至于我想安装詹金斯（前哈德森）来操作我连续一体化进程在AWS上魔豆，我需要一个自定义的AMI，因为在Tomcat中与​​放一些参数; Linux的必须为詹金斯改变

As I want to install Jenkins (ex-Hudson) to operate my continuous integration processes on AWS Beanstalk, I need a custom AMI because some parameters in Tomcat & Linux have to be changed for Jenkins

我运行安装和自定义实例的进程始于最初由魔豆，直到结束，詹金斯的工作原理就像一个魅力就可以了。

I run the process of installing and customizing the instance started initially by Beanstalk until the end and Jenkins works like a charm on it.

不过，我不能做的是重用我在我定制的结束产生的AMI：由魔豆做没有看到EC2实例，虽然魔豆开始了它，它工作正常的健康检查

But, what I can't do is reuse the AMI that I generated at the end of my customization: the health check done by BeansTalk doesn't see the EC2 instance although Beanstalk started it and it works fine.

为了了解我的问题，我减少了我的失败过程如下：

In order to understand my issue, I reduced my failing process to the following:

一）创建基于亚马逊提供的样本一个新的BT应用程序/环境（即我只有参数是密钥对SSH我的EC2实例） b）在EC2实例启动时，我使用EC2闪烁的AMI 由我genrated在（B） 1的Tomcat 6 64位） - 100fff79：在三）我改变了原来的AWS阿美（ID修改BT ENV配置 四）BT重建当我改变了AMI ID E）重建重新启动EC2实例。 F），它开始精（可以ssh的话），但健康检查失败，我的ENV变为红色状态。 a) I create a new BT application / environment based on sample provided by Amazon (only parameter that I had is a keypair to SSH my EC2 instance) b) when the EC2 instance is started, I use the EC2 to flash the AMI c) I modify the BT env config by changing the original AWS Ami (id: 100fff79 - Tomcat 6 64 bits) by the 1 that I genrated in (b) d) the BT rebuilds when I change the ami id e) the rebuild restarts the EC2 instance. f) It starts fine (can ssh to it) but the health checking fails and my env turns to red status.

有人可以重复这一过程，并告诉我什么，我做错了什么？ （我想用的AMI（B）为起点，我的詹金斯定制？

Can somebody replicate this process and tell me what I am doing wrong ? (I would like to use the AMI of (b) as starting point for my Jenkins customization.?

其他信息，我可以提供： 当SSH-荷兰国际集团的EC2实例，一个grep的为Apache，JAVA，薄放大器; bluepilld于底部描述 HTTPS的：//forums.aws.amazon ？.COM / thread.jspa主题ID = 59027＆放大器; TSTART = 25 显示，预计4流程消失。因此，故障

Additional info that I can provide: when ssh-ing to the EC2 instance, a grep for apache, java, thin & bluepilld as described at bottom of https://forums.aws.amazon.com/thread.jspa?threadID=59027&tstart=25 shows that the 4 expected processes disappeared. Hence, the failure.

请帮助！

关于

迪迪埃

推荐答案

会回答我的问题：正确的方式来获得一个工作的定制AMI的魔豆是不是要尝试更新由魔豆推出了一个正在运行的实例，而是启动模板AMI的魔豆（AMI-100fff79为Tomcat 6 64位在我的情况），从EC2控制台并从那里定制它，闪烁它，你就大功告成了。

will answer my own question: the right way to obtain a working customized ami for Beanstalk is not to try to flash a running instance launched by Beanstalk but rather start the template ami for Beanstalk (ami-100fff79 for Tomcat 6 64 bits in my case) from EC2 console and customize it from there, flash it and you're done.

您可以再编辑配置为你的BT环境改变AMI到新的，它工作正常。

You can then "edit configuration" for your BT environment by changing the ami to the new one and it works fine.

问候 迪迪埃