Mongo Benchmark

Benchmark a set of collection schemas by testing CRUD speed.

Getting started

Run Benchmark

To use this app, there are few steps you need to do.

Download the latest build of this program at the release page.
Unzip your downloaded file.
Open benchmark.config.json file, and make sure your configuration satisfy your needs. The configuration follows these rules.
Also make sure that your data directory (the one defined in benchmark.config.json > app > dataDirectory) satisfy your needs; and it's structure follows these rules.
Run the application.

Data Directory

Structure

data
│
└───<FirstDatabaseName>
│   │   flows.json
│   │   
│   └───<nameOfCollectionA>
│   │   │
│   │   │   <data1>.create.json
│   │   │   <data1>.update.json
│   │
│   └───<nameOfCollectionB>
│       │
│       │   <data1>.create.json
│       │   <data2>.create.json
│       │   <data2>.update.json
│
└───<SecondDatabaseName>
│   │
│   └───<nameOfCollectionC>
│       │   
│       │   <data1>.create.json
│       │   <data1>.update.json
│   │   │   <data2>.create.json
│   │   │   <data2>.update.json

Rules

Database Directory - Data directory should have child directories, which directories' name represent database name. These directories are called database directories. Data directory should have one or more database directory.
Collection Directory - Database directory should have child directories, which directories' name represent collection name. These directories are called collection directories. Database directory should have one or more collection directory.
Flows Configuration File - Database directory may have a flows configuration file named flows.json, written in JSON format standard according to these rules. Flow configuration files defines of the program operates on READ operations.
Data File Set - Collection directory should have data file sets. Data file set is a set of files that defines the data operated on CREATE and UPDATE operation. The set consist of data create file (named: <dataA>.create.json) and an optional data update file (named: <dataA>.update.json). When optional data update file is present, it will be used as the update data for the data created by data create file which has the same name. When optional data update file is not present, it will use it data create file counterpart to use as the update data. The contents of data file is free and depends on your needs. Collection directory should have one or more data file set.

Flows Configuration

The flows configuration file, named flows.json, must satisfy following formats.

{
   "collections": {
       "<collectionName>": {
           "<propertyName>": {
               "refCollection": "Referenced collection",
               "findById": "The row ID to be referenced with. Should not present when findOne is present.",
               "findOne": "Condition to find the row to be referenced with. Should not present when findById is present."
            }
       }
   },
   "readOtherCollections": "true or false; whether or not to read any other collections, others than the ones which defines in this flows configuration file"
}

The value of refCollection, findById, and findOne may have special variable inserted. These variables are provided and they should be written in enclosed percentage {}.

{value} - Return the value of the property.
{row} - Return the object of the row.

When the return value of variable is an object (for example {row}), you may specify the property you'd like to return like this {row.column_first}.

Below is an example of flows configuration file.

{
   "collections": {
       "logs": {
           "ref_one": {
               "refCollection": "collection_{value}",
               "findById": "{row.ref_id}"
            },
           "ref_two": {
               "refCollection": "collection_{value}",
               "findOne": {
                   "id": "{row.ref_id}"
               }
            }
       },
       "services": {}
   },
   "readOtherCollections": false
}

Build from Source

Preparation

To build this program from source, make sure you already have NodeJS and Yarn Package Manager installed.

First clone this repository.

git clone https://github.com/danang-id/mongo-benchmark.git
cd mongo-benchmark

Then, install dependencies needed for program to run.

yarn install

On Development

On development process, you may want to start the app. To do so, use the dev script.

yarn dev

Build into JavaScript executable

To build the application into a single JavaScript executable, use the build script.

yarn build

The executable files is then located on bin folder of root directory. It consist of:

mongobenchmark: JavaScript executable file (run directly on Linux/macOS, or using node mongobenchmark)
mongobenchmark.cmd: Wrapper for Windows
mongobenchmark.sh: Wrapper for Linux/macOS

You may run one of these files to start the built app. You may also use yarn start script to start bin/mongobenchmark file.

Remember: To start the app using this method, you still need NodeJS installed on the host system. Yet, no need to satisfy the dependency.

Build into binary executable files for distribution

Even is you're able to execute bin/mongobenchmark, it was not written in a binary executable format. Thus, it will not works when you have no NodeJS installed and is not appropriate for binary distribution.

To make binary executable files for distribution, you have to run the dist script.

yarn dist

This command will generate binary executable files with NodeJS 10 included for Linux (64 bit) and Windows (both 32 and 64 bit).

You may also run one of these commands to build of specific OSes accordingly.

yarn dist:linux (Linux 64 bit)
yarn dist:win (Windows, both 32 and 64 bit)
yarn dist:win32 (Windows 32 bit)
yarn dist:win64 (Windows 64 bit)
yarn dist:macos (macOS, see Note on macOS Build below)

The distribution binary will be generated at dist folder.

Note on macOS Build

yarn dist:macos

To build macOS binary, you need to run on macOS. Otherwise, if you try to build macOS binary not using macOS, an error shall appear.

Configuration

Configuration for this program is saved on benchmark.config.json. This file can be located on:

besides the binary executable file; or
the root of this project.

The configuration file consist of these settings.

{
   "app": {
       "dataDirectory": "Absolute path to data directory."
   },
   "database": {
       "uriString": "MongoDB URI String to connect to."
   }
}

Example of the configuration file is provided here.

Contribution

To contribute, simply fork this project, and issue a pull request.

Versioning

This project is using SemVer for versioning. For the versions available, see the tags on this repository.

Authors

Danang Galuh Tegar Prasetyo - Initial work - danang-id

See also the list of contributors who participated in this project.

License

This project is licensed under the Apache License version 2.0 (Apache-2.0). See LICENSE for details.