Gremlin.js: Throw a thousand monkeys at your page

mitchellh 12 years ago |

For those who don't know: this is PROBABLY named after Palm OS Gremlins. Palm OS had a debugging tool called "gremlins" that would randomly tap and type and gesture all over your Palm OS app. This debugging tool would often find edge cases that would end up crashing your app.

As someone who used to program for Palm OS a lot, I was pretty happy to see this name. :)

apinstein 12 years ago | |

Back in the very early days of Palm OS, Palm had a "Quality Certification" you could apply for, and if your app passed, you'd get a bunch of marketing benefits. So we went for the highest-level certification on our products, which required surviving 1,000,000 gremlins.

At first I was very skeptical that it would find any meaningful bugs, but in fact it was very good at finding memory leaks and edge cases. In the end it took us several weeks to finally be able to survive the 1M gremlins without a crash. A single run of 1M gremlins would take about 24 hours.

In later years they added the capability to supply your own dictionary of values to apply to certain inputs so that gremlins could get past validation, and it would also save the entire state of the device every 10K events so that you could easily reproduce errors that Gremlins found.

I could definitely see this as a good way to stress-test certain parts of your app, though it takes a ton of time, and frankly without surrounding tooling (like resource monitoring, saving "state") it wouldn't be practical.

fzaninotto 12 years ago | |

As the author of this lib, I can confirm: it was named after the movie.

Semaphor 12 years ago | |

Or even more probable, they are both simply named after Gremlins [0]

[0]: http://www.etymonline.com/index.php?search=gremlin

quesera 12 years ago |

The 1984 Macintosh had a similar testing tool called The Monkey.

Andy Hertzfeld: http://www.folklore.org/StoryView.py?story=Monkey_Lives.txt

xhrpost 12 years ago |

I think this is essentially "Fuzz testing"[1]. Something I read about in a book years ago but never saw done in real-life. The term may be antiquated at this point. [1] http://en.wikipedia.org/wiki/Fuzz_testing

dleen 12 years ago | |

Chromium uses fuzzing extensively: http://blog.chromium.org/2012/04/fuzzing-for-security.html

billyhoffman 12 years ago | |

I wouldn't call this fuzz test because it focuses on simulating legitimate actions. Do random clicking. Do random scrolling. Do random mouseovers. Do random [any DOM event]. Yes, they are doing this at a massive scale, but the key here is that germlin.js can only do things a user can do. In short, Gremlin is just automating actions; basic legal, legitimate actions. Still cool, and great for stress testing or looking for race conditions in your UI code, but not fuzz testing.

Fuzz testing, on the other hand, is about modifying the data, to interacting with the UI. To use an analogy, fuzz testing Microsoft word might involve corrupting various structures in the DOC or DOCX file or in an OLE embed with malformed data and seeing how the parser/program reacts. The Gremlin.js equivlent would be just clicking on a bunch of buttons in the Word UI really really fast.

Both are helpful, but they are testing different things

yeukhon 12 years ago | | |

Not entirely true.

Fuzzing is not about modifying data. I don't think that's the right way to describe fuzzing.

It's about testing the durability of a program by trying all kinds of data, as greedy as possible. This includes known problematic inputs and random inputs to match some expectations. By random it can either be totally random (any length, any pattern) or protocol-aware.

Random clicking is a form of random data testing because you are trying random input stream to a program. Your argument is not entirely wrong either. By fuzzing his UI he may trigger the browser to crash. He may trigger his monkey to crash. Fuzzing is a very general technique. I can write a fuzzer that fuzz Firefox's UI Australis. What will I look for? Maybe after opening 100 tabs and closing the 45th tab the titlebar disappeared. Or resize the browser from range s to range w I will find some range will cause the UI to look ugly (style overflow, etc). Or after clicking on the scrollbar several times consecutively the browser crashed.

Barton Miller, the "Father of Fuzzing" did UI fuzzing by simulating actual keystrokes and mouse clicks. See ftp://ftp.cs.wisc.edu/paradyn/technical_papers/fuzz-nt.pdf

In his Forwards page, he even mentioned "Monkey" http://pages.cs.wisc.edu/~bart/fuzz/Foreword1.html

Mutjake 12 years ago | | |

One nice way of describing fuzzing is that it is running a search problem (searching for bugs) using more or less guided Monte Carlo method for the program's input space. Generally the testing should also be automated (test case injection and testing oracle), so that one can beat the program with enough test cases for the fuzzing to be viable.

Relating to the practical example: if one is trying to find bugs from the UI code (or it is the only way to feed inputs to the program), monkey method of fuzzing is the way to go. But if one tries to test the deeper layers of the program, it is beneficial to try to find the lowest layer of inputs we can access, since it enables faster input of test cases and thus makes the fuzzing more effective.

One way of the other, my opinion is that both are fuzzing, by the definition I gave for it :-)

yeukhon 12 years ago | | |

I like the way you treat fuzzing as search program.

Mutjake 12 years ago | | |

It's a nice way of summing up the general principle. Unfortunately I can't take the credit for coming up with that one :-)

Mutjake 12 years ago | |

It's not antiquated at all, it is still very relevant for security testing. If you're hunting memory handling bugs, for example, fuzzing is probably the most cost effective way of doing it, if you can automate instrumentation (sample input and testing oracle, IMHO AddressSanitizer is the best option for the latter ATM). If you're interested, you might want to check the wiki page of Radamsa, a general-purpose fuzzer (shameless plug for my collague): https://code.google.com/p/ouspg/wiki/Radamsa

I'm not sure how up to date the CVE list is, but it probably gives you an idea, if the fuzzing is still relevant or not :-)

sergiotapia 12 years ago |

Netflix also has a famous division of monkeys that they unleash to take down servers in the hive, and generally muck about.

They want to break as many things as possible while keeping uptime and that sounds freaking awesome! :D

midas007 12 years ago | |

Chaos monkeys meet entropy gremlins.

(Watch out for water.)

madeofpalk 12 years ago | |

More info (the open sourced it): http://techblog.netflix.com/2012/07/chaos-monkey-released-in...

bhhaskin 12 years ago | | |

That is pretty darn cool!

fzaninotto 12 years ago |

Hi! I'm one of the authors of Gremlins.js. Glad to see that some of you find it interesting.

I wrote a blog post a while ago about the purpose of the lib, it's called "Completing the JavaScript Test Stack". Worth reading if you wonder what this is all about.

http://redotheweb.com/2014/01/07/completing-the-js-test-stac...

imkevinxu 12 years ago |

lol I'm gonna call this "Twitch Plays the Internet"

wildpeaks 12 years ago | |

You read my mind, that totally looks like anarchy mode applied to to the web :)

presty 12 years ago |

For Android: http://developer.android.com/tools/help/monkey.html

atom-morgan 12 years ago |

The GIF in Github makes me think TwitchPlaysYourApplication.js would be a better name :)

Houshalter 12 years ago |

That looks really cool, though I think it would have a hard time navigating past the first few menus, and could hit features that are only rarely meant to be touched (e.g. "delete everything".) Perhaps a searching algorithm that intentionally tries to explore every menu and feature, navigate to them and then test them. Or intentionally tries things it knows are more likely to find bugs.

fzaninotto 12 years ago | |

All the gremlins are configurable, you can specify which elements they can interact on (and therefore exclude "dangerous" elements).

Relevant doc: https://github.com/marmelab/gremlins.js#configuring-gremlins

taybin 12 years ago |

Now have this play the 2048 game.

toolslive 12 years ago |

So the person who named this is old enough to have seen (and liked) http://en.wikipedia.org/wiki/Gremlins

I guess 40-something.

reubenmorais 12 years ago | |

The gremlin is older (and broader) than that movie: https://en.wikipedia.org/wiki/Gremlin

minikomi 12 years ago | |

Or older perhaps.. https://www.youtube.com/watch?v=D1xqrdtJs8w

ianbicking 12 years ago |

A similar library that I wrote: https://github.com/ianb/walkabout.js

espeed 12 years ago |

Not to be confused with the Gremlin graph traversal language (https://github.com/tinkerpop/gremlin/wiki) or the JavaScript version of Gremlin.

metaedge 12 years ago |

I just give my laptop or iPad to 2 year old son and it does the same trick!