Actual throughput vs reported

[Aaron Toponce]

First off, before answering the questions, I plugged one key into my T61,
and am getting the following results:

    $  pv -a < /dev/random > /dev/null 
    [ 3.6kB/s]

The previous results are on my Raspberry Pi, connected via a USB Octopus.

On Mon, Oct 01, 2012 at 11:24:28PM -0400, Joshua Newton wrote:
> * Are the specs clear on whether that's 32Kbps before or after the
> entropy is tested and some possibly discarded?

Per the website results using pv(1), it's after. 32 Kbps = 4 KBps.

> * Have you adjusted kernel.random.poolsize,

It's not writable.

    $ ls -l /proc/sys/kernel/random/poolsize
    -r--r--r-- 1 root root 0 Aug 13 14:40 /proc/sys/kernel/random/poolsize
    $ cat /proc/sys/kernel/random/poolsize
    4096

> kernel.random.read_wakeup_threshold,

Not sure what to change this to, however, here are the contents:

    $ cat /proc/sys/kernel/random/read_wakeup_threshold
    64

> kernel.random.write_wakeup_threshold? (I'm assuming Linux.)

    You assuming correctly:

    $ cat /proc/sys/kernel/random/write_wakeup_threshold
    1024

> * Are you certain the kernel random pool management code isn't the
> bottleneck?

At this point, I'm sure it's hardware. Now just to pinpoint where:

    1) CPU
    2) FSB
    3) USB controller
    4) USB hub
    5) Something else on the Raspberry Pi

-- 
. o .   o . o   . . o   o . .   . o .
. . o   . o o   o . o   . o o   . . o
o o o   . o .   . o o   o o .   o o o
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 519 bytes
Desc: not available
URL: <http://lists.simtec.co.uk/pipermail/entropykey-users/attachments/20121001/4cc470cc/attachment.pgp>