Tag Archives: Redis

SignalR Scaleout with Redis Backplane

Introduction

In “Getting Started with SignalR“, I provided a gentle introduction to SignalR with a few simple and practical examples. The overwhelming response showed that I’m not alone in thinking this is an awesome technology enabling real-time push notifications over the web.

Web applications often face the challenge of having to scale to handle large amounts of clients, and SignalR applications are no exception. In this example, we’ll see one way of scaling out SignalR applications using something called a backplane.

Scaleout in SignalR

SignalRScaleout

Introduction to Scaleout in SignalR” (official documentation) describes how SignalR applications can use several servers to handle increasing numbers of clients. When a server needs to push an update, it first pushes it over a message bus called a backplane. This delivers it to the other servers, which can then forward the update to their respective clients.

According to the official documentation, scaleout is supported using Azure, Redis or SQL Server as backplanes. Third-party packages exist to support other channels, such as SignalR.RabbitMq.

Scaleout Example using Redis

Introduction to Scaleout in SignalR” (official documentation) describes how to use SignalR as a backplane. To demonstrate this, I’ll build on the Chat Example code from my “Getting Started with SignalR” article.

All we need to scaleout using Redis is install the Microsoft.AspNet.SignalR.Redis NuGet package, and then set it up in the Startup class as follows:

        public void Configuration(IAppBuilder app)
        {
            GlobalHost.DependencyResolver.UseRedis("192.168.1.66", 6379, null, "SignalRChat");
            app.MapSignalR();
        }

In the code above, I am specifying the host and port of the Redis server, the password (in this case null because I don’t have one), and the name of the pub/sub channel that SignalR will use to distribute messages.

To test this, you can get a Redis server from the Redis download page. Redis releases for Windows exist and are great for testing stuff, but remember they aren’t officially supported for production environments.

Now to actually test it, I’ve set up the same scaleout-enhanced chat application on two different machines, and subscribed to the Redis pub/sub channel:

signalr-scaleout-computer1

Watching the pub/sub channel reveals what SignalR is doing under the hood. There are particular messages going through when the application initializes on each machine, and you can also see the actual data messages going through. So when you write a message in the chat, you can also see it in the pub/sub channel.

But even better than that, you’ll also see it on the client (browser) that’s hooked up to the other machine:

signalr-scaleout-computer2

The magic you need to appreciate here is that these aren’t two browsers connected to the same server; they are actually communicating with different servers on different machines. And despite that, the messages manage to reach all clients thanks to the backplane, which in this case is Redis.

Caveats

So I’ve shown how it’s really easy to scale out SignalR to multiple servers: you need to install a NuGet package and add a line of code. And I’ve actually tested it on two machines.

stash-1-244250d58073b0ed1

But that’s not really scaleout. I don’t have the resources to do large-scale testing, and only intended to show how scaleout is implemented with this article. The actual benefits of scaleout depend on the application. As the official documentation warns, the addition of a backplane incurs overhead and can become a bottleneck in some scenarios. You really need to study whether your application is a good fit for this kind of scaleout before going for it.

Setting up a Connection with StackExchange.Redis

Update 25th October 2016: Just looking to quickly set up a Redis connection? Check out the followup article. Read on for a more detailed article on the topic.

StackExchange.Redis is a pretty good .NET client for Redis. Unfortunately, it can be a little bit tricky to use, and the existing documentation is far from comprehensive.

After installing StackExchange.Redis via NuGet, a Redis connection can be obtained via a special ConnectionMultiplexer object. Working with this is already tricky in itself, and many get this wrong. For instance, check out the implementation in this answer:

public static ConnectionMultiplexer RedisConnection;
public static IDatabase RedisCacheDb;

protected void Session_Start(object sender, EventArgs e)
    {
        if (ConfigurationManager.ConnectionStrings["RedisCache"] != null)
        {
            if (RedisConnection == null || !RedisConnection.IsConnected)
            {
                RedisConnection = ConnectionMultiplexer.Connect(ConfigurationManager.ConnectionStrings["RedisCache"].ConnectionString);
            }
            RedisCacheDb = RedisConnection.GetDatabase();
        }
    }

As I pointed out in my question, this is a bad form of lazy initialization because it lacks thread safety: multiple threads may get through the checks and initialize multiple connections, resulting in connection leaks.

It is not hard to prove that this code is leaky in multithreaded environments. First, let’s set up the ConfigurationOptions with a client name so that we can identify connections coming from our program:

        private static Lazy<ConfigurationOptions> configOptions
            = new Lazy<ConfigurationOptions>(() => 
            {
                var configOptions = new ConfigurationOptions();
                configOptions.EndPoints.Add("localhost:6379");
                configOptions.ClientName = "LeakyRedisConnection";
                configOptions.ConnectTimeout = 100000;
                configOptions.SyncTimeout = 100000;
                return configOptions;
            });

Then, we provide a property with the faulty lazy initialization:

        private static ConnectionMultiplexer conn;

        private static ConnectionMultiplexer LeakyConn
        {
            get
            {
                if (conn == null || !conn.IsConnected)
                    conn = ConnectionMultiplexer.Connect(configOptions.Value);
                return conn;
            }
        }

Finally, we write some code that runs some Redis stuff in parallel:

        static void Main(string[] args)
        {
            for (int i = 0; i < 3; i++)
            {
                Task.Run(() =>
                    {
                        var db = LeakyConn.GetDatabase();
                        Console.WriteLine(i);

                        string key = "key" + i;

                        db.StringSet(key, i);
                        Thread.Sleep(10);
                        string value = db.StringGet(key);
                    }
                );
            }

            Console.WriteLine("Done");
            Console.ReadLine();
        }

When the program does its work, even with just 3 iterations, we get a total of six connections (when normally a single ConnectionMultiplexer should have at most 2 physical connections):

redis-leaky-connections

Another approach from this answer is to use an exclusive lock:

private static ConnectionMultiplexer _redis;
private static readonly Object _multiplexerLock = new Object();

private void ConnectRedis()
{
    try
    {
        _redis = ConnectionMultiplexer.Connect("...<connection string here>...");
    }
    catch (Exception ex)
    {
        //exception handling goes here
    }
}


private ConnectionMultiplexer RedisMultiplexer
{
    get
    {
        lock (_multiplexerLock)
        {
            if (_redis == null || !_redis.IsConnected)
            {
                ConnectRedis();
            }
            return _redis;
        }
    }
}

However, since Redis is often used as a cache in highly concurrent web applications, this approach essentially forces code to degrade into something sequential, and has obvious performance implications.

The correct approach to using ConnectionMultiplexer is described by this answer. It involves use of Lazy<T> for thread-safe lazy initialization (see Jon Skeet’s article on Singletons). Additionally:

  • It sets “abortConnect=false”, which means if the initial connect attempt fails, the ConnectionMultiplexer will silently retry in the background rather than throw an exception.
  • It does not check the IsConnected property, since ConnectionMultiplexer will automatically retry in the background if the connection is dropped.

With this info, we can now fix our code:

        private static Lazy<ConfigurationOptions> configOptions
            = new Lazy<ConfigurationOptions>(() => 
            {
                var configOptions = new ConfigurationOptions();
                configOptions.EndPoints.Add("localhost:6379");
                configOptions.ClientName = "SafeRedisConnection";
                configOptions.ConnectTimeout = 100000;
                configOptions.SyncTimeout = 100000;
                configOptions.AbortOnConnectFail = false;
                return configOptions;
            });

        private static Lazy<ConnectionMultiplexer> conn
            = new Lazy<ConnectionMultiplexer>(
                () => ConnectionMultiplexer.Connect(configOptions.Value));

        private static ConnectionMultiplexer SafeConn
        {
            get
            {
                return conn.Value;
            }
        }

        static void Main(string[] args)
        {
            for (int i = 0; i < 3; i++)
            {
                Task.Run(() =>
                    {
                        var db = SafeConn.GetDatabase();
                        Console.WriteLine(i);

                        string key = "key" + i;

                        db.StringSet(key, i);
                        Thread.Sleep(10);
                        string value = db.StringGet(key);
                    }
                );
            }

            Console.WriteLine("Done");
            Console.ReadLine();
        }

If you run this, you’ll find that there are now only two physical connections generated by the application, which is normal.

redis-safe-connections

Double Buffering in Redis

This article deals with a specific situation. You have a program that is continuously building a data structure from scratch. This data structure takes long (i.e. several seconds) to build, but is continuously accessed by clients. While it is being built, clients will end up retrieving partial data. This article explains how a well-known technique from computer graphics can be used in Redis to address this problem. The full source code is available at BitBucket.

The Problem

Like just about anything, this scenario is best demonstrated with an example. So we’ll first start with some code to set up a client connection to Redis:

        private static Lazy<ConnectionMultiplexer> LazyConnection
            = new Lazy<ConnectionMultiplexer>(() =>
            {
                var config = new ConfigurationOptions();
                config.EndPoints.Add("localhost:6379");
                config.AbortOnConnectFail = false;

                return ConnectionMultiplexer.Connect(config);
            }
        );

        public static ConnectionMultiplexer Connection
        {
            get
            {
                return LazyConnection.Value;
            }
        }

The code above is based on the feedback I received from this StackOverflow question and is the proper way to get a connection set up using StackExchange.Redis. You’ll need to install the StackExchange.Redis package via NuGet before you can use this code.

Our actual program code is extremely simple. We establish a connection (via the property above) and then continuously call a method that populates our data structure:

        static void Main(string[] args)
        {
            var connection = Connection;
            var database = connection.GetDatabase(0);

            while (true)
                PopulateAwesomeCompanies(database);
        }

We use this infinite loop because the nature of the data may be dynamic, for example items may be removed, and it may be hard to keep track of them. So the data structure is simply rebuilt regularly.

For this simple example, our data structure is going to be a Redis sorted set that stores the most awesome companies to work for, sorted by awesomeness. Here is the data we’re going to be using:

        private static string[] companies = new string[]
        {
            "Google",
            "Apple",
            "Amazon",
            "GFI",
            "Blizzard",
            "IBM"
        };

        private static int[] companyScores = new int[] { 95, 15, 80, 0, 100, 56 };

The method that builds the data structure is also very simple. It starts off by erasing the existing data, and then for each company, it calculates their awesomeness (simulated by a simple delay) and adds them to the sorted set:

        public static void PopulateAwesomeCompanies(IDatabase database)
        {
            var key = "awesomecompanies";

            Console.WriteLine("Starting afresh...");

            database.KeyDelete(key); // start with a fresh sorted set

            for (int i = 0; i < 6; i++)
            {
                Console.WriteLine("Calculating {0}", companies[i]);

                // simulate expensive computation
                Thread.Sleep(5000);

                // add company with respective score
                database.SortedSetAdd(key, companies[i], companyScores[i]);
            }
        }

The problem with this approach is that anytime a client accesses this data structure while it’s still being built, they will get partial data:

redis-doublebuffering-partialdata

That sucks, because we want the data to be fresh, but we also want it to be complete when it is handed to clients.

Double Buffering in Computer Graphics

There is a technique in computer graphics called double buffering which solves a similar problem. Here’s some background from Wikipedia:

“In computer graphics, double buffering is a technique for drawing graphics that shows no (or less) flicker, tearing, and other artifacts.

“It is difficult for a program to draw a display so that pixels do not change more than once. For instance to update a page of text it is much easier to clear the entire page and then draw the letters than to somehow erase all the pixels that are not in both the old and new letters. However, this intermediate image is seen by the user as flickering. In addition computer monitors constantly redraw the visible video page (at around 60 times a second), so even a perfect update may be visible momentarily as a horizontal divider between the “new” image and the un-redrawn “old” image, known as tearing.

“A software implementation of double buffering has all drawing operations store their results in some region of system RAM; any such region is often called a “back buffer”. When all drawing operations are considered complete, the whole region (or only the changed portion) is copied into the video RAM (the “front buffer”); this copying is usually synchronized with the monitor’s raster beam in order to avoid tearing. Double buffering necessarily requires more memory and CPU time than single buffering because of the system memory allocated for the back buffer, the time for the copy operation, and the time waiting for synchronization.”

In short, double buffering involves the following steps:

  1. Draw the next frame (screen image) in a temporary location in memory that isn’t directly visible to the user (the back buffer)
  2. Swap the front buffer (the actual image on the screen) with the back buffer

That way, the change between the old image and the new one is instantaneous and the user will never notice the difference.

Double Buffering in Redis

We can apply a similar technique in Redis as follows:

  1. Build the new sorted set using a separate, temporary key
  2. Rename the temporary key to the existing key (implicitly deletes the temporary key)

This is how it works out in practice:

        public static void PopulateAwesomeCompanies(IDatabase database)
        {
            var tempKey = "awesomecompaniestemp";
            var key = "awesomecompanies";

            Console.WriteLine("Starting afresh...");

            for (int i = 0; i < 6; i++)
            {
                Console.WriteLine("Calculating {0}", companies[i]);

                // simulate expensive computation
                Thread.Sleep(5000);

                // add company with respective score
                database.SortedSetAdd(tempKey, companies[i], companyScores[i]);
            }

            // replace actual data in key with fresh data from tempKey
            database.KeyRename(tempKey, key);
        }

Here, we aren’t deleting any data structure before we begin, as we did before. Instead, we build our fresh sorted set using the temp key, and then copy it over to the actual key (the one that clients will be requesting). We do this using the Redis RENAME command, which copies the data from the temp key to the actual key, and implicitly also deletes the temp key. Since Redis is single-threaded, this operation is atomic and there is no race condition risk while the rename operation is happening.

The result of this is that the client will always have access to the latest complete data, even while the data structure is being rebuilt:

redis-doublebufferingfulldata

On Redis Desktop Manager and Redis Keys

Update 11th August 2015: As from version 0.8.0 beta 1, Redis Desktop Manager now supports the SCAN command (rather than KEYS) for Redis 2.8 onwards. Although this limits the applicability of this article to older servers, I am more than happy that this shortcoming has been addressed. The original article below remains both for historical purposes and to warn developers of incorrect use of the KEYS command.

There’s a tool called Redis Desktop Manager which can sometimes be useful to inspect the keys in a Redis database. Indeed one of its features is that it presents a treeview showing a structured representation of the keys in the database:

redis-desktop-manager-treeview

But how is this treeview built? That’s easy to find out, by using the Redis MONITOR command to see the incoming commands:

redis-desktop-manager-keys

 

The first two commands are executed when Redis Desktop Manager connects to the Redis server, and the other two are executed when the database is expanded in the treeview, revealing the keys in that database.

You’ll notice that the last command is a KEYS command, which with its wildcard argument (*) is effectively retrieving every key in the database. We can see an example of what this gives us by running the KEYS command ourselves:

redis-desktop-manager-keys-response

Now, in this case I only have a handful of keys in my Redis database, but it’s pretty normal for real Redis databases to have very large numbers of keys. Retrieving all that data places a large burden on the Redis server, which due to its single-threaded nature will not be able to serve other requests while it is stuck retrieving every key in the database.

In fact, the KEYS command documentation particularly warns against its use in production environments:

“Warning: consider KEYS as a command that should only be used in production environments with extreme care. It may ruin performance when it is executed against large databases. This command is intended for debugging and special operations, such as changing your keyspace layout. Don’t use KEYS in your regular application code. If you’re looking for a way to find keys in a subset of your keyspace, consider using SCAN or sets.”

I understand why Redis Desktop Manager uses the KEYS command: it needs to retrieve all the keys in the database in order to determine how the tree structure will be displayed (since each delimited part of the key is rendered as a node). That’s the whole point of having a treeview.

However, what seems to be a useful feature can actually be very dangerous, especially when used on production servers. So do take care when using Redis Desktop Manager in such environments.

My recommendation to developers using Redis is to keep good documentation of your keys, so you won’t need any Redis command to tell you what keys are in the database. That’s not what Redis is for.

But if you do ever need to inspect your Redis keys on the server, at least follow the advice in the documentation and use SCAN instead. While this may still be expensive to retrieve the entire set of keys, it can be done in small batches, thus allowing other requests to be serviced in between iterations.

A Redis Analogy in .NET

In this article, we’re going to create a simple Console Application that feels a little bit like Redis. The intention is simply to illustrate what you can do with Redis, not to create anything serious. For this reason, the code will be greatly simplified: we’ll have only a subset of Redis commands, no multithreading, no client/server code, (almost) no parameter validation, and no architecture whatsoever.

Introduction

Redis is a key-value store. You can think of it as a dictionary. In its simplest sense, it could be represented like this:

var store = new Dictionary<string, string>();

However, values in Redis can be more than just strings. In fact, Redis currently supports five main data types:

  • Strings
  • Lists
  • Hashes
  • Sets
  • Sorted Sets

Since in .NET these data types don’t share a common interface, we will be representing our values as plain objects:

var store = new Dictionary<string, object>();

This will result in some unfortunate type checking code, but this is necessary in this case. Maintaining a separate dictionary per type is not feasible since keys must be globally unique (i.e. you can’t use the same key for a list and a hash).

Note also how the use of a dictionary as an in-memory store means we aren’t persisting anything to disk. This might seem to be an oversimplification, but Redis’ default configuration is actually to store data only in memory. It does support persistence if you need it, but we won’t go there.

Redis supports a simple protocol which feels a bit like issuing commands in a command line if you use something like IMAPTalk. Thus, in our Console Application, we’ll use the following code to simulate Redis’ command processing logic:

            string input = string.Empty;

            while (true)
            {
                input = Console.ReadLine();

                if (!string.IsNullOrEmpty(input))
                {
                    var tokens = input.Split(); // split the input into words or tokens

                    if (tokens.Length >= 1)
                    {
                        string command = tokens.First().ToLowerInvariant();

                        try
                        {
                            switch(command)
                            {
                                    // TODO command handling code goes here
                                default:
                                    Console.WriteLine("Unrecognized command.");
                                    break;
                            }
                        }
                        catch (Exception)
                        {
                            Console.WriteLine("Error!");
                        }
                    }
                }
                else
                    Console.WriteLine("Invalid command.");
            }

The above code just accepts input, checks what the command is, and takes action accordingly. In the next sections, we’ll replace that TODO with code to actually handle some of the commands.

Strings

redis-imaptalk-strings

The simplest data type in Redis is the string. You can easily assign a value to a key using the SET command, and then retrieve that value using the GET command, as shown above. The SET command is just a matter of assigning the value in the dictionary:

                                case "set":
                                    {
                                        string key = tokens[1];
                                        string value = tokens[2];
                                        store[key] = value;
                                        Console.WriteLine("Done.");
                                    }
                                    break;

The GET command similarly involves basic retrieval from the dictionary, but we also need to cater for non-existent keys, and values that are not strings:

                                case "get":
                                    {
                                        string key = tokens[1];
                                        if (store.ContainsKey(key))
                                        {
                                            var value = store[key] as string;
                                            if (value != null)
                                                Console.WriteLine(value);
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

We can also support key removal using the DEL command. Again, this is a simple operation against the dictionary:

                                case "del":
                                    {
                                        string key = tokens[1];
                                        store.Remove(key);
                                        Console.WriteLine("Done.");
                                    }
                                    break;

We can now test our three basic string commands:

redis-analogy-strings

Lists

redis-imaptalk-lists

Redis also supports Lists. You can use the LPUSH and RPUSH commands to add an item to the beginning or end of a list respectively – allowing you to easily use the list as a stack or a queue as well. These operations also automatically create the list if it doesn’t already exist.

Here’s how we can implement LPUSH:

                                case "lpush":
                                    {
                                        string key = tokens[1];
                                        string value = tokens[2];
                                        List<string> list = null;

                                        // create/get list

                                        if (!store.ContainsKey(key))
                                        {
                                            list = new List<string>();
                                            store[key] = list;
                                        }
                                        else
                                            list = store[key] as List<string>;

                                        // insert new value in list

                                        if (list != null)
                                        {
                                            list.Insert(0, value);
                                            Console.WriteLine("Done");
                                        }
                                        else
                                            Console.WriteLine("Invalid data type!");
                                    }
                                    break;

The first thing we do here is retrieve the list we’re going to work with. If it doesn’t exist, we create it. We can then proceed to insert the new item into the list – in this case using List<T>.Insert() to put the item at the beginning of the list.

Note that in Redis, LPUSH actually supports insertion of multiple items with the same command. I’m not doing that here to keep things as simple as possible.

RPUSH is pretty much the same thing, except that you call List<T>.Add() instead of .Insert(), so that the new item ends up at the end of the list.

To remove items from the list, we’ll implement LPOP and RPOP, which remove items from the beginning and end of the list respectively. Here’s LPOP:

                                case "lpop":
                                    {
                                        string key = tokens[1];
                                        
                                        if (store.ContainsKey(key))
                                        {
                                            var list = store[key] as List<string>;
                                            if (list != null)
                                            {
                                                if (list.Count > 0)
                                                {
                                                    Console.WriteLine(list.First());
                                                    list.RemoveAt(0);
                                                }
                                                else
                                                    Console.WriteLine("Empty!");
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

Aside from all the validity checks, all we’re doing here is returning the first item in the list, and then removing it. If there are no items in the list, we simply return “Empty!”.

For RPOP, the only difference is that we retrieve and remove the last item instead:

                                                    Console.WriteLine(list.Last());
                                                    list.RemoveAt(list.Count - 1);

Finally, we need something that can retrieve items in the list (without removing them). For this we have LRANGE, which retrieves a specified range of items (e.g. item #2 till item #4) from the beginning of the list. Indices may be out of range without causing errors; the indices that actually exist will be returned.

Here is the implementation for LRANGE:

                                case "lrange":
                                    {
                                        string key = tokens[1];
                                        int start = Convert.ToInt32(tokens[2]);
                                        int stop = Convert.ToInt32(tokens[3]);

                                        if (start > stop)
                                            Console.WriteLine("Empty!");
                                        else if (store.ContainsKey(key))
                                        {
                                            var list = store[key] as List<string>;
                                            if (list != null)
                                            {
                                                if (start < 0)
                                                    start = 0;

                                                if (stop > list.Count - 1)
                                                    stop = list.Count - 1;

                                                var items = list.GetRange(start, stop - start + 1);
                                                if (items.Any())
                                                {
                                                    foreach(var item in items)
                                                        Console.WriteLine(item);
                                                }
                                                else
                                                    Console.WriteLine("Empty!");
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

We can now test our list functionality:

redis-analogy-lists

Hashes

redis-imaptalk-hashes

Hashes are like dictionaries in themselves. Rather than mapping a value directly to a key, hashes map values onto a field of a key. This allows you to represent objects with a number of attributes (e.g. a customer having separate values for Name, Age, etc).

With HSET, we can assign a value to a key-field, as shown in the screenshot above. Like with Lists, this operation creates the Hash if it doesn’t already exist. Here’s the HSET implementation:

                                case "hset":
                                    {
                                        string key = tokens[1];
                                        string field = tokens[2];
                                        string value = tokens[3];
                                        Dictionary<string, string> hash = null;

                                        // create/get hash

                                        if (!store.ContainsKey(key))
                                        {
                                            hash = new Dictionary<string, string>();
                                            store[key] = hash;
                                        }
                                        else
                                            hash = store[key] as Dictionary<string, string>;

                                        // set field in hash

                                        if (hash != null)
                                        {
                                            hash[field] = value;
                                            Console.WriteLine("Done");
                                        }
                                        else
                                            Console.WriteLine("Invalid data type!");
                                    }
                                    break;

The code is very similar to that of LPUSH, with the difference that the key now maps to a dictionary (the hash), and the value is assigned to a field on that hash. Think of it as: key -> field -> value.

After using HSET to set a value, we can retrieve it with HGET:

                                case "hget":
                                    {
                                        string key = tokens[1];
                                        string field = tokens[2];

                                        if (store.ContainsKey(key))
                                        {
                                            var hash = store[key] as Dictionary<string, string>;

                                            if (hash != null)
                                            {
                                                if (hash.ContainsKey(field))
                                                    Console.WriteLine(hash[field]);
                                                else
                                                    Console.WriteLine("Field not found!");
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

HKEYS can be used to retrieve all fields of a hash, given the key:

                                case "hkeys":
                                    {
                                        string key = tokens[1];

                                        if (store.ContainsKey(key))
                                        {
                                            var hash = store[key] as Dictionary<string, string>;

                                            if (hash != null)
                                            {
                                                foreach (var field in hash.Keys)
                                                    Console.WriteLine(field);
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

HGETALL, on the other hand, gets all fields of a hash and their corresponding values:

                                case "hgetall":
                                    {
                                        string key = tokens[1];

                                        if (store.ContainsKey(key))
                                        {
                                            var hash = store[key] as Dictionary<string, string>;

                                            if (hash != null)
                                            {
                                                foreach (var kvp in hash)
                                                {
                                                    Console.WriteLine(kvp.Key);
                                                    Console.WriteLine(kvp.Value);
                                                }
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

Let’s test out our hash functionality:

redis-analogy-hashes

Sets

redis-imaptalk-sets

Sets in Redis are the mathematical kind: all items in a set are distinct, and multiple sets may be combined via standard set operations (union, intersection and difference).

Use SADD to add one or more items to a set. If the set does not exist, it will be created automatically. Here’s the SADD implementation:

                                case "sadd":
                                    {
                                        string key = tokens[1];
                                        var members = tokens.Skip(2); // sadd key member [member ...]
                                        HashSet<string> set = null;

                                        // create/get set

                                        if (!store.ContainsKey(key))
                                        {
                                            set = new HashSet<string>();
                                            store[key] = set;
                                        }
                                        else
                                            set = store[key] as HashSet<string>;

                                        // add member to set

                                        if (set != null)
                                        {
                                            foreach (var member in members)
                                                set.Add(member);

                                            Console.WriteLine("Done");
                                        }
                                        else
                                            Console.WriteLine("Invalid data type!");
                                    }
                                    break;

SMEMBERS gives you all the members of the set:

                                case "smembers":
                                    {
                                        string key = tokens[1];

                                        if (store.ContainsKey(key))
                                        {
                                            var set = store[key] as HashSet<string>;

                                            if (set != null)
                                            {
                                                foreach (var member in set)
                                                    Console.WriteLine(member);
                                            }
                                            else
                                                Console.WriteLine("Invalid data type!");
                                        }
                                        else
                                            Console.WriteLine("Key not found!");
                                    }
                                    break;

Standard set operations (union, intersection and difference) require two sets. If either doesn’t exist, these operations will return an empty set rather than giving any kind of error. In Redis, these operations may work on multiple sets at once; but in these examples we’re going to limit the scenario to two sets at a time for the sake of simplicity.

Set union (SUNION) returns all distinct items in the specified sets:

                                case "sunion":
                                    {
                                        List<HashSet<string>> sets = new List<HashSet<string>>();

                                        var key1 = tokens[1];
                                        var key2 = tokens[2];

                                        // let's assume the keys exist and are the correct type

                                        var set1 = store[key1] as HashSet<string>;
                                        var set2 = store[key2] as HashSet<string>;

                                        // get union

                                        var union = set1.Union(set2);
                                        foreach(var member in union)
                                            Console.WriteLine(member);
                                    }
                                    break;

Set intersection (SINTER) returns those items which are common to both sets:

                                case "sinter":
                                    {
                                        List<HashSet<string>> sets = new List<HashSet<string>>();

                                        var key1 = tokens[1];
                                        var key2 = tokens[2];

                                        // let's assume the keys exist and are the correct type

                                        var set1 = store[key1] as HashSet<string>;
                                        var set2 = store[key2] as HashSet<string>;

                                        // get intersection

                                        var union = set1.Intersect(set2);
                                        foreach (var member in union)
                                            Console.WriteLine(member);
                                    }
                                    break;

Finally, set difference (SDIFF) returns those items which are in the first set but which are not in the second:

                                case "sdiff":
                                    {
                                        List<HashSet<string>> sets = new List<HashSet<string>>();

                                        var key1 = tokens[1];
                                        var key2 = tokens[2];

                                        // let's assume the keys exist and are the correct type

                                        var set1 = store[key1] as HashSet<string>;
                                        var set2 = store[key2] as HashSet<string>;

                                        // get intersection

                                        var union = set1.Except(set2);
                                        foreach (var member in union)
                                            Console.WriteLine(member);
                                    }
                                    break;

Let’s test that out:

redis-analogy-sets

Sorted Sets

redis-imaptalk-sortedsets

Sorted Sets in Redis are similar to sets, but their members are sorted by a score. This may be used for any kind of ordering from the latest 5 comments to the top 5 posts with most likes.

We can represent a Sorted Set similarly to a hash, using the scheme key -> score -> value. However, there are two main differences from a hash:

  • For each score, there may be multiple values; these are sorted lexicographically.
  • The scores are ordered (sorted).

The implementation of a Sorted Set in .NET is not as trivial as the other data types, and requires a combination of collections. While a simple representation could be made by combining a SortedDictionary (score -> values) with a SortedSet (distinct and sorted values), this would not allow all Sorted Set commands to be supported (e.g. ZRANK can find the index of a given value, requiring a reverse lookup).

Sorted Sets are thus beyond the scope of this article, which is intended to provide a simple mapping between Redis data types and .NET collections.

Summary

This article explained how the Redis data types work, and showed how some of their operations may be implemented using standard .NET collections. A typical mapping could be:

  • Strings – strings
  • Lists – Lists
  • Hashes – Dictionaries or ConcurrentDictionaries
  • Sets – HashSets
  • Sorted Sets – a custom data structure

Source Code

The source code for this article is available on BitBucket.