[178181] in North American Network Operators' Group
Re: OT - Small DNS "appliances" for remote offices.
daemon@ATHENA.MIT.EDU (Rob Seastrom)
Thu Feb 19 08:13:12 2015
X-Original-To: nanog@nanog.org
To: Denys Fedoryshchenko <denys@visp.net.lb>
From: Rob Seastrom <rs@seastrom.com>
Date: Thu, 19 Feb 2015 08:13:09 -0500
In-Reply-To: <ca6435401865cad5c24de6233cb2a222@visp.net.lb> (Denys
Fedoryshchenko's message of "Thu, 19 Feb 2015 14:38:01 +0200")
Cc: geoff@proto6.com, nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
Denys Fedoryshchenko <denys@visp.net.lb> writes:
> Beaglebone has gigabit mac, but due some errata it is not used in
> gigabit mode, it is 100M (which is maybe enough for small office). But
> it is "hardware" mac.
The Beaglebone Black rev C BOM calls out the ethernet phy chip as
LAN8710A-EZC-TR which is 10/100 so there's your constraint. The MAC
is built into the SoC and according to the datasheet the AM3358B is
10/100/1000.
> Another hardware MAC on inexpensive board it is Odroid-C1.
Difficulty: hardware MAC tells you nothing about how it's connected,
either on the board or internally in the SoC. Ethernet on Multibus
and Ethernet on PCIe (neither likely on an embedded ARM ;-) are both
"hardware MAC" yet the bus-constrained bandwidths will differ by
several orders of magnitude.
-r
