Dan Smith wrote:
# HG changeset patch # User Dan Smith <danms@us.ibm.com> # Date 1205785376 25200 # Node ID 802a5fbe87ed0dff07f2dc75622a9e674d891156 # Parent 47f5fc1ac66dad65785dd050d6b4b9fe95948f8a During a migrate, grab xml, do migration, define at the end ...in a more consistent manner that makes offline migration an effective NOP.
Signed-off-by: Dan Smith <danms@us.ibm.com>
+static CMPIStatus complete_migrate(virDomainPtr ldom, + virConnectPtr rconn, + const char *xml) +{ + CMPIStatus s = {CMPI_RC_OK, NULL}; + virDomainPtr newdom = NULL; + + if (virDomainUndefine(ldom) == -1) { + CU_DEBUG("Undefine of local domain failed"); + }
And we're okay with that failing? Not questioning the validity, just confirming intent. Other than that, not much to talk about. A bit of it falls under the category of me not seeing glaring code issues and not knowing enough about migration to question the deeper logic behind the change, although I do like the more coherent switch block. So assuming this CU_DEBUG doesn't also need to do some error handling, I'll +1 this. -- -Jay