Conference azur::mcc

Hi Bob,

Unfortunatly V1.2 will support only Digital's bridges.  V1.2 will support
both major spanning tree algorithms (DEC LANbridge 100 and IEEE 802.1d).

Future versions of the STM FM will support third party bridges via SNMP.   
I realize this is an issue with many of our customers that have multivendor 
network but we decided that this was all we could do in the V1.2 time frame.

I am working on workarounds for V1.2 to map around bridges in the tree that 
cannot be polled by the bridge AM.   We should be able to get a reasonable
looking map for customers that have mostly Digital bridges.


Bob

    
    
       Hi Bob,
    
    
    
            When you say that you are working on workarounds to 
        build a map, does that mean your going to show third party
        bridges as unsupported bridge or as a reference or what?
    
           You mention that later releases of ELAM will support third 
       party gear via SNMP.  Why can't we do that now?  For instance,
       when building the Spanning Tree Map, couldn't the script
       detect an unsupported bridge and ask the user to specify
       an SNMP management preference and then display the bridge as
       a SNMP IP BRIDGE ICON.This would facilitate1) A true graphical
       map and 2) registrartion of the SNMP device in a one step
       operation. This is just a suggestion.
    
          Please give us the ICONIC something to show the true STM.
    
    
    
       Thanks,
    
        Bob
    
     

    I just had a customer ask me about this.  He was most impressed with
    the the LAN AutoTopology feature, but was quite dismayed to disover
    that it worked for only DEC bridges.
    
    I understand the time constraints limited you to only supporting DEC
    bridges in V1.2; however, I am also interested in the workarounds you
    have in mind.  Could you go into a little more detail?
    
    Thanks,
    Dan

Dan,

The workarounds will attempt to map bridges that can be polled, but can't
be located in the spanning tree because their designated bridge cannot
by polled for its spanning tree attributes.   A bottom up approach will
be used to build these sub-trees.   You still won't know where these
sub-trees fit into the main tree, but at least all of the LANbridges and
DECbridges will be mapped relative to one another. 

In addition, I'm working on a fix for the case when the root bridge cannot
be polled.   This one really isn't too bad as long as there are bridges
directly below the root that can be polled.

Mapping 3rd party bridges via SNMP will have to wait for the next release.

Regards,

Bob

Bob

What's the latest on this?  Like many others, we have currently have a
Vitalink bridge serving as our root, so can not do the whole map exercise.

Thanks!
tl

>  we have currently have a
> Vitalink bridge serving as our root...

    Why ?   In a previous life I had to manage a big E-LAN and I went
    through *all sorts* of contortions to make CERTAIN that a Translan
    would *never* be elected root....

    curious,
    /doug

tl,

The workarounds as described in .4 are implemented in the SSB version of
the STM FM.   Give it a try and let me know how it goes.

Bob

    just tried the spanning tree using SSB kit on Ultrix. 

    Yes We have a Vitalink as a root bridge and yes, we have Vitalink
    lower in the tree.

    The result.  Seems to have found all LAN bridges in the Elan, placed
    the Vitalink as the root bridge, other bridges below it.

    At the point where the second Vitalink was seen another "tree:
    structure started.  Yes, we don't quit see where this fits in relation
    to the whole tree -- as per .4

    This is not bad, all things considered.

    Two questions, all bridges get registered but also we get domain
    (related to the .LAN icons) also registered, is there any way to rename
    these.  The idea is to be able to click on the place where the ,LAN
    icon is and go into a domain, this domain would contain the nodes on
    that segment -- say.

    2nd Q, how did the algorithm work out that there is a loop from the
    root segment to a lower segment ? ...
                         +---------------+
                    08007c070e40         |
                         |               |
                    08002b046374    08002b08e016
                         |               |
                    08002b02f750         |
                         |               |
                         +---------------+
                         |

Re .7

Bob, worked for me, too.

Thanks!
tl