2012虚拟机闪锁解决

5年前
2016年1月30日
5回复
1299的浏览量

Userlevel 2

+ 13

斯顿
开拓者
15日回复

我曾经遇到过一个间歇性的问题，我迁移的一些2012年的vm在一段随机时间后被锁定，通常是100% CPU和所有IO停止。虚拟机控制台无响应和网络不可达，复位是唯一的解决方案带回联机。然后它会再运行几分钟，几个小时，甚至一天或更久，只是在白天或晚上的随机时间做同样的事情，所以它没有逻辑模式。

我最后跟踪到一个时区配置问题，当重新启动有问题的虚拟机时，时区会将它们的时钟设置为虚拟机管理程序的主机时间，而该主机VM TZ的时区是不同的(默认的TZ节点是附带的)。我过去也遇到过类似的问题，不过不是主机锁定的问题，在我们的Citrix XenServer集群管理程序主机上，许多主机会失去时间同步，并且由于主机和VM之间的TZ差异而失效。因为我们的虚拟机在AD域中，时间差异太大，超出了NTP的调整范围。

作为集群设置的一部分，我将集群时区更改为与我的时区相同(澳大利亚/珀斯)。但未能做AHV主机TZ。

在改变AHV主机纠正TZ和重新启动虚拟机后，所有看起来现在都很好，没有更多的主机锁定，也在每次重新启动后的正确时间

\n
\nI endup tracking it down to a timezone config issue, when rebooting the VMs in question the timezone would set their clock to the hypervisior host time, which was in a different timezone (default TZ nodes were shipped with) to the the host VM TZ. I had a similar problem in the past, although not with the host lockups, where a number of hosts would lose time sync and be out by the TZ difference between host and VM on our Citrix XenServer Cluster hypervisor host. Because our VM are on an AD domain they will be out of NTP adjustment range becasue time difference was too big.
\n
\nAs part of my cluster setup I had changed the cluster timezone to be same as mine (Australia\/Perth). but failed to do the AHV hosts TZ.
\n
\nAfter changing AHV hosts to correct TZ and Rebooting VMs, all appears to be fine now and no more host lockups, also on the correct time after reboots each and every time 😉","quoteUsername":"seaton","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">