kvm常見故障及解決方案
- 2020 年 2 月 1 日
- 筆記
一、啟動虛擬機Connection reset by peer
# virsh start vmhost1error: Failed to start domain vmhost1error: Unable to read from monitor: Connection reset by peer
在虛擬機運行過程中關閉宿主伺服器就有可能導致這種情況出現，由於宿主伺服器中的kvm虛擬機控制器與安裝在kvm中的虛擬機會話被異常重置，所以我們可以如下解決：
# virsh managedsave-remove vmhost1# virsh start vmhost1
如果啟動查看/var/log/libvirt/qemu/vmhost1.log下log還報如下錯誤：
Cannot set up guest memory 'pc.ram': Cannot allocate memory
這個問題可能是分配給vmhost1分配的記憶體過大（甚至超過的物理主機的記憶體大小），或者可能是宿主機沒有足夠的記憶體分配給此虛擬機，導致無法啟動！
二、重Define虛擬機時無/usr/bin/kvm
error: Failed to define domain from hostname.xmlerror: Cannot find QEMU binary /usr/bin/kvm: No such file or directory
解決方法：
# ln -s /usr/libexec/qemu-kvm /usr/bin/kvm
三、error: internal error process exited while connecting to monitor
# virsh start vmhost1
error: Failed to start domain vmhost1error: internal error process exited while connecting to monitor: kvm: -drive file=/dev/sp1368155439693/v1368544020461,if=none,id=drive-virtio-disk0,format=qcow2: could not open disk p_w_picpath /dev/sp1368155439693/v1368544020461: Invalid argument
分析：鏡像格式錯誤，用qemu-img info 檢查鏡像和xml配置文件中指定的type是否一致！
四、Unable to load library 'virt': libvirt.so
Unable to load library 'virt': libvirt.so: cannot open shared object file: No such file or directoryLinux下解決： ln -s /usr/lib/libvirt.so.0 /usr/lib/libvirt.so windows下解決： 將libvirt-0.dll改名為virt.dll
五、error: Refusing to undefine while domain managed save p_w_picpath exists
# virsh undefine vmhost1error: Refusing to undefine while domain managed save p_w_picpath existshttp://www.redhat.com/archives/libvir-list/2011-July/msg01219.html
解決方法：virsh undefine $domain –managed-save
六、啟動libvirtd進程出錯
# /usr/local/sbin/libvirtd -d -l --config /usr/local/etc/libvirt/libvirtd.conf （編譯安裝的啟動方式）
error：/usr/local/sbin/libvirtd: initialization failed try to install libpcap-devel RPM and rebuild libvirt http://comments.gmane.org/gmane.comp.emulators.libvirt/58218apt-get install libpcap-dev 上面的方法好像都沒有效果，但是嘗試了http://wiki.libvirt.org/page/The_daemon_cannot_be_started說的，把配置文件里的 listen_tls = 0注釋取消（更奇怪的問題，在我的客戶端鏈接不對）
七、啟動虛擬機報錯
# virsh start vmhost1error: Failed to start domain vmhost1error: internal error process exited while connecting to monitor: Could not access KVM kernel module: No such file or directoryfailed to initialize KVM: No such file or directoryNo accelerator found!
上面的提示資訊就是因為QEMU在初始化階段因為無法找到kvm內核模組。
# modprobe kvm #載入指定的模組
重啟電腦，進入bios介面，設置advance選項裡面的virtualization標籤為Enabled
通過命令 lsmod | grep kvm #顯示已載入的模組
八、虛擬機遷移
# virsh migrate --live 1 qemu+tcp://192.168.0.121 --p2p --tunnelled --unsafe error: operation failed: Failed to connect to remote libvirt URI qemu+tcp://192.168.0.121（在URI後面加上/system，『system』相當於root用戶的訪問許可權) #virsh migrate --live 2 qemu+tcp://192.168.0.121/system --p2p --tunnelled error: Unsafe migration: Migration may lead to data corruption if disks use cache != none（加上--unsafe參數）#virsh migrate --live 2 qemu+tcp://192.168.0.121/system --p2p --tunnelled --unsafe error: Timed out during operation: cannot acquire state change lock (啟動虛擬機有時也會遇此錯誤）,需要重啟libvirtd進程
九、virsh
error: Failed to connect socket to '/var/run/libvirt/libvirt-sock': Connection refused(libvirtd 進程沒有啟動，libvirtd是一個監聽客戶端請求的進程）# virsh -c qemu:///system listerror: Failed to connect socket to '/var/run/libvirt/libvirt-sock': Permission denied error: failed to connect to the hypervisor （當前用戶沒有許可權，修改/etc/libvirt/libvirtd.conf,unix_sock_rw_perms = 0777,使所有用戶都有許可權讀寫）
轉：http://openskill.cn/article/146
