RS队列满，并且报内部错误

guzhenbin

环境:HP-UNIX
       sybase 12.5.4
       rs 12.6
复制服务器报错，服务启动后一分钟就报错并自动就down掉，而且有内部错误，一开始怀疑是rs的bug和队列满造成的，后给rs打了最新补丁，并加了2个队列，但rs仍然报内部错误，不知道是什么原因，请高人指点。下面是rs报错日志。
H. 2007/07/20 16:10:50. THREAD FATAL ERROR #5049 DSI EXEC(50067(1) TH111.DB_db) - dsiqmint.c(3041)
        The DSI thread for database 'TH111.db' is being shutdown. DSI received data server error #2601 which is mapped to STOP_REPLICATION. See logged data server errors for more information. The data server error was caused by output command #1 mapped from input command #16 of the failed transact
I. 2007/07/20 16:10:50. The DSI thread for database 'TH111.db' is shutdown.
N. 2007/07/20 16:12:14. INTERNAL ERROR #23 DSI(50064 TH111.DB_k) - dsisched.c(2655)
        Failed to assert '!isLarge' at 2655, 'neric/dsi/dsisched.c'.
T. 2007/07/20 16:12:14. (175): Dumping memory trace.
T. 2007/07/20 16:12:14. (175): ***************STACK TRACE***************
T. 2007/07/20 16:12:14. (175): *****thread DSI(50064 TH111.DB_k)*******
(0) 0x000000000442e4e0  dbg_dump_stack + 0x2b0 [/export/home/sybase/REP-12_6/bin/repserver]
(1) 0x00000000041bf780  exc__appDumpStack + 0x120 [/export/home/sybase/REP-12_6/bin/repserver]
(2) 0x00000000041c38f0  exc_terminate + 0x730 [/export/home/sybase/REP-12_6/bin/repserver]
(3) 0x00000000041c1d80  exc_reraise_exception + 0x180 [/export/home/sybase/REP-12_6/bin/repserver]
(4) 0x00000000041c1560  exc_raise_exception + 0x5d0 [/export/home/sybase/REP-12_6/bin/repserver]
(5) 0x000000000416bfa0  dsi__GroupXact + 0x940 [/export/home/sybase/REP-12_6/bin/repserver]
(6) 0x00000000043b0190  _sqt_close_tran + 0x550 [/export/home/sybase/REP-12_6/bin/repserver]
(7) 0x00000000043a20e0  _sqt_add_to_cache + 0x2470 [/export/home/sybase/REP-12_6/bin/repserver]
( 0x0000000004393ea0  sqt_LoadCache + 0x5c0 [/export/home/sybase/REP-12_6/bin/repserver]
(9) 0x0000000004169540  dsi__SchedulerBody + 0x440 [/export/home/sybase/REP-12_6/bin/repserver]
(10) 0x0000000004167970  dsi__SchedulerWrapper + 0xa30 [/export/home/sybase/REP-12_6/bin/repserver]
(11) 0x000000000452a100  srv__start_function + 0x230 [/export/home/sybase/REP-12_6/bin/repserver]
(12) 0x60000000c0335d20  __pthread_bound_body + 0x170 [/usr/lib/hpux32/libpthread.so.1]
T. 2007/07/20 16:12:14. (175): *****End of stack trace.****
T. 2007/07/20 16:12:14. (175): Internal error. Attempting to produce a core file

guzhenbin

各位高人给提个方向吧，我现在一点办法都没有了。

Eisen

呵呵。修改一下rs_config里面的memory_limit和num_threads之后重起试试看啦

aliking

先把RS队列满的问题解决了，否则就会内部错误。

解决队列满的几种办法：
1、跳过阻塞的事务resume connect to dbxxx skip transaction；
2、清空队列，sysadmin purge_queue;
3、重建队列，rebuild queue;
其中，2、3需要做丢失检测。
解决这些问题的时候，要注意相关数据库和相关RS队列的情况。
因为有可能你刚把队列清空，数据库日志或其他RS里堆积了很久的消息又会把它塞满。