[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: not much throughput



     On Sat, 1 Jan 2011 20:56:44 -0500 grarpamp <grarpamp@xxxxxxxxx> wrote:
To: tor-relays@xxxxxxxxxxxxxx
X-To-Get-Off-This-List: mail majordomo@xxxxxxxxxxxxxx, body unsubscribe tor-relays

>>> > that advertises 20KB/s. Probably part of that has to do with selecting
>>> > your 'bandwidthburst' at 20KB also.
>>> level, I have 1.5 Mbit/s download and 384 kbit/s upload,
>>> up is 10 Mbit/s down and 1 Mbit/s up.
>>> The web server gets some hits from robots, but I have seen 20 kB/s
>> 384/8 = 48KB/s. So my first suggestion would be to set your BandwidthBurst
>
>Can we all use correct bandwidth units please?

     Ahem.

>384kbps / 8 = 48kbps, NOT 48KBps [1] (or its equivalent: 384000bps)
>Further, bandwidth is, at least in the US, commonly sold and measured
>in bits per second, in multiples of decimal thousands. ie: b, kbps, Mbps, Gbps.
>Not in powers of two of disk storage space. ie: B, KiB, MiB, GiB.
>And please don't mix decimal with binary (kB or Kib).

     Frankly, the above is all bullshit, IMO.  I've been dealing with bytes
for over 43 years and was dealing with decimal-based machines before that.
Bytes were understood from the start to be measured in powers of 2, whereas
decimal-based systems used strictly decimal numbers and terminology.  The
Greek prefixes "kilo-" and "mega-" were understood to be 10^3 and 10^6,
respectively, for the decimal-based systems, but for the byte-based systems
were the powers of 2 that were closed to those powers of 10 used in the
decimal-based systems.  I.e., a kilobyte was 2^10 = 1024 bytes, and a megabyte
was 2^20 = 1048576 bytes.  No distinction was ever made as to where said bytes
were located (e.g., core (RAM), magnetic disk, magnetic tape, magnetic drum,
transmission rates for multiplexor, selector, or block multiplexor channels,
unit-record devices, burst-mode devices).
     The only grey area was bits, which, because they applied to storage of
any base, were often treated using the decimal-based terminology.  This
mattered little because the only time they were normally referred to in "large"
quantities was as "bits audible/s" or "baud".  A problem rarely discussed is
the correlation between byte rates and bit rates, even when not using the
Greek prefixes, namely, whether a byte rate includes parity bits and, if so,
how many per byte.

>[1] Upper case K is an ambiguous non-standard prefix,
>don't use it, assumed 1000 here.

     They are and have long been used interchangeably.  I see no problem with
either usage.

>http://en.wikipedia.org/wiki/Binary_prefix
>Tor is a network application, not a storage application.
>Hopefully all its references to bandwidth are using the SI and bit format.

     I've looked at those in the past, IIRC, and remained unfazed, thanks to
my own work experience.
     In summary, the contrived distinction between prefixes like "K" and "Ki"
is bad revisionism.  I suspect it is the outcome of advertising propaganda
manipulation by disk drive manufacturers because they have historically been
those who have been the most incensed by people correcting their bogus claims
of outsized disk drive capacities.  I never encountered this distinction until
hard disk drives began being sold for personal computers.  Many programmers
were outraged at the time and continued to be for nearly a decade afterward.
     "Bandwidth" is a thoroughly misused term in this context because it has
nothing whatsoever to do with the width of a band (of, for example, a frequency
or wavelength or wavenumber spectrum).  Data rate capacity is a more accurate
and appropriate expression.
     As noted above, transmission rates expressed in b/s have long been in
decimal terms, but expressed in B/s should be in powers of 2 most nearly
approaching a relevant power of 10.

>Thanks and carry on :)
>
     Likewise, I'm sure. :-}


                                  Scott Bennett, Comm. ASMELG, CFIAG
**********************************************************************
* Internet:       bennett at cs.niu.edu                              *
*--------------------------------------------------------------------*
* "A well regulated and disciplined militia, is at all times a good  *
* objection to the introduction of that bane of all free governments *
* -- a standing army."                                               *
*    -- Gov. John Hancock, New York Journal, 28 January 1790         *
**********************************************************************