最近遇到这个问题，其实提示很简单了。
去oracle 论坛看下，有人回复了
Extract was off for two long. All redo logs have been switched.
You have to provide archive logs to Oracle default destination or configure destination in GG parameter file.
If you don't have archive logs you have to resynchronize the two databases.
大概意思
由于线程2为1的归档日志已经被删除，导致goldengate 抽取进程停止：
解决方法：
1、恢复删除的归档日志，重新启动抽取进程；
rman target ggs/ggs@orcl
RMAN>restore archivelog from logseq 1 until logseq 587 ;
GGSCI>start ext ext
fuck恢复报错
3、通过v$archived_log查询到下个日志的时间点，重新指定抽取进程的时间点,启动抽取进程；
GGSCI>alter extract extjms2 ,begin 2014-11-20 00:20:15
GGSCI>start extjms2
4.实在不行就从现在启动。
这个会缺失数据的！！！！
GGSCI> alter extract ext begin now
GGSCI> start ext ext