英文:
What are the web_merged.xml and ibm-metadata.xml files generated by websphere deployment?
问题
我们有一个自动化流程，在 WebSphere 8.5 上部署 EAR 应用程序。
在此过程中，web_merged.xml 和 ibm-metadata.xml 这两个文件是以一种不利于应用程序正常运行的方式生成的。
如果我们使用常规的手动更新方式在 WAS 控制台中部署应用程序，这两个文件会以正确的方式重新生成，应用程序就能正常运行。
- 这两个文件是什么？
- 在部署 EAR 包之前是否将它们包含在包中会是正确的做法？
- 还是它们必须由部署过程生成，然后会被覆盖？
- 如果是这样的话，也许应该由监督自动化部署流程的团队审查流程，以正确创建它们？
编辑：
我基本上同意 F Rowe 的回答，目前我们无法在服务器上安装补丁。目前，我临时解决方案是在另一个可以直接访问的 WebSphere 测试服务器中安装 EAR，然后从 Eclipse 导出 EAR 文件。这个 EAR 包含了生成的文件，并且已修改了 web.xml，以在部署时不会重新生成它们。我将其用于最终部署……在等待更新服务器并寻找真正解决方案的同时，这就是全部。
英文:
We have an automated process to deploy ear application on websphere 8.5.
During this process the files web_merged.xml and ibm-metadata.xml are generated in a way which does not let the application working well.
If we deploy the application using usual manual update in WAS Console the two files are newly generated in correct way and the application works well.
- What are those two files?
- Would be correct to include them in package before the EAR is deployed?
- Or they have to be generated by deployment process, they would be overwritten?
- In this case maybe the team supevisioning the automated process of deploy should review the process to create them correctly?
EDIT:
I generally agree with F Rowe answer, at this time it's not possible for us to install patch on server. For now I'm temporarily solving installing the ear in another websphere test server in which I have direct access, then export the ear file from eclipse. This ear has the files generated and the web.xml modified to not regenerate them newly when deploying. I'm using it for final deploy... that's all in the meanwhile that I wait for updating the server and check for real solution...
答案1
得分: 2
那两个文件是什么？ibm-metadata.xml 包含了注释扫描的结果，而 web_merged.xml 是将 web.xml 部署描述符与元数据合并的结果。
在将 EAR 文件部署之前，将它们包含进去是正确的吗？根据这个 IBM 支持页面，
> IBM 的建议是在部署应用程序之前移除任何预生成的元数据。
还是说它们必须由部署过程生成，会被覆盖吗？它们将会由部署过程生成并被覆盖。
你是否安装了包含了上述页面中所描述的 APAR 的修补程序级别？如果没有，你应该安装其中一个，看看是否可以解决你的问题，因为你的问题描述似乎与该 APAR 中解决的问题相匹配。
英文:
What are those two files? ibm-metadata.xml contains the result of the annotation scan, web_merged.xml is the result of merging the web.xml deployment descriptor and that metadata.
Would be correct to include them in package before the EAR is deployed? According to this IBM support page,
> the IBM recommendation is to remove any pre-generated metadata from an
> application before deploying the application.
Or they have to be generated by deployment process, they would be overwritten? They will be generated by the deployment process and overwritten.
Do you have one of the fix pack levels containing the APAR described in the above referenced page? If not, you should install one of them and see if it fixes your problem since your description seems to match the problem solved in that APAR.
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论