Hi István,
Our limit is 6 hour, because good raid etc the impact is minimal.
Sample configuration
25 VM's on 2x Dell R515 - 4170HE H700+BBU SATA - DRBD (4x Gbit)
DRBD A - 864 GB used
Local disk - 339 GB used (backup) - QNAP NFS (offside location! 2 versions VM) - 678 GB - 5 hours
DRBD B - 758 GB used
Local disk - 334 GB used (backup) - QNAP NFS (offside location! 2 versions VM) - 668 GB - 6 hours
We also backup in the same time the OS with there own backup solution.
Because we are using DRBD, we can backup to the local disk without interrupt the VM or blocking I/O on DRDB storage.
In the past we got also a lot of locking errors QNAP/NFS/SAMBA etc. so first backup to local and then copy to the QNAP is working perfect for 2 years+
Thanks
-----Oorspronkelijk bericht-----
Van: Pongrácz István [mailto:***@gmail.com]
Verzonden: maandag 24 februari 2014 9:31
Aan: Bart Lageweg | Bizway; 'Lindsay Mathieson'; ProxMox Users
Onderwerp: Re: [PVE-User] Backup to samba share stalling/locking up
Hi Bart,
May I ask, how much time does it take to complete the backup and how much disk space does it use?
Thank you!
Best regards,
István
----------------eredeti üzenet-----------------
Feladó: "Bart Lageweg | Bizway" ***@bizway.nl
Címzett: "'Lindsay Mathieson'"
, "ProxMox Users"
Dátum: Mon, 24 Feb 2014 08:00:30 +0000
-------------------------------------------------
Post by Bart Lageweg | BizwayHi Lindasy,
We are using the rack QNAP on 2 locations with NFS, backup first to
local and copy to QNAP (40+ nodes).
/etc/vzdump.conf
Script: /usr/local/bin/backup-hook.pl
#!/usr/bin/perl -w
# hook script for vzdump (--script option use strict;
# config
my $phase = shift;
my $mode = shift; # stop/suspend/snapshot my $vmid = shift; my $vmtype
= $ENV{VMTYPE}; # openvz/qemu my $dumpdir = $ENV{DUMPDIR}; my
$hostname = $ENV{HOSTNAME}; my $tarfile = $ENV{TARFILE}; my $logfile =
$ENV{LOGFILE}; my $backuplocatie = "/mnt/pve/sharenameNFS/dump";
my %dispatch = (
"job-start" => \&nop,
"job-end" => \&nop,
"job-abort" => \&nop,
"backup-start" => \&nop,
"backup-end" => \&backup_end,
"backup-abort" => \&nop,
"log-end" => \&log_end,
"pre-stop" => \&nop,
"pre-restart" => \&nop,
);
sub upload {
system("mv -f $backuplocatie/vzdump-qemu-$hostname-$vmid.vma.lzo
$backuplocatie/vzdump-qemu-oud-$hostname-$vmid.vma.lzo");
system("cp $tarfile
$backuplocatie/vzdump-qemu-$hostname-$vmid.vma.lzo") == 0 || die
"upload cp $tarfile
$backuplocatie/dump/vzdump-qemu-$hostname-$vmid.vma.tar.lzo
backup-host failed";
print "HOOK: upload " . $tarfile . " to
$backuplocatie/vzdump-qemu-$hostname-$vmid.vma.lzo done\n"; }
sub nop {
# nothing
}
sub backup_end {
upload($tarfile);
}
sub log_end {
# upload($logfile);
}
-----Oorspronkelijk bericht-----
Verzonden: maandag 24 februari 2014 7:42
Aan: ProxMox Users
Onderwerp: Re: [PVE-User] Backup to samba share stalling/locking up
On 24 February 2014 16:38, Lindsay Mathieson
Post by Lindsay MathiesonI'm testing backing up a vm to a USB Drive attached to a Qnap NAS,
shared out via samba.
It consistently stalls at 17% and can't be stopped via the web gui.
The only thing that works is a reboot of the server and the vm has to
be unlocked.
Seeing this in dmesg
CIFS VFS: No writable handles for inode
--
Lindsay
_______________________________________________
pve-user mailing list
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
_______________________________________________
pve-user mailing list
http://pve.proxmox.com/cgi-bin/