurements shaped like this:



|concurrency|throughput|

|-----------|----------|

|          1|    955.16|

|          2|   1878.91|

|          3|   2688.01|

|          4|   3548.68|

|          5|   4315.54|

|          6|   5130.43|

|          7|   5931.37|

|          8|   6531.08|



For simplicity's sake, let's assume you're storing this as a `double[][]`. Now you can build a model

and begin estimating things:



```java

import com.codahale.usl4j.Measurement;

import com.codahale.usl4j.Model;

import java.util.Arrays;



class Example {

  void buildModel() {

    final double[][] points = {{1, 955.16}, {2, 1878.91}, {3, 2688.01}}; // etc.

  

    // Map the points to measurements of concurrency and throughput, then build a model from them. 

    final Model model = Arrays.stream(points)

                              .map(Measurement.ofConcurrency()::andThroughput)

                              .collect(Model.toModel());

    for (int i = 10; i < 200; i+=10) {

      System.out.printf("At %d workers, expect %f req/sec\n", i, model.throughputAtConcurrency(i));

    }

  }

}

```



## Performance



Building models is pretty fast:



```

Benchmark         (size)  Mode  Cnt   Score   Error  Units

Benchmarks.build      10  avgt    5   0.507 ± 0.061  us/op

Benchmarks.build     100  avgt    5   1.242 ± 0.266  us/op

Benchmarks.build    1000  avgt    5   7.499 ± 0.157  us/op

Benchmarks.build   10000  avgt    5  72.321 ± 2.681  us/op

```



## Further reading



I strongly recommend [Practical Scalability Analysis with the Universal Scalability Law][PSA], a

free e-book by [Baron Schwartz][BS], author of [High Performance MySQL][MySQL] and CEO of

[VividCortex][VC]. Trying to use this library without actually understanding the concepts behind

[Little's Law][LL], [Amdahl's Law][AL], and the [Universal Scalability Law][USL] will be difficult

and potentially misleading.



I also [wrote a blog post about this library][usl4j].



## License



Copyright © 2017 Coda Hale



Distributed under the Apache License 2.0.



[NJG]: http://www.perfdynamics.com/Bio/njg.html

[AL]: https://en.wikipedia.org/wiki/Amdahl%27s_law

[LL]: https://en.wikipedia.org/wiki/Little%27s_law

[PSA]: https://www.vividcortex.com/resources/universal-scalability-law/

[USL]: http://www.perfdynamics.com/Manifesto/USLscalability.html

[BS]: https://www.xaprb.com/

[MySQL]: http://shop.oreilly.com/product/0636920022343.do

[VC]: https://www.vividcortex.com/

[DDogleg]: http://ddogleg.org/

[wrk2]: https://github.com/giltene/wrk2

[usl4j]: https://codahale.com/usl4j-and-you/