EBS11.5.10 @redhat4 安装后客户端连接DB 出现ORA-12547
作者: moonsoft(http://moonsoft.itpub.net)发表于: 2008.03.26 08:12
分类: Linux , Oracle_DBA
出处: http://moonsoft.itpub.net/post/15182/458248
---------------------------------------------------------------
ORA-12547并不常见？先GOOGLE 一下，得到说是少一个包，可是我查看了我的LINUX 为UPDATE5，且非64位， 如果是LINUX3以下的倒是有下载：
http://oss.oracle.com/projects/compat-oracle/files/RedHat/
| binutils-2.15.92.0.2-13.0.0.0.2.x86_64.rpm | 2005.10.05 | 7362f50479940072e79136125ea32321 |
| RHEL 4 Update 1 patched binutils neccessary for 10gR2 install on x86_64. IMPORTANT: The problem with binutils has been fixed in RHEL4 Update 3, so starting from that version this patch is no longer required. | ||
| binutils-2.15.92.0.2-15.0.0.0.2.x86_64.rpm | 2006.05.11 | a0c38cca671745ae66f14957ef086091 |
RHEL 4 Update 2 patched binutils neccessary for 10gR2 install on x86_64. IMPORTANT: The problem with binutils has been fixed in RHEL4 Update 3, so starting from that version this patch is no longer required. 奇怪了，肯定不是这个包的问题！ | ||
明确几个前提：
安装过程很仔细（因为已经失败过多次，所以也成熟许多）
网络连接没有问题
DB， APPS 服务都已经正常开启
TNSNAME 没有问题
为什么还是不行，METALINK 也查看了很多：
FROM METALINK :
VALIDNODE_CHECKING REFUSES CONNECTIONS WITH ORA-12547 OR ORA-12571
Currently, when we activate validnode_checking, and a connection comes from one
of the excluded nodes or one node not in the invited nodes, client gets a
ORA-12547 error message:
12547, 00000, "TNS:lost contact"
// *Cause: Partner has unexpectedly gone away, usually during process
// startup.
// *Action: Investigate partner application for abnormal termination. On an
// Interchange, this can happen if the machine is overloaded.
.
Also, there is a ORA-12206 error on the stack
12206, 00000, "TNS:received a TNS error during navigation"
// *Cause: Internal navigation error because of an unexpected TNS error.
// *Action: Look at the log file to find the TNS error. If necessary, turn
// on tracing and repeat the operation.
.
Basically, we need a better error message for this type of situations
ORA-12547 is not a very helpful error message, and it's not clear if connection
is failing due to being reject on validnode_checking or due to other error.
A better error message, something like:
"Your node is not allowed to connect to this service"
would be much better, since it will not allow any doubts why the connection is
not allowed
*** 11/22/96 07:06 am *** (CHG: Sta->11 Pri->3 Asg->NEW OWNER)
Connected from a Windows 95 client to the server will give:
ORA-12571: TNS:packet writer failure
This has *nothing* do to with the actual problem.
*** 11/22/96 05:12 pm ***
*** 12/19/96 02:00 pm *** (CHG: Asg->NEW OWNER)
*** 12/19/96 02:00 pm ***
I agree -- the message could be better.
*** 01/29/99 09:20 am *** (CHG: Asg->NEW OWNER)
*** 03/17/99 06:16 pm *** (CHG: Asg->NEW OWNER)
*** 01/25/02 11:42 am *** (CHG: Asg->NEW OWNER)
*** 06/28/04 12:54 pm *** (CHG: Asg->NEW OWNER)
*** 02/28/05 04:30 pm *** (CHG: Asg->NEW OWNER)
. 以上是METALINK 的官方解释，但是觉得方法不见，郁闷，找到了一篇中文解释，摘录一段：
客户端计算机无法使用工具直接连接ERP数据库，提示报错处理方法：修改服务端 sqlnet.ora 文件，具体如下
cd /u1/VIS/visdb/9.2.0/network/admin/VIS_erp
修改 sqlnet.ora，令 tcp.validnode_checking = no
我修改后，连接OK，感谢！
