License: MIT License

Java 100.00%

androiddatabaselibrarycomparison's Introduction

AndroidDatabaseLibraryComparison

A test between a few of the popular libraries running a speed test on how fast they load and save data.

Note: SugarORM and ActiveAndroid and Sprinkles were removed because they're too slow. If you're curious how well they run, you can uncomment the lines in MainActivity. SugarORM was a lot slower on load/save. ActiveAndroid and Sprinkles were a lot slower than the remaining comparisons on save.

Benchmark Description

There are two benchmarks. The Simple trial uses a flat schema for an address book so each row is composed of name, address, city, state, and phone columns. The Simple trial does 25000 copies of the AddressItem.

The Complex trial is hierarchical and has support for multiple address books where each address book has contacts and addresses. The Complex trial does 50 copies of the AddressBook.

Results

These are the results for the Simple trial:

And these are the results for the Complex trial:

Older Results

These are the results for the Simple trial with the DBs that are no longer included currently:

And these are the results for the Complex trial:

Blog Post:

The original blog post can be found here

androiddatabaselibrarycomparison's People

Contributors

Stargazers

Watchers

androiddatabaselibrarycomparison's Issues

please fix it. just couldn't compile it

https://raw.githubusercontent.com/Raizlabs/maven-releases/master/ seems not working.
https://raw.github.com/xujiaao/mvn-repository/master/releases too.

Add Realm comparison

Great information but would like to see how realm.io compares

Fix broken schema images for README

At the moment both UML schemas (for simple and complex trials) at README are broken.

Add missed benchmarks results

At the moment there are benchmarks results with 0 (zero) result time which is confusing. Very unlikely that it's the real results, so most likely it means the benchmark wasn't launched for that cases.
Cases with 0 results:

Sprinkles "save" for Complex Trial
Sprinkles "load" for Complex Trial
DBFlow "load" for Complex Trial

Please share some results

Hi,

can you share some results about the benchmark?

In complex trial, why is copyToRealmOrUpdate() being called for contacts and address items?

On Realm Complex Trial code(RealmTester.java), if you call realm.copyToRealm(finalAddressBooks);, all contacts and address items referenced by these address books will be inserted as well. So why do you need to call copyToRealmOrUpdate() for contacts and items in each address book? This seems to have a big impact on Realm's performance.

That change was added on this commit: f79b824#diff-254e581a054b5c06b307166748225016

I don't understand the reason for this change. To me it seems like the code before this commit was correct.

Consider to label the axes

please. Have no idea what all the numbers mean.

"Complex Load" does not make much sense

The tests for "load" in the "complex trial", loads 50 address books and then walks along the relations potentially resolving them on the fly (doing database requests). I had a closer look why DbFlow takes just a few milliseconds (usually single digit) while all others take hundred of milliseconds. I think the test set up is questionable: after inserting the entities, they stay around in memory initialized with all relations already set up for DbFlow. So what happens is that DbFlow makes a single query and is done, while others must hit the database multiple times. So you test rather caching than loading. I could send a trivial PR for greenDAO in order to make the same caching work, which would also result in single digit ms test runs.

However, I'd rather propose to clear all caches before doing the complex load test to actually benchmark loading. For greenDAO, you can clear the memory "cache" using daoSession.clear(). I don't did not find the equivalent in DbFlow, so I could not try myself.

rightpoint / androiddatabaselibrarycomparison Goto Github PK