On Fri, Mar 9, 2018 at 3:46 PM, Matt Riedemann <[email protected]>
wrote:
On 3/9/2018 6:26 AM, Balázs Gibizer wrote:
The instance-action REST API has already provide the traceback to
the user (to the admin by default) and the notifications are also
admin only things as they are emitted to the message bus by
default. So I assume that security is not a bigger concern for the
notification than for the REST API. So I think the only issue we
have to accept is that the traceback object in the ExceptionPayload
will not be a well defined field but a simple string containing a
serialized traceback.
If there is no objection then Kevin or I can file a specless bp to
extend the ExceptionPayload.
I think that's probably fine. As you said, if we already provide
tracebacks in instance action event details (and faults), then the
serialized traceback in the error notification payload also seems
fine, and is what the legacy notifications did so it's not like there
wasn't precedent.
I don't think we need a blueprint for this, it's just a bug.
I thought about a bp because it was explicitly defined in the original
spec not have traceback so for me it does not feels like a bug.
Cheers,
gibi
--
Thanks,
Matt
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe:
[email protected]?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
__________________________________________________________________________
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: [email protected]?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
