[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference netcad::hub_mgnt

Title:DEChub/HUBwatch/PROBEwatch CONFERENCE
Notice:Firmware -2, Doc -3, Power -4, HW kits -5, firm load -6&7
Moderator:NETCAD::COLELLADT
Created:Wed Nov 13 1991
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:4455
Total number of notes:16761

1819.0. "huge FDDI/ETHERNET configuration questions ?" by MARS03::THOMAS () Fri Dec 23 1994 03:59

	My customer uses a big ETHERNET LAN on a site of 2 km * 2 km, with
about 20 ETHERNET segments that are linked with optical fibers (100/140 and
62.5/125) and LB 100/150. He has some automatic backup links with TRANSLAN at
256 or 512 kb/sec.

	Here is the main part of his topology.

===== 	ETHERNET segment
I	optical fiber
!	256 or 512 kb/sec
B	LB 100/150
T	TRANSLAN
R	DR 90/100/200
			=====	=====	=====
			R	R	R
			I	I	I
			I	I	I
			R	R	R
			=====================
				T	B
				!	I
				!	I
				T	B
		     ================================
			T	R		T	
			!	I		!
			!	I		!
			T	R		T
		=====================   ==============
		T		B		T
		!		I		!
		!		I		!
		!		B		T
		!	=============	==============
		!		 R		R
		!		 I		I
		!		 I		I
		T		 B		R
		=========================================
			T		R
			!		I
			!		I
			T		B
			======		=======


	I want to migrate this ETHERNET LAN to an FDDI + ETHERNET LAN.
The ETHERNET segments that will be changed to FDDI concentrator + FDDI/Ethernet
bridge are replaced by FFFFFF on the topology. 

	I will have an FDDI logical ring, but not a physical one, at the
beginning.
	
	I will use DEChub 900 with DECswitch900-EF, DECRepeater900-FP and
DECconcentrator900-MX.

FFFF 	DECconcentrator900 + DECswitch900-EF
II	62,5/125 Optical fibers used in a FDDI ring
B	DECswitch900 ou LB 100/150
R	DR100,200,90,900

			=====	=====	=====
			R	R	R
			I	I	I
			I	I	I
			R	R	R
			=====================
				T	B
				!	I
				!	I
				T	B
		     FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
				II		B	B	II
				II		I	I	II
				II		I	I	II
				II		B	I	II
		FFFFFFFFFFFFFFFFFFFFFFFF     =======	I	II
				B			I	II
				I			I	II
				I			I	II
				B1			B2	II
			=============		     =======	II
				 R			R	II
				 I			I	II
				 I			I	II
				 B			B	II
		FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
			V		I
			!		I
			!		I
			V		B
			======		=======


	If there is a choice, i want that the spanning tree in this
FDDI/ETHERNET LAN use FDDI links by default and only ETHERNET links for backup
purposes, that is to say that B1 and B2 are in STAND BYE.

	I want that the backup of B1 and/or B2 are automatic.

	Is it possible to configure the LB100/150 and DECswitch900 bridges on
this LAN to have this behaviour ?


	Thank you for your answers.


	MERRY CHRISTMAS and HAPPY NEW YEAR.


	Regards,


	Bernard THOMAS
T.RTitleUserPersonal
Name
DateLines
1819.1NETCAD::SLAWRENCETue Dec 27 1994 10:1033
    
    > If there is a choice, i want that the spanning tree in this
    > FDDI/ETHERNET LAN use FDDI links by default and only ETHERNET links for
    > backup purposes, that is to say that B1 and B2 are in STAND BYE.
    
    That's the default; just plug-n-play :-)
    
    The default port costs on all the DEChub switches is set to 10 for
    Ethernet ports and 4 for FDDI, so the cost to the root will be much
    less on the FDDI link than on the Ethernets, putting the Ethernet ports
    into backup.
    
    A couple of recommendations:
    
    1) Both B1 and B2 in your plan are on one end of a point-to-point link
    that has a bridge at both ends; this is redundant in a network such as
    yours where the bridge/bridge link is a high speed link - you could
    replace these with repeaters.
    
    2) If possible, leave the LB100s out of it or upgrade them to LB150 so
    that you avoid the spanning tree migration issues.
    
    3) Since you have quite a bit of redundancy in this plan, I would
    recommend setting choosing one of the FDDI/Ethernet switches and
    setting its Priority to 0 so that you know which one should be the
    root; that way you can know in advance what the spanning tree topology
    should be (if you don't set the priority it will depend on the MAC
    addresses of the bridges, so you'll find out quickly anyway, but I'm a
    control freak).  Then you can monitor each bridge to see whether or not
    the topology has changed - you want your redundant links to take over,
    but you want to know that your primary has failed so that you can fix
    it.
    
1819.2ISSUE WITH LB100 ???MARS03::THOMASTue Dec 27 1994 10:2414
    
    	Thank you for your answer and your comments.
    
    C1:	B1 and B2 are bridges because the optical fiber length are about 2
    km, so i can't use repeaters.
    
    C2:	What are the spanning migration issues with LB100 in an
    FDDI/ETHERNET LAN ?
    
    C3:	I agree with you about the selection of the root bridge.
    
    	Regards,
    
       	Bernard
1819.3LB100's are too inflexible relative to STP useage....NETCAD::BATTERSBYTue Dec 27 1994 10:3617
    > C2: What are the spanning migration issues with LB100 in an
    > FDDI/ETHERNET LAN ?
         
    The issue is that the LB100 only plays epoch 1 spanning tree,
    that is the old DEC spanning tree algorythm. The LB150 will
    play both the old DEC spanning tree, and the IEEE 802.1d., and
    can be locked in 802.1d mode. The LB100 cannot.
    The HUB bridges DECswitch 900EF (also known as the DB900MX), 
    DECswitch 900EE, and PEswitch 900TX, default to the IEEE 802.1d
    but are capable of autoselecting between that and DEC spanning tree.
    If the extended LAN is not a pure DEC shop, and there are other
    vendor network products that don't implement DEC spanning tree,
    properly, you will end up with an extended LAN which will oscillate
    between epochs causing interrution in network manegement, and normal
    network traffic.
    
    Bob