ksbsrv: No startup acknowledgement from forked process after 30 seconds

数据库alert.log报错：

Sat JUN 5 14:23:10 2013
ksbsrv: No startup acknowledgement from forked process after 30 seconds

数据库listener.log报错：

TNS-12500: TNS：监听器未能启动专用的服务器进程
TNS-12560: TNS: 协议适配器错误
TNS-00530: 协议适配器错误

应用服务器客户端报错：

客户端连接连接数据库提示：TNS-12500

环境： windows 2003 64位 、oracle 9.2.0.1 非核心数据库OA系统

[@more@]

开发公司报，应用服务器登录失败，报错：TNS-12500

下面是数据库服务器后台告警日志显示：

Sat JUN 5 14:25:44 2012
ksbsrv: No startup acknowledgement from forked process after 30 seconds
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn
skgpspawn failed:category = 27143, depinfo = 9261, op = spdcr, loc = skgpspawn

metalink有一篇参考文档:ID388781.1

Goal
Fix
References

--------------------------------------------------------------------------------

This document is being delivered to you via Oracle Support's Rapid Visibility (RaV) process and therefore has

not been subject to an independent technical review.

Applies to:
Oracle Server - Enterprise Edition - Version 8.1.7.4 and later
Information in this document applies to any platform.
***Checked for relevance on 03-May-2012***

Goal
From time to time, messages like:

ksbsrv: No startup acknowledgement from forked process after 30 secondsappear in the database. The timeout can

have various values.

Fix
1.
When the message appears at startup and the database instance cannot be started, the solution is to remove the

.blocked files that reside in the $ORACLE_HOME/database directories.

2.
If the message is left in the alert log by a starting process (in most situations this is done by a slave

process, such as jxxx or by a shared server process):
Basically, this means that a process has started but it didn't provide an acknowledgment for the success of the

operation. If this ack signal is not given, the process dies after 30 (or other interval, depending on the

process type and oracle release) seconds.
The process did not hit a resource limit at startup, because when that happens, the message is different, ie:

"skgpspawn failed category".
This actually means that the process was delayed in delivering the acknowledge message due to a performance

problem, such as high CPU usage. In most situations the message would just not reoccur, since the performance

problem was particular to that moment.
When the message still reoccurs, to improve the situation one could improve the overall performance and try to

detect the bottleneck.
+ increase the available memory for the oracle process by enabling the 3 GB switch
+ tune the system IO
+ monitor the CPU usage and find out the thread that is causing the situation (use the qslice tool for that).

Generally, this is not a reason for concern when it is only sporadically occurring.
However, when occurring too often, it is recommended to monitor and tune the system, as the cause of the problem is slow performance, most likely due to CPU overloading.

References
@ BUG:3556455 - KSBSRV: NO STARTUP ACKNOWLEDGEMENT FROM FORKED PROCESS AFTER 30 SECONDS

本案应该是一个bug所致（BUG：3556455）,如果该事故是偶发性的发生，则可以忽略，如果该事故发生的频率较高，则就需要分析了。因为我这里是偶发性的，几年都没有出现这个错误，我给忽略了，因此我这里的解决方法是重启数据库服务，重启监听即可.

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/28227905/viewspace-1060006/，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/28227905/viewspace-1060006/