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

Conference 7.286::fddi

Title:	FDDI - The Next Generation

Moderator:	NETCAD::STEFANI

Created:	Thu Apr 27 1989
Last Modified:	Thu Jun 05 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	2259
Total number of notes:	8590

1459.0. "FDDI on two nodes won't work" by NWD002::GOODELL_MA (Marianne Goodell 801-565-3174) Thu Sep 29 1994 12:40

I have a situation at a customer site where we are having problems bringing
up our FDDI network.  Problem definition:

* The customer has ethernet cards and FDDI cards in each system.

* This is a three-node cluster:  7620 (MRISF1), 7620 (MRISF2 area router),
  7630 (MRISF3).

* The customer is currently using their ethernet cards.  We are not having
  any network problems while using ethernet.

* To use the FDDI, the customer performs the following steps on one of the
  three nodes:

     -- Shutdown PATHWORKS
     -- define circ and line states off for the ethernet controller
     -- define circ and line states ON for the FDDI controller
     -- set the exec state off
     -- through SYSMAN run startnet for that one node
     -- start PATHWORKS
     -- FDDI on this first node seems to work fine

* The customer then repeats the same steps on the second node.  Within 
  minutes, communication on this second node grinds to a halt.  LAT com-
  munications get slower and slower.  The existing node cannot see the
  new node.  PATHWORKS, of course, does not start.

Does anybody out there have any ideas as to why this is happening?  My
customer is anxious to have this problem resolved.

Marianne

T.R	Title	User	Personal Name	Date	Lines
1459.1		NPSS::MDLYONS	Michael D. Lyons - Young enough and dumb enough	`Thu Sep 29 1994 13:57`	5
	It sounds very much like you have not completely shut down the all the protocols using that MAC address, and you have duplicate MAC addresses on both interfaces. Look at the interfaces and check the MAC address. If it's the same, then shut down the network on the ethernet interface and see which protocols are still active on that interface.
1459.2		NPSS::MDLYONS	Michael D. Lyons - Young enough and dumb enough	`Thu Sep 29 1994 14:00`	1
	P.S. Did you shut down LAT, for example?
1459.3	LAT & Decnet were stopped...	NWD002::GOODELL_MA	Marianne Goodell 801-565-3174	`Thu Sep 29 1994 15:22`	10
	Yes. I got more of the story... Another Digital consultant came on-site last week, rebooted the systems and had commented out LAT and DECNET in the system startup file. He then turned on the FDDI controller and started them both up manually. Any more ideas? Also, excuse my ignorance, but how do we check the MAC address to see if it's duplicate on the interfaces? Thanks, Marianne
1459.4		NPSS::MDLYONS	Michael D. Lyons - Young enough and dumb enough	`Thu Sep 29 1994 18:47`	63
	...assuming VMS from your previous comments: 1) run SDA ($ Ana/sys 2) Show lan/dev=(device) ...you will eventually get to a screen something like: LAN Data Structures ------------------- -- XQA Device Information 29-SEP-1994 18:38:44 -- LSB address 80509820 Active unit count 4 LAN version 00000006 05052025 Driver version 0A000001 05050006 LAN code address 80503999 Driver code address 80500C18 Device name XQ_DELQA Device type 33 Device version 00000000 00000000 DLL type CSMACD Data chaining ON All multicast state OFF Controller mode NORMAL Promiscuous mode OFF CRC generation mode ON Hardware mode 0000 Physical address AA-00-04-00-15-FB Hardware address 08-00-2B-11-BC-CB Flags: 0000 Characteristics: 0000 Status: 0013 Inited,Run,Timer DAT stage 00000000 DAT xmt status 0000001A 001A0001 DAT number started 1 DAT xmt complete 12-JUL 13:47:40 DAT number failed 0 DAT rcv found None ...If the Physical Address field is one of the AA-00-04-xx-xx-xx entries, and matches for both interfaces, then that is your problem. ...continuing along: LAN Data Structures ------------------- -- XQA Multicast Address Information 29-SEP-1994 18:40:42 -- AB-00-00-04-00-00 FF-FF-FF-FF-FF-FF 09-00-2B-00-00-0F 09-00-2B-02-01-04 09-00-2B-02-01-07 -- XQA Unit Summary 29-SEP-1994 18:40:47 -- UCB UCB Addr Fmt Value Client State --- -------- --- ----- ------ ----------- XQA0 805091E0 XQA2 805329C0 Eth 60-03 DECNET 0017 Strtn,Len,Uniq,Strtd XQA3 80550BF0 Eth 08-00 TCPIP 0015 Strtn,Uniq,Strtd XQA4 80550EB0 Eth 08-06 ARP 0015 Strtn,Uniq,Strtd XQA5 80563F00 Eth 60-04 LAT 0015 Strtn,Uniq,Strtd ^ \| ...look here ---------------------------/ to see which protocol is still operating...
1459.5	May have to purge and clear Ethernet line and circuit	SSDEVO::PARRIS	RAID-5 vs. RAID-1: n+1 << 2n, in $$$	`Thu Sep 29 1994 19:23`	11
	> -- define circ and line states off for the ethernet controller > -- define circ and line states ON for the FDDI controller Should probably be: -- purge and clear circ and line info for the ethernet controller -- define and set circ and line info for the FDDI controller I've seen reports that DECnet will change the address of a LAN controller from the 08-00-2B-* form to the AA-00-04-* form when it starts up, even if the line and circuit state are off, and that you need to purge and clear the info in DECnet before it will leave it alone.
1459.6	Bad controller :(	NWD002::GOODELL_MA	Marianne Goodell 801-565-3174	`Wed Nov 09 1994 12:23`	7
	It finally ended up being another bad FDDI controller in the VAX. We were doing most of the things recommended in this note at one time or another with no success. Finally, we tried a new controller and we're running on our FDDI now. Thanks for all of your help. Marianne