0 积分	1 好友	2 主题

发消息

[SQL调优] ORA - 12537: TNS: Connection Closed

1^#

发表于 2014-7-16 21:34:45 | 查看: 4085| 回复: 3

aix 10.2.0.5的rac 和一台web服务器C
C 能正常连接rac1 节点
C不能访问rac2 节点但能ping通 tnsping不通
rac1 和 rac2 相互sqlplus连接都正常
三台机子在同一个网段里面
监听配置如下：
LISTENER_XYZXRAC2 =
  (DESCRIPTION_LIST =
(DESCRIPTION =
   (ADDRESS = (PROTOCOL = TCP)(HOST = xyzxrac2-vip)(PORT = 1521)(IP = FIRST))
   (ADDRESS = (PROTOCOL = TCP)(HOST = 10.20.10.156)(PORT = 1521)(IP = FIRST))
)
  )

SID_LIST_LISTENER_XYZXRAC2 =
  (SID_LIST =
(SID_DESC =
   (SID_NAME = PLSExtProc)
   (ORACLE_HOME = /oracle/product/10.2.0/db_1)
   (PROGRAM = extproc)
)
  )

客户端版本32位  10.2.0.1
tnsping 报错如下：
C:\Documents and Settings\xyzx>tnsping oracle_157

TNS Ping Utility for 32-bit Windows: Version 10.2.0.1.0 - Production on 16-7月 -
2014 19:59:46

Copyright (c) 1997, 2005, Oracle.  All rights reserved.

已使用的参数文件:
D:\oracle\product\10.2.0\db_1\network\admin\sqlnet.ora

已使用 TNSNAMES 适配器来解析别名
Attempting to contact (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)
(HOST = 10.20.10.157)(PORT = 1521))) (CONNECT_DATA = (SID = credit1)))
TNS-12537: TNS: 连接关闭

开启客户端跟踪如下：
[16-7月 -2014 17:25:29:343] nsdo: rank=64, nsctxrnk=0
[16-7月 -2014 17:25:29:343] nsdo: nsctx: state=2, flg=0x4005, mvd=0
[16-7月 -2014 17:25:29:343] nsdo: gtn=10, gtc=10, ptn=10, ptc=2011
[16-7月 -2014 17:25:29:343] nscon: entry
[16-7月 -2014 17:25:29:343] nscon: recving a packet
[16-7月 -2014 17:25:29:343] nsprecv: entry
[16-7月 -2014 17:25:29:343] nsprecv: reading from transport...
[16-7月 -2014 17:25:29:343] nttrd: entry
[16-7月 -2014 17:25:29:343] nttrd: exit
[16-7月 -2014 17:25:29:359] ntt2err: entry
[16-7月 -2014 17:25:29:359] ntt2err: Read unexpected EOF ERROR on 1760
[16-7月 -2014 17:25:29:359] ntt2err: exit
[16-7月 -2014 17:25:29:359] nsprecv: error exit
[16-7月 -2014 17:25:29:359] nserror: entry
[16-7月 -2014 17:25:29:359] nserror: nsres: id=0, op=68, ns=12537, ns2=12560; nt[0]=507, nt[1]=0, nt[2]=0; ora[0]=0, ora[1]=0, ora[2]=0
[16-7月 -2014 17:25:29:359] nscon: error exit
[16-7月 -2014 17:25:29:359] nsdo: nsctxrnk=0
[16-7月 -2014 17:25:29:359] nsdo: error exit
[16-7月 -2014 17:25:29:359] nscall: unexpected response
[16-7月 -2014 17:25:29:359] nsclose: entry
[16-7月 -2014 17:25:29:359] nstimarmed: entry
[16-7月 -2014 17:25:29:359] nstimarmed: no timer allocated
[16-7月 -2014 17:25:29:359] nstimarmed: normal exit
[16-7月 -2014 17:25:29:359] nttctl: entry
[16-7月 -2014 17:25:29:359] nttctl: entry
[16-7月 -2014 17:25:29:359] nsdo: entry
[16-7月 -2014 17:25:29:359] nsdo: cid=0, opcode=98, *bl=0, *what=0, uflgs=0x440, cflgs=0x2
[16-7月 -2014 17:25:29:359] nsdo: rank=64, nsctxrnk=0
[16-7月 -2014 17:25:29:359] nsdo: nsctx: state=1, flg=0x4001, mvd=0
[16-7月 -2014 17:25:29:359] nsbfr: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: normal exit
[16-7月 -2014 17:25:29:359] nsbfr: normal exit
[16-7月 -2014 17:25:29:359] nsbfr: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: normal exit
[16-7月 -2014 17:25:29:359] nsbfr: normal exit
[16-7月 -2014 17:25:29:359] nsdo: nsctxrnk=0
[16-7月 -2014 17:25:29:359] nsdo: normal exit
[16-7月 -2014 17:25:29:359] nsiocancel: entry
[16-7月 -2014 17:25:29:359] nsiofrrg: entry
[16-7月 -2014 17:25:29:359] nsbfr: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: entry
[16-7月 -2014 17:25:29:359] nsbaddfl: normal exit
[16-7月 -2014 17:25:29:359] nsbfr: normal exit
[16-7月 -2014 17:25:29:359] nsiofrrg: exit

分享0

收藏0 回复只看该作者道具举报

jason_yh

4^#

发表于 2014-7-17 20:56:29

今天又做了一个tnsping的跟踪，大致可以确认是防火墙上的限制，坐等工程师回复感谢刘大

回复只看该作者道具举报

Maclean Liu(刘相兵

3^#

发表于 2014-7-17 20:39:49

1、是否存在防火墙端口限制或者规则
2、“ rac1 和 rac2 相互sqlplus连接都正常” ，就你所说的这一点来看，第一个点的可能性很高
3、如果无法确认，那么继续调试

回复只看该作者道具举报

Maclean Liu(刘相兵

2^#

发表于 2014-7-17 20:38:27

odm finding:

Several possible situations can cause this to happen:

client changed its mind and closed the connection immediately after initiating it
client crashed
firewall kills the connection
some oracle timeout set on client
From all the above causes only the last one can be checked by Oracle Support, because all the others are outside oracle product, they are not under our control/possibility for investigation.

Because the entry from listener.log contains only CONNECT_DATA and CID related information we need to check the client configuration for any sqlnet timeouts:

possible timeouts in sqlnet.ora in client oracle home:
sqlnet.outbound_connect_time
sqlnet.recv_timeout
sqlnet.send_timeout
tcp_connect_timeout

possible timeout in client connect descriptor (hardcoded in client application or in client tnsnames.ora):
connect_timeout

回复只看该作者道具举报

返回列表

		自动登录	找回密码
密码			注册