Skip to content
Open Source NoSQL Database for Unique Key Tables
Branch: develop
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


A New Open Source NoSQL Database for Unique Key Tables

The basic concept of a scattered database is that all data is stored as JSON files scattered across the file system in a logical hierarchy. One of the oldest and most tuned parts of the operating system is the file system. If you ask for a single file with a full path, you can open and read a file almost instantly, and as disks get faster and larger, the file system makes a lot of sense as an index to information. It made sense on floppies, on today's disks, it is amazing. ZarahDB makes storing data in JSON files simple, structured, reliable and robust.


There is no installation. Seriously, it's a database server, without a database server! The Operating System IS the database. If you can boot a windows server, a desktop, or even a laptop, really any Windows device of any kind will do, if there is a disk of some sort, it will do.

If you can share the disk, as a shared folder, NAS, SAN or any other way to share disk space, then you can add as many nodes as the drives can handle. With today's fastest SANs that's a LOT of nodes! Need to scale even bigger? Break the scattering across physical disks to offer even more speed and handle even more nodes.


If you are thinking infrastructure, start with any PC, even old PCs perform pretty well for most uses. If you are thinking architecture, you have two approaches.

The first is a 50K DLL that you add easily to any project as a NuGet package. You then have a rich API at your disposal.

The second is to use the WebAPI/Swagger based interface for an SOA architecture. The Web API uses the 50K DLL, so in the end, you can use the DLL directly, or you can use the Web API wrapper, it all depends on what fits your application best.


  1. Fork it! Please, I encourage you. I promise, to be honest, and open, merging any changes that make sense, from any contributor.
  2. Create your feature branch: git checkout -b my-new-feature
  3. Commit your changes: git commit -am 'Add some feature'
  4. Push to the branch: git push origin my-new-feature
  5. Submit a pull request :D

Using Visual Studio

Click HERE to learn how to contribute using Visual Studio


Starting in the late '80s I worked on a series of projects and used a Scattered Database to add performance to several of the systems I designed. I got a lot of pushback once SQL Server was released, but there were times when the Scattered Database could easily outperform SQL Server. With the growth in popularity of NoSQL databases, I decided it was time to set the scattered database free. I hope you like it!

Oh, and Zarah is Hebrew for Scattered... that's where the name comes from.


By Mike Reed of Benchmark Solutions LLC, a Kansas Corporation. Make a real, positive and consconstructive difference to the project, and I may just add your name here! Seriously!


Copyright 2015 Benchmark Solutions LLC
Originally created by Mike Reed

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
See the License for the specific language governing permissions and
limitations under the License.

Quick Start Tutorial

I'll get better instructions up before the full release, but hopefully this will help get you started. You might want to look at the Overview before diving into the tutorial.

The basic concept of a scattered database is that all data is stored as JSON files scattered across the file system in a logical hierarchy. Here are some concepts that will help you visualize how the system works:

It's a database, but it is based on scattering data in files, like this:

  1. On the disk there is a top level folder called an instance. Under the instance are sub-folders called tables.
  2. Tables have a hierarchy of folders based on the key and the selected max depth (default 5) using the first letters of the key.
  3. In the lowest folder of the hierarchy there will be a JSON file, named the same as the key (or close).
  4. In the JSON is a node named after the key, and that node has an array of column/value pairs.

Instance = Root Folder for the scattered database (Zarah is Hebrew for Scattered). Table = Sub-folder. In standard DB terms the table can be said to contain any number of rows, each row is accessed by its key. Key = A single JSON file that contains all the data associated with the key, which is stored as columns, each column having a value. Column = The name for the name/value pair to be stored or retrieved. Value = A string value, which can be anything from null, a number, a set of typed characters or even a full JSON blob.

To get started, all you need to do is create a new instance and then start storing and retrieving values. Quick Start

Using the WebAPI from the Swagger UI (Or from any language)

The Swagger UI makes using the web interface very straight forward:

  1. Use "GET /Instance/Exists" to test if we already have an instance named "Test". In swagger open the section called "Instance", and click the API for "GET /Instance/Exists". Enter "Test" as the name of the Instance and click "Try it Out!".
  2. If it exists, go to step 4.
  3. If it doesn't exist, then use the "POST /Instance" to create a "Test" instance. Note that you don't really need to create anything, when you write to the DB, that will create everything automatically.
  4. To write to the database use "POST /Value" with the instance "Test", the table "Test Table", the Key "Test Key", the column of "Test Column" and a value of "Test Value". You will find the API in the "Value" section of the API in Swagger.
  5. To read from teh databse, use "GET /Value" using "Test" as the instance, "Test Table" as the table, "Test Key" as the key and "Test Column" as the column and check that you receive the value of "Test Value" in return.

Using the DLL in your C# project

The ZarahDB_Library.dll file is all you need to add a scattered database to your C# project. Add this DLL and add a reference to it in your project. You will also need the Newtonsoft.Json NuGet package. You can get the ZarahDB_Library from or using the packet manager in Visual Studio.

  1. In Visual Studio, create a new project called "ZarahDB_HelloWorld".
  2. Add the ZarahDB_Library NuGet package from to your project.
  3. Click "Program.cs" in the Solution Explorer and replace the contents of that file with this:
using System;
using ZarahDB_Library;

namespace ZarahDB_HelloWorld
    class Program
        static void Main(string[] args)
            //Define where the data is stored
            var instance = new Uri(@"C:\ZDB");
            var table = "Test Table";
            var key = "Test Key";
            var column = "Test Column";
            var value = "Test Value";

            //Store a value in the database
            var putResult = ZarahDB.Put(instance, table, key, column, value);

            //Read a value from the database
            var getResult = ZarahDB.Get(instance, table, key, column);

The Results

Either using the API or the the code, the result is one row written to a table called "Test_Table" for the Key "Test Key". The row contains a single column called "Test Column" and has a value of "Test Value". The result will be a single JSON file in the C:/zdb/Test_Table/t/e/s/t/_ folder called test key.json. It's contents will look like the following JSON. (It won't be formatted, but it will contain the same structure and data)

    "Keys": [
            "Key": "Test Key",
            "ColumnValues": [
                    "Column": "Test Column",
                    "Value": "Test Value",
                    "PreviousValue": null,
                    "Updated": "636267579942863287"

You can now read and write data to a ZarahDB database. It really is that easy. Now you need to decide if you will store values as Name/Value pairs or as documents. The examples above used a Name/Value pair. For each column, we assume there will be a single value.

Going the Document DB route

We can also approch the value as a document, then it can contain any type of data. A common thing to store is a JSON blob. Using the Json.Net library from Newtonsoft we can simply serialize and deserialize any object type into a JSON Blob using the following code:

using System;
using Newtonsoft.Json;
using ZarahDB_Library;

namespace ConsoleApplication1
    class Program
        public class ComplexObject //This could be any variable or structure in our program, from simple to very complex.
            public string FirstName { get; set; }
            public string LastName { get; set; }
            public DateTime DateOfBirth { get; set; }

        static void Main(string[] args)
            //Define where the data is stored
            var instance = new Uri(@"C:\ZDB");
            var table = "Test Table";
            var key = "Test Key";
            var column = "Test Column";

            //Create a value that is a document (JSON in this case)
            var complexObject1 = new ComplexObject
                FirstName = "Mike",
                LastName = "Reed",
                DateOfBirth = DateTime.Parse("06/24/1962")
            var value = JsonConvert.SerializeObject(complexObject1);

            //Store a value (JSON Blob serialized from a POCO) in the database
            var putResult = ZarahDB.Put(instance, table, key, column, value);

            //Read a value from the database
            var getResult = ZarahDB.Get(instance, table, key, column);
            //Place the value in a complex object
            var complexObject2 = JsonConvert.DeserializeObject<ComplexObject>(getResult.Value);

Imagine how easy this makes storing your settings or other program data, no longer do you need to map to and from tables, think about what data goes where or many of the complexities of storing data in a typical SQL database.

Where would I use this?

Anywhere that you have a unique ID that maps to a set of data. That's not every case, but it happens quite frequetly. If you have objects in your program that you reference by ID (number, word, anything unique) then you can save the entire object into a ZarahDB and extract it again, each in a single line of code. How cool is that?

There are many uses for a ZarahDB, so how will you use it?

You can’t perform that action at this time.