slove ‘Lock table overflow' process

出错情况：

用户报称修改QT时出错：‘Lock table overflow';

...

检查日志SM21,

Internal lock administration error
> EqSetCluster(); rtc=16 enxxmenq1905
Lock table overflow
Internal lock administration error
> EqSetCluster(); rtc=16 enxxmenq1905
Lock table overflow
Internal lock administration error
> EqSetCluster(); rtc=16 enxxmenq1905
Lock table overflow
Internal lock administration error

详细日志，

An overflow of the lock table occurred when you were setting the loc
The lock table consists of several individual tables each of which
can overflow individually. The subsequent line indicates which of th
tables has overflowed.
The lock table is contained in the memory of the lock server
(-> enqueue server). Only the total size can be set. This is set
with the profile parameter "enque/table_size" in the instance
profile of the central locking server. Its value is entered in KB.
The setting only becomes effective when the system is restarted.
The default size of the lock table is 4096 KB.
Please note that when increasing the size of the lock table, a disk
requirement of approx. 1.5 x the size of the lock table is necessary
in each case in the global directory for the reporting file and in
the log directory for the backup file.
Before you increase the lock table after an overflow, you should
always check whether the overflow resulted from a program error
or an operating error.

检查profile parameter "enque/table_size"　with RZ10,

Short description: Size of lock table

by this parameter:

Parameter description :

This parameter defines the size of the lock table that is held in
main memory by the enqueue server. The lock table contains all of
the information about which locks are currently held, and by whom.

The size of the lock table limits the maximum number of lock
entries, the maximum number of elementary lock names, and the
maximum number of lock owners.
In the case of an overflow of the lock table, you should check
whether the update server is working correctly, since the lock table
can grow in size very quickly if updates stop. If the cause is not
update problems, you can use this parameter to increase the size of
the lock table.
The Computing Center Management System (CCMS) constantly monitors
the status of the lock table, and issues warnings if the free space
runs short.

The Computing Center Management System (CCMS) constantly monitors
the status of the lock table, and issues warnings if the free space
runs short.
You can view the fill level of the lock tables using the statistics
functions of lock management (transaction SM12 -> Extras ->
Statistics).

实施时考虑过公司的运行情况，将实际大小设成10MB，

enque/table_size = 10000

check LOCK status with SM12,

发现total locks居然达到8K有余，Nr_of_Locks: 8701,有异于平时系统正常运行状态，而绝大部分是一个用户锁定的；

查看用户进程SM50,确认用户在运行CORUAFFW，询问用户运行Tcode：COGI,确认用户在做物料替换时中断了会话，从而导致系统锁定的大批物料而没有释放。

解决方案：end session in SM50.

[@more@]

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/167898/viewspace-1008902/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/167898/viewspace-1008902/