Re: Ideally BER should be a customer controlled option
- To: bill.st.arnaud@xxxxxxxxxx
- Subject: Re: Ideally BER should be a customer controlled option
- From: Rich Taborek <rtaborek@xxxxxxxxxxxxxxxx>
- Date: Mon, 31 May 1999 01:04:56 -0700
- CC: HSSG_reflector <stds-802-3-hssg@xxxxxxxx>
- Organization: Transcendata, Inc.
- References: <NBBBJIMEPHPGCNGAHPMFIEMOEMAA.bill.st.arnaud@xxxxxxxxxx>
- Reply-To: rtaborek@xxxxxxxxxxxxxxxx
- Sender: owner-stds-802-3-hssg@xxxxxxxxxxxxxxxxxx
I have to disagree with you on this one.
Ethernet is a highly successfull data transport in part because of its
inherent simplicity and scalability. The simplicity, in turn, results
from developing standards with very few options. I'm afraid that making
BER a customer controlled option is a step in the wrong direction.
As for scalability, the BER, being essentially proportional to the frame
error rate due to transmission link noise, has scaled along with
increases in throughput. It only makes sense to scale BER
objectives/frame error rate objectives for 10 GbE.
Setting a BER objective of 10 E-13 may be reasonable and cost effective
for the following reasons:
The component whose inherent BER affects the BER of the system in which
is employed the most is the transceiver. These complex components have
routinely been tested to meet BER's of approximately 10 E-12 since the
days of FDDI and ESCON (tm) where bit rates were in the range of 100-200
Mbps. BER testing of transceivers added considerably to their cost at
that time. BER test methods have been cost optimized considerably since
then. I am assuming that testing of 10 GbE transceivers to a BER of 10
E-13 rather than 10 E-12 would not add significantly to their costs.
10 GbE will be 'super-backbone' technology for some. It may also have
new applications as a WAN feed or even a long-haul protocol in some
instances. BERs greater than their GbE and Fast Ethernet feeds surely
are not in line with good engineering practice. This suggests that 10
GbE BER objectives be increased from those of GbE.
Bill St. Arnaud wrote:
> Realizing that there are some practical limitations, as much as possible,
> BER should be a customer controlled option.
> If I am running only IP with high number of a TCP transmissions I may
> deliberately want a high BER to act as layer 1 WRED. Also I may be able to
> push my repeater distance using low cost laser and a willing to suffer a
> higher BER.
> If I am running some other protocol I may require a lower BER.
> I always believe in giving the customer as much choice as possible. I don't
> we should play god and decide before hand what is the best BER for our
> Bill St Arnaud
> Director Network Projects
Richard Taborek Sr. Tel: 650 210 8800 x101 or 408 370 9233
Principal Architect Fax: 650 940 1898 or 408 374 3645
Transcendata, Inc. Email: rtaborek@xxxxxxxxxxxxxxxx
1029 Corporation Way http://www.transcendata.com
Palo Alto, CA 94303-4305 Alt email: rtaborek@xxxxxxxxxxxxx