基于redis的cas集群配置

1、cas ticket统一存储

　　做cas集群首先需要将ticket拿出来，做统一存储，以便每个节点访问到的数据一致。官方提供基于memcached的方案，由于项目需要，需要做计入redis，根据官方例子改了一个基于redis版本的。

public class RedisTicketRegistry extends AbstractDistributedTicketRegistry{
    @NotNull
    private final RedisTemplate<String,Object> reidsTemplate;

    /**
     * TGT cache entry timeout in seconds.
     */
    @Min(0)
    private final int tgtTimeout;

    /**
     * ST cache entry timeout in seconds.
     */
    @Min(0)
    private final int stTimeout;

    public RedisTicketRegistry(RedisTemplate<String,Object> reidsTemplate,int tgtTimeout,int stTimeout){
        this.reidsTemplate=reidsTemplate;
        this.tgtTimeout=tgtTimeout;
        this.stTimeout=stTimeout;
    }
    @Override
    public void addTicket(Ticket ticket) {
        logger.debug("Adding ticket {}", ticket);
        try {
            reidsTemplate.opsForValue().set(ticket.getId(),ticket, getTimeout(ticket), TimeUnit.SECONDS);
        } catch (final Exception e) {
            logger.error("Failed adding {}", ticket, e);
        }
    }

    @Override
    public Ticket getTicket(String ticketId) {
         try {
                final Ticket t = (Ticket) this.reidsTemplate.opsForValue().get(ticketId);
                if (t != null) {
                    return getProxiedTicketInstance(t);
                }
            } catch (final Exception e) {
                logger.error("Failed fetching {} ", ticketId, e);
            }
            return null;
    }

    @Override
    public boolean deleteTicket(String ticketId) {
         logger.debug("Deleting ticket {}", ticketId);
            try {
                 this.reidsTemplate.delete(ticketId);
                 return true;
            } catch (final Exception e) {
                logger.error("Failed deleting {}", ticketId, e);
            }
            return false;
    }

    @Override
    public Collection<Ticket> getTickets() {
         throw new UnsupportedOperationException("GetTickets not supported.");
    }

    @Override
    protected void updateTicket(Ticket ticket) {
     logger.debug("Updating ticket {}", ticket);
        try {
              this.reidsTemplate.delete(ticket.getId());
              reidsTemplate.opsForValue().set(ticket.getId(),ticket, getTimeout(ticket), TimeUnit.SECONDS);
        } catch (final Exception e) {
            logger.error("Failed updating {}", ticket, e);
        }
    }

    @Override
    protected boolean needsCallback() {
        // TODO Auto-generated method stub
        return true;
    }
   private int getTimeout(final Ticket t) {
        if (t instanceof TicketGrantingTicket) {
            return this.tgtTimeout;
        } else if (t instanceof ServiceTicket) {
            return this.stTimeout;
        }
        throw new IllegalArgumentException("Invalid ticket type");
    }
}

　　对应的ticketRegistry.xml的配置如下：

 <bean id="ticketRegistry" class="com.test.cas.ticket.registry.RedisTicketRegistry">
        <constructor-arg index="0" ref="redisTemplate" />
        <constructor-arg index="1" value="1800" />
        <constructor-arg index="2" value="10" />
    </bean>
    <bean id="poolConfig" class="redis.clients.jedis.JedisPoolConfig">
        <property name="maxIdle" value="200" />
        <property name="testOnBorrow" value="true" />
    </bean>  

    <bean id="connectionFactory" class="org.springframework.data.redis.connection.jedis.JedisConnectionFactory"
        p:host-name="redis_server_ip" p:port="6379"   p:pool-config-ref="poolConfig"/>
    <bean id="redisTemplate" class="org.springframework.data.redis.core.RedisTemplate"
        p:connection-factory-ref="connectionFactory">
    </bean>

　　这里既然使用了redis作为ticket存储，就需要将原来的方案给注释掉：

 <!-- Ticket Registry
  <bean id="ticketRegistry" class="org.jasig.cas.ticket.registry.DefaultTicketRegistry" />
  -->
    <!--Quartz -->
    <!-- TICKET REGISTRY CLEANER
    <bean id="ticketRegistryCleaner" class="org.jasig.cas.ticket.registry.support.DefaultTicketRegistryCleaner"
        p:ticketRegistry-ref="ticketRegistry"
        p:logoutManager-ref="logoutManager" />

    <bean id="jobDetailTicketRegistryCleaner" class="org.springframework.scheduling.quartz.MethodInvokingJobDetailFactoryBean"
        p:targetObject-ref="ticketRegistryCleaner"
        p:targetMethod="clean" />

    <bean id="triggerJobDetailTicketRegistryCleaner" class="org.springframework.scheduling.quartz.SimpleTriggerBean"
        p:jobDetail-ref="jobDetailTicketRegistryCleaner"
        p:startDelay="20000"
        p:repeatInterval="5000000" />
    -->

　　到这里，cas的改进就OK了，下面就需要将session也做集群同步。

2、tomcat session集群同步

　　这里采用开源的tomcat-redis-session-manager,git hub地址为：https://github.com/jcoleman/tomcat-redis-session-manager

　　这里只使用的是jdk1.7，tomcat7，tomcat6需要重新编译好像。

　　1）拷贝编译打包之后的tomcat-redis-session-manager-VERSION.jar，jedis-2.5.2.jar，commons-pool2-2.2.jar到tomcat/lib目录下

　　2）修改Tomcat context.xml (or the context block of the server.xml if applicable.)

<Valve className="com.orangefunction.tomcat.redissessions.RedisSessionHandlerValve"/>
<Manager className="com.orangefunction.tomcat.redissessions.RedisSessionManager"
         host="redis_server_name"
         port="6379"
         database="0"
         maxInactiveInterval="1800"/>

至此cas集群配置就可以了。