[pve-devel] [RFC v2 qemu-server 4/5] adapt to new storage_migrate activation behavior

[Fabian Ebner]

Offline migrated volumes are now activated within storage_migrate.
Online migrated volumes can be assumed to be already active.

Signed-off-by: Fabian Ebner <f.ebner at proxmox.com>
---

dependency bump needed

Sent as RFC, because I'm not completly sure if this is fine here.
Is the assumption about online volumes correct or is there some weird
edge case I'm missing?
I only found run_replication as a potential place that might need active
local volumes, but that also uses storage_migrate in the end.

 PVE/QemuMigrate.pm | 8 --------
 1 file changed, 8 deletions(-)

diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
index 2f4eec3..f2c2b07 100644
--- a/PVE/QemuMigrate.pm
+++ b/PVE/QemuMigrate.pm
@@ -251,7 +251,6 @@ sub prepare {
 
     my $vollist = PVE::QemuServer::get_vm_volumes($conf);
 
-    my $need_activate = [];
     foreach my $volid (@$vollist) {
 	my ($sid, $volname) = PVE::Storage::parse_volume_id($volid, 1);
 
@@ -266,16 +265,9 @@ sub prepare {
 	    my $plugin = PVE::Storage::Plugin->lookup($scfg->{type});
 	    warn "Used shared storage '$sid' is not online on source node!\n"
 		if !$plugin->check_connection($sid, $scfg);
-	} else {
-	    # only activate if not shared
-	    next if ($volid =~ m/vm-\d+-cloudinit/);
-	    push @$need_activate, $volid;
 	}
     }
 
-    # activate volumes
-    PVE::Storage::activate_volumes($self->{storecfg}, $need_activate);
-
     # test ssh connection
     my $cmd = [ @{$self->{rem_ssh}}, '/bin/true' ];
     eval { $self->cmd_quiet($cmd); };
-- 
2.20.1