[CLOUDSTACK-8158] After the host reboots, the system will run out vm management IP, no matter how much.

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 4.4.0, 4.4.1, 4.4.2
• Fix Version/s: Future
• Component/s: KVM
• Security Level: Public (Anyone can view this level - this is the default.)
• Labels: patch

Environment

All nodes in the system OS: Centos 6.5 x64 are
Cloudstack Version: 4.4.1 / 4.4.2 / 4.4.0
All nodes are installed on a single host, use the basic network, installation documentation is installed in accordance with the documentation provided by the official.

• Assignee: Rohit Yadav
• Reporter: liliang
• Created: 16/Jan/15 02:15
• Updated: 10/Aug/17 11:24
• Resolved: 10/Aug/17 11:24

Description

cloudstack all nodes installed on a host, all nodes in the system OS are centos6.5 x64. After using basic network build regional success, reboot the host, the system will manage IP vm exhausted. Every restart a host system vm will occupy the new management IP address, the old IP address can not be released. According to the analysis system BUG logs, restart once every host system vm will automatically rebuild destroyed repeatedly, until the system vm system up or manage IP exhausted, the system can start vm, the premise is under the management of IP abundant, management IP consumption when you do not start the system vm, database tables need to manually release the management IP. And under normal circumstances, reboot the system into the production environment is not su, senior network has not been detected this problem,

cloudstack 所有节点安装在一台资源充沛的宿主机，所有节点系统OS均为 centos6.5 x64。安装成功后，使用基本网络搭建区域成功后，重启宿主机后，系统vm会把管理IP耗尽。每重启一次宿主机，系统vm都会占用新的管理IP地址，旧的IP地址无法释放。根据系统BUG日志的分析，每重启一次宿主机，系统vm会自动反复重建销毁，直到系统vm系统起来或管理IP耗尽，系统vm可以正常启动，前提是管理IP充裕的情况下，管理IP耗尽时则无法启动系统vm ，需要到数据库表中手动释放管理IP。都未运行任何实例。而在正常的情况下，重启系统宿主机，系统vm不会自动销毁重建，或占用大量IP。而在4.3.1或4.3.0以下，包括（4.3.0/4.3.1）不会出现上述问题，安装方式是按照官方提供的文档进行的安装，安装文档是：http://cloudstack-installation.readthedocs.org/en/latest/qig.html
未投入生产环境中，高级网络中尚未检测到这种问题。