[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

3387.0. "Repeater Redundancy Question" by MFRNW1::SCHUSTER (Karl Schuster @RTO OMS Munich Germany) Thu Mar 21 1996 05:26

Note #2085 is not very clear no me.

Configuration:
1 90FS (with the masterports) linked via 2 fibrelinks to 2 different 900FP
(with a responderport each):

Is it a MUST, that the 900FPs are on the same LAN-segment(no bridging between
), or can they be on 2 different LAN-segments (segmentation via bridge).

The 900FP documentation says, that the responder ports can be on an
"extended LAN" - but what is an "extended LAN" ?

Who can help ?

Regards,
Karl
T.RTitleUserPersonal
Name		DateLines
3387.1NPSS::WADENetwork Systems SupportThu Mar 21 1996 09:096
    
    The responders need to be on the same LAN or extended LAN (which generally
    refers to a LAN that has been extended with bridges).
        
    bill
3387.2Now you've got me confused ...SNOFS1::KHOOJEANNIEHumpty Dumpty was pushedFri Mar 22 1996 01:2515
    Hi Bill
    
    This is contradictory to what is in 2085.9, which suggests that it
    doesn't matter whether the responders are on the same LAN/ext LAN or
    not.
    
    I have been involved in a number of designs where the responders were
    not in the same LAN, and so far, have not heard any complaints of the
    redundancy not working.
    
    Can you explain further?
    
    Regards
    Jeannie
3387.3MFRNW1::SCHUSTERKarl Schuster @RTO OMS Munich GermanyFri Mar 22 1996 07:1515
    Today I did tests with the failover, as described in .0.
    
    The failover works pretty good. The failover of the physical connection
    is immediate.
    But the following must be kept in mind:
    As there is bridging between the 2 responders, it lasts a significant
    amount of time until the network traffic fails over. In my environment
    it was between 20 and 50 seconds. The bridge forwarding tables have to
    learn the new configuration.
    
    It seems to work, but:  Is this an OFFICIALLY SUPPORTED and RECOMMENDED 
    configuration ?
    
    Regards,
    Karl
3387.4NPSS::WADENetwork Systems SupportFri Mar 22 1996 11:3837
    re .2
    
    Geezzzzz
    
    We are talking about a repeater being configured with a master port
    pair for the sole purpose of providing redundancy, right?
                                                      
   >     This is contradictory to what is in 2085.9, which suggests that it
   >     doesn't matter whether the responders are on the same LAN/ext LAN or
   >     not.
   >      
   >     I have been involved in a number of designs where the responders were
   >     not in the same LAN, and so far, have not heard any complaints of the
   >     redundancy not working.
   >      
    
    
    These are "REDUNDANT" links!  If they don't go to the same LAN or
    		   ---------
                                           
    
    ELAN what is the point?  The primary is the only link that is up
    whether the ports connect to two different LAN/ELANs (with no
    repeater, bridge, widget connecting them) or not.  If they are not
    connected then the stations on the other end of the secondary better be
    happy just talking among themslves because that's all they can reach.
                                                     
    There is no signaling that occurs between the two responder ports so 
    there is nothing that requires them to be locating on the same
    physically net.  It just doesn't make any sense, if we are talking 
    redundancy, for the other ends of the master port pair to be connected 
    to physically disjoint Ethernets.
    
    bill