SQE required on Sun network?

[Lap Yun Yau]

Should SQE be turned on or off in Sun network?

In our network, we have more than 100 Suns including a lot of SPARCstation
1 & 1+, 3/60s, 3/260s, 4/110s, 3/280, 4/280s, 4/490s, Solbourne Series 5,
and PCs.  All these machines are hooked up to form 1 ethernet network.
Cabletron equipments like the multiport repeater, multiport transceivers,
and the multi-media access center (a network concentrator) like this
diagram:

                          +------------+
                          | multiport  |
                          | transceiver|
                          +------------+
                           |   |   |
               ------------+   |   +-------------------------+
               |               |                             |
        +------+-------+  +----+---------+          +---------------+
        | multiport    |  | multiport    |          | multi-media   |
        | repeater     |  | repeater     |          | access center |
        +--------------+  +--------------+          +---------------+
         |  |   |   |       |   |   |   |             |   |   |   |
         Suns & PCs        Suns & PCs                Suns & PCs

All transceiver (Cabletron ST-500-03) that are hook to the Suns have SQE
turned on.  Recently we expanded our office and installed many new thin
ethernet cables to the new office and the people who installed the cable
(again, they are from Cabletron) suggested that we should turn off SQE on
the ST-500 to get a better performance.  So we have turned off SQE on the
new machines.  Now we have a mix of transceivers with and without SQE.

QUESTION:  Should SQE be on or off?  Is it "legal" to mix machines with
SQE on and machines with SQE off?  (I guess it is violating 802.3 rules)
If there is a mix, will it cause any trouble to the Sun network?

Of the 14 servers, 4 of the 4/490 crashed today without any warning.  When
a server crashed, all of the  other servers hung with the message:

		ie0: no carrier
		ie0: Ethernet jammed

Anybody know what happened?  After we reboot the problem server, the other
servers then get the ok message.  Is it possible that it is caused by SQE
or the MMAC.  Please send me your comment by email, thanks.

uunet!smosjc!lapyun