关于缓存的设计
1、什么情况下用缓存

缓存是提高应用程序性能的最好方法之一。运用缓存可以优化数据查询，避免不必要的网络数据回传，和避免执行不必要的完全相同的数据处理逻辑。在实现缓存的时候我们要确定什么时候装入缓存数据。用异步装入缓存或用批处理方式来避免出现客户端数据延迟。
一般来说在一定时间内请求了相同的业务逻辑而没有变更的话，可以采用缓存来设计。数据请求频繁的的请求不适合采用缓存，如论坛的回复，但是论坛的主题是可以采用缓存设计的。

2、缓存设计的步骤
确定缓存数据结构：即设计中哪些数据用到了缓存，设计这些数据的缓存结构
确定缓存什么数据
确定缓存过期规则和清理
确定如何装入缓存数据

3、示例 Community Server的缓存类

复制代码 代码如下:
using System;
  using System.Collections;
  using System.Text.RegularExpressions;
  using System.Web;
  using System.Web.Caching;

  namespace Larry.Cache
  {
      /// <summary>
     /// 缓存类 Community Server的缓存类
     /// </summary>
     public class BaseCache
     {
         /// <summary>
         /// CacheDependency 说明
         /// 如果您向 Cache 中添加某个具有依赖项的项，当依赖项更改时，
         /// 该项将自动从 Cache 中删除。例如，假设您向 Cache 中添加某项，
         /// 并使其依赖于文件名数组。当该数组中的某个文件更改时，
         /// 与该数组关联的项将从缓存中删除。
         /// [C#]
         /// Insert the cache item.
         /// CacheDependency dep = new CacheDependency(fileName, dt);
         /// cache.Insert("key", "value", dep);
         /// </summary>
         public static readonly int DayFactor = ;
         public static readonly int HourFactor = ;
         public static readonly int MinuteFactor = ;
         public static readonly double SecondFactor = 0.;

         private static readonly System.Web.Caching.Cache _cache;

         private static int Factor = ;

         /// <summary>
         /// 单件模式
         /// </summary>
         static BaseCache()
         {
             HttpContext context = HttpContext.Current;
             if (context != null)
             {
                 _cache = context.Cache;
             }
             else
             {
                 _cache = HttpRuntime.Cache;
             }
         }

         /// <summary>
         /// 一次性清除所有缓存
         /// </summary>
         public static void Clear()
         {
             IDictionaryEnumerator CacheEnum = _cache.GetEnumerator();
             ArrayList al = new ArrayList();
             while (CacheEnum.MoveNext()) //逐个清除
             {
                 al.Add(CacheEnum.Key);
             }

             foreach (string key in al)
             {
                 _cache.Remove(key);
             }

         }

 

         public static void RemoveByPattern(string pattern)
         {
             IDictionaryEnumerator CacheEnum = _cache.GetEnumerator();
             Regex regex = new Regex(pattern, RegexOptions.IgnoreCase | RegexOptions.Singleline | RegexOptions.Compiled);
             while (CacheEnum.MoveNext())
             {
                 if (regex.IsMatch(CacheEnum.Key.ToString()))
                     _cache.Remove(CacheEnum.Key.ToString());
             }
         }

         /// <summary>
         /// 清除特定的缓存
         /// </summary>
         /// <param name="key"></param>
         public static void Remove(string key)
         {
             _cache.Remove(key);
         }

         /// <summary>
         /// 缓存OBJECT.
         /// </summary>
         /// <param name="key"></param>
         /// <param name="obj"></param>
         public static void Insert(string key, object obj)
         {
             Insert(key, obj, null, );
         }

        /// <summary>
        /// 缓存obj 并建立依赖项
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="dep"></param>
        public static void Insert(string key, object obj, CacheDependency dep)
        {
            Insert(key, obj, dep, MinuteFactor * );
        }

        /// <summary>
        /// 按秒缓存对象
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="seconds"></param>
        public static void Insert(string key, object obj, int seconds)
        {
            Insert(key, obj, null, seconds);
        }

        /// <summary>
        /// 按秒缓存对象 并存储优先级
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="seconds"></param>
        /// <param name="priority"></param>
        public static void Insert(string key, object obj, int seconds, CacheItemPriority priority)
        {
            Insert(key, obj, null, seconds, priority);
        }

        /// <summary>
        /// 按秒缓存对象 并建立依赖项
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="dep"></param>
        /// <param name="seconds"></param>
        public static void Insert(string key, object obj, CacheDependency dep, int seconds)
        {
            Insert(key, obj, dep, seconds, CacheItemPriority.Normal);
        }

        /// <summary>
        /// 按秒缓存对象 并建立具有优先级的依赖项
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="dep"></param>
        /// <param name="seconds"></param>
        /// <param name="priority"></param>
        public static void Insert(string key, object obj, CacheDependency dep, int seconds, CacheItemPriority priority)
        {
            if (obj != null)
            {
                _cache.Insert(key, obj, dep, DateTime.Now.AddSeconds(Factor * seconds), TimeSpan.Zero, priority, null);
            }

        }

        public static void MicroInsert(string key, object obj, int secondFactor)
        {
            if (obj != null)
            {
                _cache.Insert(key, obj, null, DateTime.Now.AddSeconds(Factor * secondFactor), TimeSpan.Zero);
            }
        }

        /// <summary>
        /// 最大时间缓存
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        public static void Max(string key, object obj)
        {
            Max(key, obj, null);
        }

        /// <summary>
        /// 具有依赖项的最大时间缓存
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        /// <param name="dep"></param>
        public static void Max(string key, object obj, CacheDependency dep)
        {
            if (obj != null)
            {
                _cache.Insert(key, obj, dep, DateTime.MaxValue, TimeSpan.Zero, CacheItemPriority.AboveNormal, null);
            }
        }

        /// <summary>
        /// Insert an item into the cache for the Maximum allowed time
        /// </summary>
        /// <param name="key"></param>
        /// <param name="obj"></param>
        public static void Permanent(string key, object obj)
        {
            Permanent(key, obj, null);
        }

        public static void Permanent(string key, object obj, CacheDependency dep)
        {
            if (obj != null)
            {
                _cache.Insert(key, obj, dep, DateTime.MaxValue, TimeSpan.Zero, CacheItemPriority.NotRemovable, null);
            }
        }

        public static object Get(string key)
        {
            return _cache[key];
        }

        /// <summary>
        /// Return int of seconds * SecondFactor
        /// </summary>
        public static int SecondFactorCalculate(int seconds)
        {
            // Insert method below takes integer seconds, so we have to round any fractional values
            return Convert.ToInt(Math.Round((double)seconds * SecondFactor));
        }
    }
}

其实这个类就是一个单件模式的设计 和缓存的公共操作方法，其中CacheDependency表示建立缓存依赖项，CacheItemPriority表示缓存的优先级。S使用如下

复制代码 代码如下:
 public static CardShop.Model.Systems GetConfig()
     {
         const string cacheKey = "WebConfig";
         CardShop.Model.Systems sampleCacheTable = Larry.Cache.BaseCache.Get(cacheKey) as CardShop.Model.Systems;
         if (sampleCacheTable == null)
         {
                                   OprationCheck.Message("第一次加载使用缓存");
             sampleCacheTable = model;
             Larry.Cache.BaseCache.Insert(cacheKey, sampleCacheTable, 24 * Larry.Cache.BaseCache.MinuteFactor);
        }
        else
        {
            OprationCheck.Message("已经加载了缓存不需要再加载");
        }
        return sampleCacheTable;
    }