X.25 MIB changes

["Dean D. Throop" <[e-mail]>]

As I study the X.25 MIB I've noticed some things the MIB doesn't 
handle well.  

We need a value for the X.25 Mode to specify XID negotation.  This 
means that X25AdmnInterfaceMode would have values for dte, dce, and 
dxe.  

We need to allow values of zero for Lowest and Highest channel 
numbers in the X25 channel table if no channels are configured for 
a range.  

We need to clarify that the channel table refers only to 
configuration values and not the current operating values.  This 
means there is no way to read the current operating value if the 
configuration values were changed after the interface was started.  
(We could remove the x25channel table and put the objects in the 
Admn and Oper tables but this is a rather major change at this late 
date).  

Unless someone has a better idea, I'll make the above changes to 
the X.25 MIB.  



I've also noticed the MIB doesn't have objects for some commonly 
required configuration information.  The following are some 
observations I've noticed but I won't add any objects unless I get 
a couple of people saying they are needed.  


The X.25 MIB includes the number of data packets received but
doesn't have the complimentary number of data packets sent (but
the ISO draft for X.25 does have it).

The last draft of the X.25 MIB has a default maximum packet size.  
This means we don't have objects for the maximum size the PLE will 
accept.  

We don't have an object for whether or not to use a flow control 
facility.  

We don't have objects to tell when Closed User Groups, Closed User 
groups with out Going Access, or Bilaterial Closed User group 
facilities are allowed.  Nor whether they are the short or extended 
version of the facility.  

We don't have objects for the max & min, in & out throughput class 
objects to negotiate.  

One last thing:

The last draft of the X.25 MIB allowed the CircuitTable to be 
written to configure PVCs (but only when the interface is down).  It 
might be be better to add a separate table to configure PVCs.  This 
would make the circuit table read-only and always refer only to 
existing circuits.  This requires adding a lot of objects but 
simplifies what the objects refer to.  I doubt it's worth it but 
I'll listen for your requests.


Any comments on the above list of things?

Dean Throop		[e-mail]