Hi all,

draft-ietf-netmod-system-config-09 - System-defined 
Configuration<https://datatracker.ietf.org/doc/draft-ietf-netmod-system-config/>
 proposes an origin of "system" for nodes in the operational datastore that 
came from the system DS.

But I'm wondering if we may want to consider differentiating between:

  *   Nodes that use the system origin in the original NMDA spec (before this 
new system DS work), but may not actually exist in the new system DS
  *   Nodes that come from the system DS

The way it stands currently, there could be nodes in the operational DS with an 
origin of system that are not in the system DS. Is that confusing?  Or are we 
OK with:

  1.  Any node in the system DS shows up as origin "system" in the operational 
DS, but
  2.  Not all nodes with origin "system" in the operational DS are in the 
system DS

Somewhat related to this issue:  NMDA allows a config true (CT) node to have 
one value in <running> and a different value in <operational>.  The 
<operational> view is supposed to return what is actually in use (which can in 
theory differ from what was configured in running/intended).

If a node "foo" has value 1500 in <intended>, but in operational the value is 
1492, what should the origin be for node foo?  System?
[note in this question/example, node "foo" is not in the system DS. It is just 
a standard config node]

Jason (he/him)

_______________________________________________
netmod mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to