I tried pulling over the logic for building ids from 2.1.3, but ran into a bit
of a catch. In 3.1.3 part of the logic was to only set id for existing modeled
proceses, which it did by walking the model. That worked when it could call
Device.getDmdRoot("Process"). But in 2.2, the collector is holding a
DeviceProxy that lacks getDmdRoot.
Either the fix is to create a substiture for the call to getDmdRoot, ie some
sort of remote call to the model, or to alter the logic in
ApplyDataMaps._updateRelationship deletes the id from relmaps when
_dreateRelObject creates the same object id. While I suspect the former is the
more likely, I'm not 100% sure I understand the intent of the changes to sub in
a DeviceProxy and to set classname in HRSWRunMap (thus always building the
object from the ObjectMap and not the id). Thus I'm not sure I can fix this
right without understanding the intended behavior.
Any advice? Am I on the right path here or completely off base? Assuming I'm
on the right path: is this something that is being fixed, or can I get a
pointer on the perfered way to fix it and help, or is this something that has
to wait (less important than other issues or significantly complicated)?
-------------------- m2f --------------------
Read this topic online here:
http://community.zenoss.com/forums/viewtopic.php?p=20474#20474
-------------------- m2f --------------------
_______________________________________________
zenoss-users mailing list
[email protected]
http://lists.zenoss.org/mailman/listinfo/zenoss-users
