为什么KVM计算机点无故重启?
一.故障1:机器hangs
本地一台cloudstack计算节点无故连不上了,cloudstack也坏了,后查看有一台系统虚拟机在这台计算节点上,导致cs挂了。去找到这台机器后,发现这台机器卡住了,重启后卡在starting udev,等好久也不行,即使进入单用户也是一样,重启很多次也是卡在这。最后查到一篇文章,原话:
Remove quiet
from the kernel command line and you should get enough output to see the cause of the hang.
然后进入系统菜单,在编辑kernel行,quiet静默模式。相当于设置"loglevel=4"(WARNING)。将quiet删除后,等待进入系统。大概一个小时后,进去系统了,能ssh了,问题解决了一个,得查一下机器为什么重启。
二.故障2:kvm计算机点为什么无故重启
通过查看cloudstack-agent的日志。
[root@kvm204 ~]# tail -1000f /var/log/cloudstack/agent/cloudstack-agent.out
看到了这里有报警,并且执行了reboot命令。继续往上面翻,又看到一条信息,如下:
看到很多这样的报错,但最后一次报错是下面这个,可能是它导致了kvm检测脚本执行了reboot命令,然后就出现了本文第一张图片日志里reboot the host。
现在知道了是谁重启了电脑,但kvm为什么会重启这台计算节点呢?带着疑问,我去查看了一下那个脚本,也就是kvmheartbeat.sh脚本。内容如下:
#!/bin/bash
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License. help() {
printf "Usage: $0
-i nfs server ip
-p nfs server path
-m mount point
-h host
-r write/read hb log
-c cleanup
-t interval between read hb log\n"
exit
}
#set -x
NfsSvrIP=
NfsSvrPath=
MountPoint=
HostIP=
interval=
rflag=
cflag= while getopts 'i:p:m:h:t:rc' OPTION
do
case $OPTION in
i)
NfsSvrIP="$OPTARG"
;;
p)
NfsSvrPath="$OPTARG"
;;
m)
MountPoint="$OPTARG"
;;
h)
HostIP="$OPTARG"
;;
r)
rflag=
;;
t)
interval="$OPTARG"
;;
c)
cflag=
;;
*)
help
;;
esac
done if [ -z "$NfsSvrIP" ]
then
exit
fi #delete VMs on this mountpoint
deleteVMs() {
local mountPoint=$
vmPids=$(ps aux| grep qemu | grep "$mountPoint" | awk '{print $2}' > /dev/null)
if [ $? -gt ]
then
return
fi if [ -z "$vmPids" ]
then
return
fi for pid in $vmPids
do
kill - $pid &> /dev/null
done
} #checking is there the same nfs server mounted under $MountPoint?
mounts=$(cat /proc/mounts |grep nfs|grep $MountPoint)
if [ $? -gt ]
then
# remount it
mount $NfsSvrIP:$NfsSvrPath $MountPoint -o sync,soft,proto=tcp,acregmin=,acregmax=,acdirmin=,acdirmax=,noac,timeo=,retrans= &> /dev/null
if [ $? -gt ]
then
printf "Failed to remount $NfsSvrIP:$NfsSvrPath under $MountPoint"
exit
fi
if [ "$rflag" == "" ]
then
deleteVMs $MountPoint
fi
fi hbFolder=$MountPoint/KVMHA/
hbFile=$hbFolder/hb-$HostIP write_hbLog() {
#write the heart beat log
stat $hbFile &> /dev/null
if [ $? -gt ]
then
# create a new one
mkdir -p $hbFolder &> /dev/null
touch $hbFile &> /dev/null
if [ $? -gt ]
then
printf "Failed to create $hbFile"
return
fi
fi timestamp=$(date +%s)
echo $timestamp > $hbFile
return $?
} check_hbLog() {
now=$(date +%s)
hb=$(cat $hbFile)
diff=`expr $now - $hb`
if [ $diff -gt $interval ]
then
return
fi
return
} if [ "$rflag" == "" ]
then
check_hbLog
if [ $? == ]
then
echo "=====> ALIVE <====="
else
echo "=====> DEAD <======"
fi
exit
elif [ "$cflag" == "" ]
then
reboot
exit $?
else
write_hbLog
exit $?
fi
上面这个脚本是我们现在的计算节点上的版本。从上面可以看出,有判断reboot命令,也就是知道了为什么会重启。但这很坑爹呀,你检测检测,你还重启。。。。
下面这个脚本是我在github上看到的。而且版本已经更新。距离现在8月。应该是最新的,从下面看出,这个坑爹的reboot命令已经被修改了。
#!/bin/bash
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License. help() {
printf "Usage: $0
-i nfs server ip
-p nfs server path
-m mount point
-h host
-r write/read hb log
-c cleanup
-t interval between read hb log\n"
exit
}
#set -x
NfsSvrIP=
NfsSvrPath=
MountPoint=
HostIP=
interval=
rflag=
cflag= while getopts 'i:p:m:h:t:rc' OPTION
do
case $OPTION in
i)
NfsSvrIP="$OPTARG"
;;
p)
NfsSvrPath="$OPTARG"
;;
m)
MountPoint="$OPTARG"
;;
h)
HostIP="$OPTARG"
;;
r)
rflag=
;;
t)
interval="$OPTARG"
;;
c)
cflag=
;;
*)
help
;;
esac
done if [ -z "$NfsSvrIP" ]
then
exit
fi #delete VMs on this mountpoint
deleteVMs() {
local mountPoint=$
vmPids=$(ps aux| grep qemu | grep "$mountPoint" | awk '{print $2}' > /dev/null)
if [ $? -gt ]
then
return
fi if [ -z "$vmPids" ]
then
return
fi for pid in $vmPids
do
kill - $pid &> /dev/null
done
} #checking is there the same nfs server mounted under $MountPoint?
mounts=$(cat /proc/mounts |grep nfs|grep $MountPoint)
if [ $? -gt ]
then
# remount it
mount $NfsSvrIP:$NfsSvrPath $MountPoint -o sync,soft,proto=tcp,acregmin=,acregmax=,acdirmin=,acdirmax=,noac,timeo=,retrans= &> /dev/null
if [ $? -gt ]
then
printf "Failed to remount $NfsSvrIP:$NfsSvrPath under $MountPoint"
exit
fi
if [ "$rflag" == "" ]
then
deleteVMs $MountPoint
fi
fi hbFolder=$MountPoint/KVMHA/
hbFile=$hbFolder/hb-$HostIP write_hbLog() {
#write the heart beat log
stat $hbFile &> /dev/null
if [ $? -gt ]
then
# create a new one
mkdir -p $hbFolder &> /dev/null
touch $hbFile &> /dev/null
if [ $? -gt ]
then
printf "Failed to create $hbFile"
return
fi
fi timestamp=$(date +%s)
echo $timestamp > $hbFile
return $?
} check_hbLog() {
now=$(date +%s)
hb=$(cat $hbFile)
diff=`expr $now - $hb`
if [ $diff -gt $interval ]
then
return
fi
return
} if [ "$rflag" == "" ]
then
check_hbLog
if [ $? == ]
then
echo "=====> ALIVE <====="
else
echo "=====> DEAD <======"
fi
exit
elif [ "$cflag" == "" ]
then
/usr/bin/logger -t heartbeat "kvmheartbeat.sh rebooted system because it was unable to write the heartbeat to the storage."
sync &
sleep
echo b > /proc/sysrq-trigger
exit $?
else
write_hbLog
exit $?
fi
为什么KVM计算机点无故重启?的更多相关文章
- ACPI引起linux系统无故重启
新装机器无故重启多次. centos6 64bit uname -a Linux Eos 2.6.32-71.el6.x86_64 #1 SMP Fri May 20 03:51:51 BST 201 ...
- centos7无故重启-内核升级
最近有一台物理服务器,centos7操作系统,无故重启,每天都会发生这种情况: 解决: 升级内核 CentOS 允许使用 ELRepo,这是一个第三方仓库,可以将内核升级到最新版本,使用ELRepo升 ...
- PowerShell工作流学习-4-工作流中重启计算机
关键点: a)工作流中重新启动计算机,请使用Restart-Computer的Wait参数,Wait参数不仅适用于本地计算机,也适用于远程计算机. b)重启运行工作流的计算机,手工恢复请使用Resum ...
- Win10自动重启原因怎么查Windows10无故自动重启
电脑偶尔自动重启,可能很少用户会在意,若电脑经常无故重启,那么应该怎么办,怎么查找电脑无故自动重启原因呢?下面就以Windows10系统自动重启为例,来查查WIN10无故重启是什么原因导致.百度经验: ...
- kvm 使用入门详解
kvm 是虚拟化技术的一个典型实现,功能非常强大,使用很方便.kvm 本身主要实现对 CPU 的虚拟化,内存和IO的虚拟化使用了开源软件 qemu,qemu 是纯软件层面的虚拟化,其实就是个模拟器.k ...
- Linux 修改计算机名
查看计算机名:在终端输入hostname 修改的话 hostname +计算机名(重启后失效) 要永久修改的话要修改配置文件/etc/sysconfig/network 修改hostname=你要改的 ...
- centos linux中怎么查看和修改计算机名/etc/sysconfig/network
centos linux中怎么查看和修改计算机名 查看计算机名:在终端输入hostname 修改的话 hostname +计算机名(重启后失效)要永久修改的话要修改配置文件/etc/sysconfig ...
- cmd 更改计算机名
bat 更改计算机名 不用重启电脑就生效^_^ @Echo off Color 0A title --更改计算机名 :A cls echo. echo. [0]退出 echo. echo. 不用重启 ...
- KVM之五:KVM日常管理常用命令
1.查看.编辑及备份KVM 虚拟机配置文件 以及查看KVM 状态: 1.1.KVM 虚拟机默认的配置文件在 /etc/libvirt/qemu 目录下,默认是以虚拟机名称命名的.xml 文件,如下,: ...
随机推荐
- 我也说说Emacs吧(7) - lisp基础
lisp基础 Lisp是仅次于Fortran的第二古老的著名计算机语言. Lisp从一开始就与众不同的一点在于,它是基于S-表达式的语言.也就是说,代码和数据是用同一种方式表达出来的. S-表达式,我 ...
- C# Thrift 实战开发 从PLC到Thrift再到客户端集成开发
About Thrift: 本文并不是说明Thrift设计及原理的,直接拿Thrift来开发一个Demo程序,如果想要了解Thrift的细节,可以访问官方网站:https://thrift.apach ...
- iOS9的新特性以及适配方案-----转载
2015年9月8日,苹果宣布iOS 9操作系统的正式版在太平洋时间9月16日正式推出,北京时间9月17日凌晨1点推送. 新的iOS 9系统比iOS8更稳定,功能更全面,而且还更加开放.iOS 9加入了 ...
- Eclipse远程调试Java程序
1. 在服务器上运行jar包时加入参数 -Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=n,address= 2. 在Eclipse中操作 ...
- Cannot setup mail box on Android
Error: “You don’t have permission to sync with this server” Solution: “You have 10 phone partnershi ...
- BZOJ1407 NOI2002 Savage 【Exgcd】
BZOJ1407 NOI2002 Savage Description Input 第1行为一个整数N(1<=N<=15),即野人的数目. 第2行到第N+1每行为三个整数Ci, Pi, L ...
- hadoop2.2使用手册2:如何运行自带wordcount
问题导读:1.hadoop2.x自带wordcount在什么位置?2.运行wordcount程序,需要做哪些准备? 此篇是在hadoop2完全分布式最新高可靠安装文档 hadoop2.X使用手册1:通 ...
- biginteger转Long
pjProDocumentFolderTreeNodeBySeqnoDTO.setFolderId(((java.math.BigInteger) objects[0]).longValue()); ...
- Mac OS安装php-redis扩展
下载php-redis(用于php5.x的版本),地址:https://nodeload.github.com/nicolasff/phpredis/zip/master. 如果是php7.2,选择p ...
- BZOJ4565 [Haoi2016]字符合并
题意 有一个长度为\(n\)的\(01\)串,你可以每次将相邻的\(k\)个字符合并,得到一个新的字符并获得一定分数.得到的新字符和分数由这\(k\)个字符确定.你需要求出你能获得的最大分数. \(n ...