LURemoveInstance: fix op.ignore_failures usage
authorIustin Pop <iustin@google.com>
Fri, 30 May 2008 10:51:22 +0000 (10:51 +0000)
committerIustin Pop <iustin@google.com>
Fri, 30 May 2008 10:51:22 +0000 (10:51 +0000)
commitf074ef73047a4856217b6ef8864d05474586b317
treefe08a80c9d2e1df7fce189ada80618d12f4d0b27
parent48992b07ef07eab075c73b6d2a47bf402171eccc
LURemoveInstance: fix op.ignore_failures usage

Currently: the LURemoveInstance.Exec() method uses the ignore_failures
attribute of the OpRemoveInstance opcode, but it doesn't check for its
existence. The patch adds this attribute to _OP_REQP and to all the
places where this opcode was created.

This attributes is always passed by gnt-instance, but burnin didn't pass
it so it can fail if it enters the 'fail to remove disks' branch of the
method (which is why it was not triggered until now).

Reviewed-by: ultrotter, imsnah
lib/cmdlib.py
tools/burnin