java.util.concurrent.TimeoutException: Timed out(1000) waiting for operation

[GoogleCodeExporter]

我照你们的文档用spring3配置了服务端和客户端，但在get/set数
据的时候却总是抛出java.util.concurrent.TimeoutException: Timed 
out(1000) waiting for 
operation的异常，我xmemcached和memcached都是用的最新版本
如果我在程序里直接用MemcachedClientBuilder builder = new 
XMemcachedClientBuilder( AddrUtil.getAddresses("192.168.1.24:12000 
192.168.1.88:12000"))的方式却不会报错，会是什么问题呢？怎么��
�决？

Original issue reported on code.google.com by sw5201...@gmail.com on 29 Jun 2012 at 4:56

[GoogleCodeExporter]

将你的xml配置贴出来，我估计是你设置了太大的连接池了

Original comment by killme2...@gmail.com on 29 Jun 2012 at 4:57

[GoogleCodeExporter]

<bean name="memcachedClientBuilder" 
class="net.rubyeye.xmemcached.XMemcachedClientBuilder">
        <constructor-arg>
            <list>
                <bean class="java.net.InetSocketAddress">
                    <constructor-arg>
                        <value>192.168.1.24</value>
                    </constructor-arg>
                    <constructor-arg>
                        <value>12000</value>
                    </constructor-arg>
                </bean>
                <bean class="java.net.InetSocketAddress">
                    <constructor-arg>
                        <value>192.168.1.88</value>
                    </constructor-arg>
                    <constructor-arg>
                        <value>12000</value>
                    </constructor-arg>
                </bean>
            </list>
        </constructor-arg>
        <constructor-arg>
            <list>
                <value>1</value>
                <value>2</value>
            </list>
        </constructor-arg>
        <!-- <property name="authInfoMap"> <map> <entry key-ref="server1"> <bean 
            class="net.rubyeye.xmemcached.auth.AuthInfo" factory-method="typical"> <constructor-arg 
            index="0"> <value>cacheuser</value> </constructor-arg> <constructor-arg index="1"> 
            <value>123456</value> </constructor-arg> </bean> </entry> </map> </property> -->
        <!-- nio connection pool size -->
        <property name="connectionPoolSize" value="5"></property>
        <!-- Use binary protocol,default is TextCommandFactory -->
        <property name="commandFactory">
            <bean class="net.rubyeye.xmemcached.command.TextCommandFactory"></bean>
        </property>

        <property name="sessionLocator">
            <bean class="net.rubyeye.xmemcached.impl.KetamaMemcachedSessionLocator"></bean>
        </property>
        <property name="transcoder">
            <bean class="net.rubyeye.xmemcached.transcoders.SerializingTranscoder" />
        </property>
        <property name="bufferAllocator">
            <bean class="net.rubyeye.xmemcached.buffer.SimpleBufferAllocator"></bean>
        </property>
    </bean>
    <bean name="memcachedClient" class="net.rubyeye.xmemcached.MemcachedClient" factory-bean="memcachedClientBuilder"
        factory-method="build" destroy-method="shutdown" />

对了memcached那边我用了magent做代理

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:00

[GoogleCodeExporter]

<property name="connectionPoolSize" value="5"></property>
将它设置1试试，并且设置opTimeout达到3000左右，1000通常来说��
�不够的，测试还是可以。

Original comment by killme2...@gmail.com on 29 Jun 2012 at 5:08

[GoogleCodeExporter]

还是一样

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:16

[GoogleCodeExporter]

郁闷，我重启下magent就Ok了，会是什么原因呢？

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:21

[GoogleCodeExporter]

而当我尝试不同的操作时也会经常报这个错，需要重启magent

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:26

[GoogleCodeExporter]

我不知道什么是magent。这个的意思是操作超时了，默认是1秒�
��时，你有设置opTImeout到3000，也就是3秒吗？

Original comment by killme2...@gmail.com on 29 Jun 2012 at 5:28

[GoogleCodeExporter]

设置了，设置多少都没有，magent就是个memcached的代理罗https://
code.google.com/p/memagent/

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:35

[GoogleCodeExporter]

你设置成多少，我看看，怎么设置的，不可能没用的。

Original comment by killme2...@gmail.com on 29 Jun 2012 at 5:48

[GoogleCodeExporter]

memcachedClient.setOpTimeout(3000);

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 5:58

[GoogleCodeExporter]

那只能说你的数据是不是太大，或者magent有没有什么问题，��
�致经常超时。可以将这个时间调长，或者直接链接memcached看�
��。

Original comment by killme2...@gmail.com on 29 Jun 2012 at 6:03

[GoogleCodeExporter]

确实去掉magent就没问题了，我想请问下如果不用magent，怎么��
�现memcached的负载均衡和单点故障？又或者在xmemcached里已有相
关解决方案？ 

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 6:19

[GoogleCodeExporter]

搜索memcached proxy，选择很多吧。
不过memcached的分布一般也都是在客户端做，一般都不搞proxy。
当然，代价就是缓冲命中率在故障时候下降。

Original comment by killme2...@gmail.com on 29 Jun 2012 at 6:22

[GoogleCodeExporter]

多谢

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 8:13

[GoogleCodeExporter]

我改了用twitter的twemproxy做proxy，一次操作也没成功过，一直��
�是java.util.concurrent.TimeoutException，郁闷

Original comment by sw5201...@gmail.com on 29 Jun 2012 at 9:46

[GoogleCodeExporter]

Original comment by killme2...@gmail.com on 16 Jul 2012 at 11:49

• Changed state: Invalid