[pve-devel] [PATCH qemu-server 1/2] migration: avoid crash with heavy IO on local VM disk

Fiona Ebner f.ebner at proxmox.com
Tue May 28 10:01:07 CEST 2024


Am 23.05.24 um 11:08 schrieb Fabian Grünbichler:
> one nit below, otherwise very nice to have this!
> 
> On May 3, 2024 10:34 am, Fiona Ebner wrote:
>> diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
>> index 8d9b35ae..649cfec4 100644
>> --- a/PVE/QemuMigrate.pm
>> +++ b/PVE/QemuMigrate.pm
>> @@ -1142,6 +1142,15 @@ sub phase2 {
>>  	    $self->log('info', "$drive: start migration to $nbd_uri");
>>  	    PVE::QemuServer::qemu_drive_mirror($vmid, $drive, $nbd_uri, $vmid, undef, $self->{storage_migration_jobs}, 'skip', undef, $bwlimit, $bitmap);
>>  	}
>> +
>> +	my $kvm_version = PVE::QemuServer::kvm_user_version();
> 
> wouldn't this need to check the *running* qemu binary for the migrated
> VM, not the *installed* qemu binary on the system?
> 

Good catch! Yes, I copied/adapted this from the check for replication
which also does it wrong. I'll send a v2 fixing it (and remove the check
for replication for version 4.2 which is not needed nowadays anymore).




More information about the pve-devel mailing list