Search results

Shouldn't QemuServer.pm be in pve-manager package?

Is it going to enter stable repository any time soon?

I'm no expert in kvm but I don't think that would be reliable. kvm process may be sleeping if guest is not doing anything and AFAIK there is no /proc or /sys interface for kvm that You could check.

I don't think that You can check if it's paused or running without communicating with kvm monitor. What method are You using to talk to monitor? Do You use QMP (http://www.linux-kvm.org/page/MonitorProtocol)?

Tested and after settings rerror and werror it works as expected, I've run badblocks on my virtual machine and disconnected iscsi several times, each time my virtual machine was paused and badblocks did not found any errors.

Using multipath seems to fix iscsi device renumbering, device mapper maps to multipath device and multipath daemon can handle device name (and so minor/major number) changes. But I'll need to test it more tomorrow.

Regarding iscsi disk getting renamed from sda to sdb on iscsi restart or reconnect, I can fix that by hand by doing: # stop iscsi /etc/init.d/open-iscsi stop # remove all device mapper devices dmsetup remove_all # start iscsi /etc/init.d/open-iscsi start # rescan lvm using pvesm pvesm list...

P.S. this is very simple patch that adds werror and rerror options to drive (just list cache or snapshot), I will post more complete version that makes werror=stop default once I test it all.

also qm list shows kvm guest as running if it is paused, it does not check it's status using kvm monitor?

Patch for QemuServer.pm Patch for qm executable I've tested it and it works, disconnecting my iscsi storage and waiting until I/O errors occur makes my kvm guest paused, resuming it after fixing storage makes it work again without data loss, I only got warnings about tasks hanging too...

It's seems like a simple change and this is a very important feature for me so I'll try to make one.

Great, is it going to be in 2.0 or will it be added earlier?

I think that such great and hidden feature should be enabled by default (if it's safe) with possibility to disable it per vm. I don't see any need to be able to disable it per disk.

It seems that this feature is simply undocumented but present. I see rerror and werror options in kvm-0.12.4 source code and there are some references to those options in kvm changelog (http://www.spinics.net/lists/kvm/msg29883.html). Can support for those options be added to qm?

Hello, according to http://www.centos.org/docs/5/html/5.5/Technical_Notes/kvm.html using -drive "werror=stop" option virtual machine would be paused on I/O error, but I can't find that werror option in kvm or qm manual, adding werror=stop to ide0 in config file does not work, is it only...

Boot Your virtual machine using live cd (http://www.sysresccd.org for example) , mount Your partition and fix it.

It triggers an error and snapshot is not created. https://www.redhat.com/archives/linux-lvm/2010-January/msg00007.html http://archives.free.net.ph/message/20081014.081254.df482534.en.html

I don't think that would work. AFAIK if lvm volume group is going to be shared among it cluster flag must be set, lvm will treat it as shared and synchronize metadata across cluster. The trick with disabling this flag works because clustering is required only for lvm metadata sync so If You...

CLVM does not support snapshots which are very usefull for backup, do You have any workaround for this problem? The only solution I found is to disable clustering on LVM volume group before doing snapshot and re-enable it after it's done but I'm not sure if it's safe when few CLVM cluster...

Google to the rescue? It is really not that hard to paste error text to the search box. http://forum.proxmox.com/threads/2760-MMU-problem-for-memory-sharing-and-ballooning.