0

In my organization, we conduct Data Entry Operation though a 3-tier Application System. The Configuration is shown as below:

  • Operating System : Linux
  • Application Software has been built using: JSF (a Java EE Framework Technology)
  • Database : Oracle 11g

The Problem: For each entry an Enterprise bean is created and kept in a queue inside the Weblogic Sever. The Bean is being sent to the database but the memory used for the bean is not being released. As a result for a certain time duration, when the hits reach a peak and the memory is being used up to say 16GB or more, the server hangs responding and it has to be restarted in order to release the memory.

To be on the safe side, we have set the maximum simultaneous connections to the maximum. But this doesn't help.

What may be the problem and how to tune the system?

4

1 回答 1

1

Bean 被发送到数据库，但用于 bean 的内存没有被释放。

由于某些资源（例如数据库连接）未在代码中关闭，此处可能存在内存泄漏。

将堆转储到 HPROF 文件中并使用 Eclipse 内存分析器工具或VisualVM运行它或使用JConsole 监控它

于 2012-04-23T11:14:32.737 回答