Caching

Caching

As caching is an essential technology in the development of high-performance web services, Service Stack has a number of different caching options available that each share the same common client interface (ICacheClient) for the following cache providers:

• Redis - A very fast key-value store that has non-volatile persistent storage and support for rich data structures such as lists and sets.
• Memcached - The tried and tested most widely used cache provider.
• In Memory Cache - Useful for single host web services and enabling unit tests to run without needing access to a cache server.
• Azure Cache Client - For using the Azure DataCache client when your application is hosted on Azure.
• Aws Cache Client - For using with Amazon's Dynamo DB backend hosted on Amazon Web Services

To configure which cache should be used, the particular client has to be registered in the IoC container:

In-memory cache:

container.Register<ICacheClient>(new MemoryCacheClient());

Redis

container.Register<IRedisClientsManager>(c => 
    new PooledRedisClientManager("localhost:6379"));
container.Register<ICacheClient>(c => 
    (ICacheClient)c.Resolve<IRedisClientsManager>().GetCacheClient());

Memcached:

container.Register<ICacheClient>(
    new MemcachedClientCache("127.0.0.0[:11211]"); //Add Memcached hosts

Azure:

container.Register<ICacheClient>(
    new AzureCacheClient("MyAppCache"); //Add your Azure CacheName if any

Cache a response of a service

To cache a response you simply have to call ToOptimizedResultUsingCache which is an extension method existing in ServiceStack.ServiceHost.

In your service:

public class OrdersService : Service
{
    public object Get(CachedOrders request)
    {
        var cacheKey = "unique_key_for_this_request";
        return base.RequestContext.ToOptimizedResultUsingCache(base.Cache, cacheKey, () =>
        {
            //Delegate is executed if item doesn't exist in cache
        		    
            //Any response DTO returned here will be cached automatically
        });
    }
}

Tip: There exists a class named UrnId which provides helper methods to create unique keys for an object.

ToOptimizedResultUsingCache also has an overload which provides a parameter to set the timespan when the cache should be deleted (marked as expired). If now a client calls the same service method a second time and the cache expired, the provided delegate, which returns the response DTO, will be executed a second time.

var cacheKey = "some_unique_key";
//Cache should be deleted in 1h
var expireInTimespan = new TimeSpan(1, 0, 0);
return base.RequestContext.ToOptimizedResultUsingCache(base.Cache, cacheKey, expireInTimespan, ...)

Delete cached responses

If now for example an order gets updated and the order was cached before the update, the webservice will still return the same result, because the cache doesn't know that the order has been updated.

So there are two options:

• Use time based caching (and expire cache earlier)
• Cache on validility

When the cache is based on validility the caches are invalidated manually (e.g. when a user modified his profile, > clear his cache) which means you always get the latest version and you never need to hit the database again to rehydrate the cache if it hasn't changed, which will save resources.

So if the order gets updated, you should delete the cache manually:

public class CachedOrdersService : Service
{
    public object Put(CachedOrders request)
    {
        //The order gets updated...
        var cacheKey = "some_unique_key_for_order";
        return base.RequestContext.RemoveFromCache(base.Cache, cacheKey);
    }
}

If now the client calls the webservice to request the order, he'll get the latest version.

Live Example and code

A live demo of the ICacheClient is available in The ServiceStack.Northwind's example project. Here are some requests to cached services:

• /customers
• /customers/ALFKI
• /customers/ALFKI/orders

Which are simply existing web services wrapped using ICacheClient that are contained in CachedServices.cs