WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-cim

[Xen-cim] CMPI support of arrays in OUT params

To: xen-cim@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-cim] CMPI support of arrays in OUT params
From: Michael Johanssen <JOHANSSN@xxxxxxxxxx>
Date: Fri, 26 Jan 2007 16:55:50 +0100
Delivery-date: Fri, 26 Jan 2007 07:55:19 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-cim-request@lists.xensource.com?subject=help>
List-id: xen-cim mailing list <xen-cim.lists.xensource.com>
List-post: <mailto:xen-cim@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-cim>, <mailto:xen-cim-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-cim>, <mailto:xen-cim-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-cim-bounces@xxxxxxxxxxxxxxxxxxx
> It appears CMPI (1.3 or 2.0) does not support extrinsic method OUT params
that
> are arrays.  For the VirtualSystemManagementService provider, arrays of
> references are problematic.  AddResourceSettings() and
ModifyResourceSettings()
> both have OUT params that are reference arrays.  OpenWBEM does not
support this
> - segfaults when calling something like:
>
> CMAddArg(argsout, "OutSettings", (CMPIValue*)&array, CMPI_refA);
>
> I have not tried Pegasus but would expect the same -- well, hopefully it
fails
> more gracefully :-).
>
> I do not plan to support AddResourceSettings, ModifyResourceSettings and
> RemoveResourceSettings in CIM_VirtualSystemManagementService and instead
> support AddResourceSetting, ModifyResourceSetting and
RemoveResourceSetting
> (notice the singular setting) in our derived management service class.  A

> single reference will be returned in these versions, side-stepping the
above
> issue.
>
> Frankly, the plural versions suck anyway :-).  I came to this conclusion
while
> implementing AddResourceSettings. Implementation is difficult,
particularly
> handling errors.  Imagine successfully adding 2 resources but failing on
the
> 3rd.  Do you fail the whole thing and attempt to unwind (and perhaps fail
doing
> that!)? Or do you proceed forward and indicate half failure?  I don't
like it
> and considered the 'one-at-a-time' approach before hitting the 'output
array' >
> issue.
>
> I will proceed with this approach unless (strongly) convinced otherwise.
>
> Jim

Jim,

in DMTF SVPC we modelled that way
to enable the client to pass in multiple
logically related requests in one step,
that either succeeds or fails as a whole.
Admittedly that makes things more difficult
for the implementation, but consider that
with the single step approach you leave
the responsiblity for cleanup to the client,
and if he does not care about that responsibility,
then you end up with inconsistent system definitions.

Of course the issue with cmpi needs to be addressed.

Michael


_______________________________________________
Xen-cim mailing list
Xen-cim@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-cim

<Prev in Thread] Current Thread [Next in Thread>