redis相比很多人都知道,是一个内存式的key-value数据库,存取速度极快,使用非常简单,支持多种语言。本文对其使用进行一个简要说明,并进行简单测试。
1.下载与编译
可以从redis官网下载最新的源码包:http://www.redis.io/
编译十分简单make既可。
2.redis安装与配置
实际上并不需要安装。redis编译后会在src目录下生成redis-server,它是一个可执行文件,即启动redis服务。不过它需要一个配置文件。配置文件写法网上很多了,这里直接给出一个示例:
daemonize yes pidfile /tmp/redis/var/redis.pid port 6379 timeout 300 loglevel debug logfile /tmp/redis/var/redis.log databases 16 save 900 1 save 300 10 save 60 10000 rdbcompression yes dbfilename dump.rdb dir /tmp/redis/var/ appendonly no appendfsync always #glueoutputbuf yes #shareobjects no #shareobjectspoolsize 1024 |
将其保存为redis.conf
然后直接运行./redis-server redis.conf就可以启动redis服务了,是不是很方便呢?
3.C/C++访问redis
在redis源码目录下有一个deps目录,下面有一个hiredis目录。redis编译时会自动编译该目录生成libhiredis.a,通过引用hiredis.h 和 libhiredis.a就可以访问redis了。具体步骤如下:
1)创建一个redisContext
2)通过redisContext执行命令
3)从返回redisReply中获取所需数据
代码如下:
<pre code_snippet_id="151033" snippet_file_name="blog_20140110_2_9509153" name="code" class="cpp">redisContext * c = redisConnect((char *)"192.168.150.135",6379); const char * pData = "this is a test"; <pre code_snippet_id="151033" snippet_file_name="blog_20140110_2_9509153" name="code" class="cpp">redisReply *reply1 = (redisReply *)redisCommand(c,"SET 100 %s",pData); <pre code_snippet_id="151033" snippet_file_name="blog_20140110_2_9509153" name="code" class="cpp"><pre code_snippet_id="151033" snippet_file_name="blog_20140110_2_9509153" name="code" class="cpp"><pre code_snippet_id="151033" snippet_file_name="blog_20140110_2_9509153" name="code" class="cpp">freeReplyObject(reply1);</pre> <pre></pre> <pre></pre> <pre></pre> <pre></pre> <pre></pre> <p></p> <pre></pre> <pre></pre> <br> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_3_730653" name="code" class="cpp"><pre code_snippet_id="151033" snippet_file_name="blog_20140110_3_730653" name="code" class="cpp"><pre code_snippet_id="151033" snippet_file_name="blog_20140110_3_730653" name="code" class="cpp">redisReply *reply2 = (redisReply *)redisCommand(c,"GET 100");</pre> <pre></pre> <pre></pre> printf("%s\n",reply2->str);<br> freeReplyObject(reply2);<br> <pre></pre> <pre></pre> <pre></pre> |
是不是非常简单呢?
<p></p> <p>不过需要注意的是,redis接受的数据是字符串,对于二进制数据,可以通过base64编码来解决。具体可参看我的另一篇文章。</p> <p></p> <h1><a name="t3"></a>4.Java访问redis</h1> <p>redis可以支持多种语言,当然也可以支持Java。</p> <p>首先需要下载redis的java包。jedis.jar。这里提供一个下载地址:<a target="_blank" href="http://download.csdn.net/detail/jmppok/6834151">redis的Java客户端jedis</a></p> <p>使用如下:</p> <p></p> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_4_2824204" name="code" class="java">Jedis jedis = new Jedis("192.168.150.135"); jedis.set("100","this is a test"); String data = jedis.get("100"); </pre><br> <br> <p></p> <h1><a name="t4"></a>5.性能测试</h1> <p>测试方法:向redis写一个1M的数据,分别写10次,读10次,计算其耗时。分C++和Java两个版本进行测试。<br> </p> <h2><a name="t5"></a>C++测试代码</h2> <p></p> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_5_8505805" name="code" class="cpp">#include <stdio.h> #include "hiredis.h" #include <string.h> #include <time.h> int main(int argc, char **argv) { printf("CLOCKS_PER_SEC:%d\n",CLOCKS_PER_SEC); redisContext *c; redisReply *reply; c = redisConnect((char *)"one-60",6379); char * pData; reply = (redisReply *)redisCommand(c,"GET 0"); int size = strlen(reply->str); pData = new char[size+1]; strcpy(pData,reply->str); freeReplyObject(reply); clock_t start, finish; start = clock(); for(int i=0;i<10; i++) { reply = (redisReply *)redisCommand(c,"GET %d",i); freeReplyObject(reply); } finish = clock(); double duration = (double)(finish - start) / CLOCKS_PER_SEC*1000; printf("GET Time used:%f ms.\n",duration); start = clock(); for(int i=0;i<10; i++) { reply = (redisReply *)redisCommand(c,"SET %d %s",i,pData); freeReplyObject(reply); } finish = clock(); duration = (double)(finish - start) / CLOCKS_PER_SEC*1000; printf("SET Time used:%f ms.\n",duration); delete []pData; redisFree(c); } </pre>测试结果 <p></p> <p></p> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_6_1363508" name="code" class="plain">CLOCKS_PER_SEC:1000000 GET Time used:190.000000 ms. SET Time used:70.000000 ms. </pre><br> <br> <p></p> <h2><a name="t6"></a>Java测试代码<br> </h2> <p></p> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_7_633160" name="code" class="java">import java.io.BufferedReader; import java.io.File; import java.io.FileReader; import java.util.Date; import redis.clients.jedis.Jedis; public class JedisTest { public static void main(String[] args) { Jedis jedis = new Jedis("10.100.211.232"); String f = "/tmp/e2.txt.backup"; try { File file = new File(f); BufferedReader reader = new BufferedReader(new FileReader(file)); String data = reader.readLine(); reader.close(); Date start = new Date(); for(int i=0; i<10; i++) { jedis.set(i+"", data); } Date end = new Date(); System.out.println("Set used(ms):"+(end.getTime()-start.getTime())); start = new Date(); for(int i=0; i<10; i++) { String v = jedis.get(i+""); } end = new Date(); System.out.println("Get used(ms):"+(end.getTime()-start.getTime())); }catch (Exception e) { e.printStackTrace(); } jedis.disconnect(); } }</pre><br> |
测试结果
<p></p> <p></p> <pre code_snippet_id="151033" snippet_file_name="blog_20140110_8_8615272" name="code" class="plain">Set used(ms):1212 Get used(ms):1437</pre><br> |
6.总结
<p></p> <p>redis效率还是非常高的,读写1M数据的数据,耗时都在10ms左右。<br> </p> <p><br> </p> <pre></pre> <pre></pre> </pre></pre></pre></pre></pre></pre> |
最新内容请见作者的GitHub页:http://qaseven.github.io/