[转] .net core Session , Working with a distributed cache

本文转自：https://docs.microsoft.com/en-us/aspnet/core/performance/caching/distributed

By Steve Smith+

Distributed caches can improve the performance and scalability of ASP.NET Core apps, especially when hosted in a cloud or server farm environment. This article explains how to work with ASP.NET Core's built-in distributed cache abstractions and implementations.+

View or download sample code+

What is a Distributed Cache

A distributed cache is shared by multiple app servers (see Caching Basics). The information in the cache is not stored in the memory of individual web servers, and the cached data is available to all of the app's servers. This provides several advantages:+

Cached data is coherent on all web servers. Users don't see different results depending on which web server handles their request

Cached data survives web server restarts and deployments. Individual web servers can be removed or added without impacting the cache.

The source data store has fewer requests made to it (than with multiple in-memory caches or no cache at all).

+

Note
If using a SQL Server Distributed Cache, some of these advantages are only true if a separate database instance is used for the cache than for the app's source data.+

Like any cache, a distributed cache can dramatically improve an app's responsiveness, since typically data can be retrieved from the cache much faster than from a relational database (or web service).+

Cache configuration is implementation specific. This article describes how to configure both Redis and SQL Server distributed caches. Regardless of which implementation is selected, the app interacts with the cache using a common

IDistributedCache

interface.+

The IDistributedCache Interface

The

IDistributedCache

interface includes synchronous and asynchronous methods. The interface allows items to be added, retrieved, and removed from the distributed cache implementation. The

IDistributedCache

interface includes the following methods:+

Get, GetAsync+

Takes a string key and retrieves a cached item as a

byte[]

if found in the cache.+

Set, SetAsync+

Adds an item (as

byte[]

) to the cache using a string key.+

Refresh, RefreshAsync+

Refreshes an item in the cache based on its key, resetting its sliding expiration timeout (if any).+

Remove, RemoveAsync+

Removes a cache entry based on its key.+

To use the

IDistributedCache

interface:+

Specify the dependencies needed in project.json.

Configure the specific implementation of

IDistributedCache

in your

Startup

class's

ConfigureServices

method, and add it to the container there.

From the app's `Middleware or MVC controller classes, request an instance of

IDistributedCache

from the constructor. The instance will be provided by Dependency Injection (DI).

+

Note
There is no need to use a Singleton or Scoped lifetime for

IDistributedCache

instances (at least for the built-in implementations). You can also create an instance wherever you might need one (instead of using Dependency Injection), but this can make your code harder to test, and violates the Explicit Dependencies Principle.+

The following example shows how to use an instance of

IDistributedCache

in a simple middleware component:+

Copy
C#

using Microsoft.AspNetCore.Builder;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Caching.Distributed;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

namespace DistCacheSample
{
public class StartTimeHeader
{
private readonly RequestDelegate _next;
private readonly IDistributedCache _cache;

public StartTimeHeader(RequestDelegate next,
IDistributedCache cache)
{
_next = next;
_cache = cache;
}

public async Task Invoke(HttpContext httpContext)
{
string startTimeString = "Not found.";
var value = await _cache.GetAsync("lastServerStartTime");
if (value != null)
{
startTimeString = Encoding.UTF8.GetString(value);
}

httpContext.Response.Headers.Append("Last-Server-Start-Time", startTimeString);

await _next.Invoke(httpContext);
}
}

// Extension method used to add the middleware to the HTTP request pipeline.
public static class StartTimeHeaderExtensions
{
public static IApplicationBuilder UseStartTimeHeader(this IApplicationBuilder builder)
{
return builder.UseMiddleware<StartTimeHeader>();
}
}
}

In the code above, the cached value is read, but never written. In this sample, the value is only set when a server starts up, and doesn't change. In a multi-server scenario, the most recent server to start will overwrite any previous values that were set by other servers. The

Get

and

Set

methods use the

byte[]

type. Therefore, the string value must be converted using

Encoding.UTF8.GetString

(for

Get

) and

Encoding.UTF8.GetBytes

(for

Set

).+

The following code from Startup.cs shows the value being set:+

Copy
C#

public void Configure(IApplicationBuilder app,
IDistributedCache cache)
{
var serverStartTimeString = DateTime.Now.ToString();
byte[] val = Encoding.UTF8.GetBytes(serverStartTimeString);
cache.Set("lastServerStartTime", val);

app.UseStartTimeHeader();

Note
Since

IDistributedCache

is configured in the

ConfigureServices

method, it is available to the

Configure

method as a parameter. Adding it as a parameter will allow the configured instance to be provided through DI.+

Using a Redis Distributed Cache

Redis is an open source in-memory data store, which is often used as a distributed cache. You can use it locally, and you can configure an Azure Redis Cache for your Azure-hosted ASP.NET Core apps. Your ASP.NET Core app configures the cache implementation using a

RedisDistributedCache

instance.+

You configure the Redis implementation in

ConfigureServices

and access it in your app code by requesting an instance of

IDistributedCache

(see the code above).+

In the sample code, a

RedisCache

implementation is used when the server is configured for a

Staging

environment. Thus the

ConfigureStagingServices

method configures the

RedisCache

:+

Copy
C#

/// <summary>
/// Use Redis Cache in Staging
/// </summary>
/// <param name="services"></param>
public void ConfigureStagingServices(IServiceCollection services)
{

services.AddDistributedRedisCache(options =>
{
options.Configuration = "localhost";
options.InstanceName = "SampleInstance";
});
}

Note
To install Redis on your local machine, install the chocolatey package http://chocolatey.org/packages/redis-64/ and run

redis-server

from a command prompt.+

Using a SQL Server Distributed Cache

The SqlServerCache implementation allows the distributed cache to use a SQL Server database as its backing store. To create SQL Server table you can use sql-cache tool, the tool creates a table with the name and schema you specify.+

To use sql-cache tool add SqlConfig.Tools to the tools section of the project.json file and run dotnet restore.+

Copy
C#

"tools": {
"Microsoft.AspNetCore.Server.IISIntegration.Tools": {
"version": "1.0.0-preview2-final",
"imports": "portable-net45+win8+dnxcore50"
},
"Microsoft.Extensions.Caching.SqlConfig.Tools": "1.0.0-preview2-final"
},

Test SqlConfig.Tools by running the following command+

Copy
none

C:\DistCacheSample\src\DistCacheSample>dotnet sql-cache create --help

sql-cache tool will display usage, options and command help, now you can create tables into sql server, running "sql-cache create" command :+

Copy
none

C:\DistCacheSample\src\DistCacheSample>dotnet sql-cache create "Data Source=(localdb)\v11.0;Initial Catalog=DistCache;Integrated Security=True;" dbo TestCache
info: Microsoft.Extensions.Caching.SqlConfig.Tools.Program[0]
Table and index were created successfully.

The created table have the following schema:+


+

Like all cache implementations, your app should get and set cache values using an instance of

IDistributedCache

, not a

SqlServerCache

. The sample implements

SqlServerCache

in the

Production

environment (so it is configured in

ConfigureProductionServices

).+

Copy
C#

/// Use SQL Server Cache in Production
/// </summary>
/// <param name="services"></param>
public void ConfigureProductionServices(IServiceCollection services)
{

services.AddDistributedSqlServerCache(options =>
{
options.ConnectionString = @"Data Source=(localdb)\v11.0;Initial Catalog=DistCache;Integrated Security=True;";
options.SchemaName = "dbo";
options.TableName = "TestCache";
});

}

Note
The

ConnectionString

(and optionally,

SchemaName

and

TableName

) should typically be stored outside of source control (such as UserSecrets), as they may contain credentials.+

Recommendations

When deciding which implementation of

IDistributedCache

is right for your app, choose between Redis and SQL Server based on your existing infrastructure and environment, your performance requirements, and your team's experience. If your team is more comfortable working with Redis, it's an excellent choice. If your team prefers SQL Server, you can be confident in that implementation as well. Note that A traditional caching solution stores data in-memory which allows for fast retrieval of data. You should store commonly used data in a cache and store the entire data in a backend persistent store such as SQL Server or Azure Storage. Redis Cache is a caching solution which gives you high throughput and low latency as compared to SQL Cache.+

Additional resources:+

In memory caching

Redis Cache on Azure

SQL Database on Azure