On 09/14/2016 10:40 AM, Daniel P. Berrange wrote:
On Wed, Sep 14, 2016 at 10:37:07AM -0400, Corey S. McQuay wrote:
Currently Libvirt allows attempts to migrate read only disks. Qemu cannot 
handle this as read only
disks cannot be written to on the destination system. The end result is a 
cryptic error message
and a failed migration.

This patch causes migration to fail earlier and provides a meaningful error 
message stating that
migrating read only disks is not supported.

Signed-off-by: Corey S. McQuay <csmcq...@linux.vnet.ibm.com>
Reviewed-by: Jason J. Herne <jjhe...@linux.vnet.ibm.com>
Reviewed-by: Boris Fiuczynski <fiu...@linux.vnet.ibm.com>
  src/qemu/qemu_migration.c | 25 +++++++++++++++++++++++++
  1 file changed, 25 insertions(+)

diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c
index e451ef6..3311711 100644
--- a/src/qemu/qemu_migration.c
+++ b/src/qemu/qemu_migration.c
@@ -2392,6 +2392,28 @@ qemuMigrationIsSafe(virDomainDefPtr def,
      return true;

+static bool
+qemuMigrationAreAllDisksRW(virDomainDefPtr def,
+                    size_t nmigrate_disks,
+                    const char **migrate_disks)
+    size_t i;
+    for (i = 0; i < def->ndisks; i++) {
+        virDomainDiskDefPtr disk = def->disks[i];
+        if (qemuMigrateDisk(disk, nmigrate_disks, migrate_disks) &&
+            disk->src->readonly) {
+            virReportError(VIR_ERR_OPERATION_UNSUPPORTED,
+                       _("Cannot migrate read-only disk %s"),
+                       disk->dst);
+            return false;
+        }
+    }
+    return true;

We already have multiple places in the migration code which iterate
over all disks, determining which are migratable. IMHO we should
just add an readonly check in one of those, rather than adding yet
another iteration.

Hi Daniel,

I'm not seeing a suitable existing location for this new check to live.
The only place I see migration code loop over the disks for error checking
is in qemuMigrationBeginPhase.

for (i = 0; i < nmigrate_disks; i++) {
    for (j = 0; j < vm->def->ndisks; j++) {
        if (STREQ(vm->def->disks[j]->dst, migrate_disks[i]))

    if (j == vm->def->ndisks) {
                       _("disk target %s not found"),
        goto cleanup;

And putting inside a nested loop would be kind of silly :)
It seems to be that all other disk loops are in locations
that do not run before migration begins, or their purpose
is not for error checking.

It may make sense to add the check to either of the following:

But I have not researched those enough yet to know for sure.
I'll look into these on Monday, unless someone can tell me in
the meantime if this makes sense or not?

-- Jason J. Herne (jjhe...@linux.vnet.ibm.com)

libvir-list mailing list

Reply via email to