"So if 1G 1000BASE-T, pairs are bidirectional, full duplex, does that mean
each singe wire in the pair group is doing either transmit or recieve. In
that case I can understand there will be no collision, as each single wire
will be either transmitting or receiving."
That was my understanding yes. 250M of BW in both directions per pair.
From: Irfan Sid [mailto:lifeoverip_at_gmail.com]
Sent: Monday, 1 August 2011 3:02 PM
To: Joe Astorino
Cc: Carlos G Mendioroz; Aaron Riemer; Cisco certification; Cisco
certification
Subject: Re: Packet delivery on a Ethernet Wire
So if 1G 1000BASE-T, pairs are bidirectional, full duplex, does that mean
each singe wire in the pair group is doing either transmit or recieve. In
that case I can understand there will be no collision, as each single wire
will be either transmitting or receiving.
But if a single wire in the pair group is doing both transmitting and
receiving, you have frames going in both directions on a single wire. Its
like cars coming in both directions in a single lane. In this case you will
expect collisions.
Please can someone clarify .....
On 1G 1000BASE-T, pairs are bidirectional, full duplex.
(Like your POTS phone line, you can talk while you listen)
No collisions.
On Sun, Jul 31, 2011 at 9:47 PM, Joe Astorino <joeastorino1982_at_gmail.com>
wrote:
If all 4 pairs are used in 1000Base-T I wonder how POE also rolls into
that...
On Sun, Jul 31, 2011 at 1:55 PM, Carlos G Mendioroz <tron_at_huapi.ba.ar>
wrote:
On 1G 1000BASE-T, pairs are bidirectional, full duplex.
(Like your POTS phone line, you can talk while you listen)
No collisions.
Signal treatment is beyong my current understanding, but it uses 5 levels in
each pair!
Amazing...
Irfan Sid @ 31/07/2011 11:32 -0300 dixit:
On 1 Gig link pin 4,5, 7 and 8 are bi-directional meaning the same wire
will be used for Tx and Rx?
And for 10/100 those pins are not used at all, so 10/100 is serial rather
then parrallel.
Does that mean that odd collison is possible on both since they are
bidirectional packet streams??
http://pinouts.ws/rj-45-pinout.html
http://pinouts.ru/Net/Ethernet10BaseT_pinout.shtml
On Sun, Jul 31, 2011 at 2:50 PM, Carlos G Mendioroz <tron_at_huapi.ba.ar
<mailto:tron_at_huapi.ba.ar>> wrote:
I stand corrected, in 1000BASE-T there are 4 bidirectional signal paths,
so it's basically a parallel interface.
Sorry about that.
-Carlos
Aaron Riemer @ 31/07/2011 05:36 -0300 dixit:
Yep. However just doing some reading it looks like Gigabit copper
(1000BaseT) uses all four pairs and this link seems to suggest
that data is
sent in parallel.
http://www.cisco.com/en/US/__tech/tk389/tk214/tech___digest09186a0080091a86.
html
<http://www.cisco.com/en/US/tech/tk389/tk214/tech_digest09186a0080091a86.htm
l>
"1000BASE-T works by using all four of the Category 5 pairs to
achieve 1000
Mbps operation over the installed Category 5 copper cabling.
1000 Mbps data
rates are achieved by sending and receiving a 250 Mbps data
stream over each
of the four pairs simultaneously (4 X 250 Mbps = 1 Gbps)."
Cheers,
-Aaron.
-----Original Message-----
From: nobody_at_groupstudy.com <mailto:nobody_at_groupstudy.com>
[mailto:nobody_at_groupstudy.com <mailto:nobody_at_groupstudy.com>] On
Behalf Of
Irfan Sid
Sent: Sunday, 31 July 2011 1:02 PM
To: Aaron Riemer
Cc: Carlos G Mendioroz; Cisco certification; Cisco certification
Subject: Re: Packet delivery on a Ethernet Wire
So does that mean that in copper wire, you have two paths ie
full duplex,
one for transmit and one for recieve. One pair of the copper will be
dedicated to tansmit and one to recieve. Each path transmitting
serially.
On Fibre its the same, you have two fibre's one for transmit and
one for
receive and data is transmitted serially on them.
Is this correct?
On Thu, Jul 28, 2011 at 6:07 AM, Aaron Riemer
<ariemer_at_amnet.net.au <mailto:ariemer_at_amnet.net.au>> wrote:
My apologies. I must have misread.
http://en.wikipedia.org/wiki/__Data_transmission
<http://en.wikipedia.org/wiki/Data_transmission>
Thanks for clearing that up.
Cheers,
-Aaron.
-----Original Message-----
From: nobody_at_groupstudy.com <mailto:nobody_at_groupstudy.com>
[mailto:nobody_at_groupstudy.com
<mailto:nobody_at_groupstudy.com>] On Behalf Of
Carlos G Mendioroz
Sent: Thursday, 28 July 2011 12:27 AM
To: Aaron Riemer
Cc: 'Irfan Sid'; 'Cisco certification'; 'Cisco certification'
Subject: Re: Packet delivery on a Ethernet Wire
Aaron,
copper has "pairs" because you use one wire as a reference
to the other.
But there is only one signal path (each way), and it works
serially.
-Carlos
Aaron Riemer @ 27/7/2011 4:09 -0300 dixit:
Hi Irfan,
I believe with copper Ethernet (100BaseTX for example)
since there are
copper "pairs" the data is sent in parallel. I would
assume the only way
parallel data transmission would work with fibre is if a
different band
of
light is used (CWDM for example).
Queuing on the router/switch only affects packets within
the router
itself
and has no bearing on packets along the wire (unless
they are dropped).
Tools such as compression / LFI / DSCP marking are
examples of the
router
manipulating the packets in some way before transmission
out an
interface.
Cheers,
-Aaron.
-----Original Message-----
From: nobody_at_groupstudy.com
<mailto:nobody_at_groupstudy.com>
[mailto:nobody_at_groupstudy.com
<mailto:nobody_at_groupstudy.com>] On Behalf Of
Irfan Sid
Sent: Wednesday, 27 July 2011 1:48 PM
To: Cisco certification; Cisco certification
Subject: Packet delivery on a Ethernet Wire
This maybe a stupid question would appreciate if someone
can clearify:
Are Ethernet packets delivered serially on the physcial
medium ie copper
or
fibre - one after another in one single file or line if
you will. Or are
they delivered in parallel ie. two or more packets can be in
transmission
along each other in parallel.
Also when you enable queueing on Router/switch. Does
this only effect
the
behaviour of how the router treats the packet on the
inbound or outbound
interface or does it also effect the behaviour of the
packet when it is
actually in transit on the wire itself.
Blogs and organic groups at http://www.ccie.net
<http://www.ccie.net/>
<http://www.ccie.net/>
Received on Mon Aug 01 2011 - 15:15:36 ART
This archive was generated by hypermail 2.2.0 : Thu Sep 01 2011 - 06:05:56 ART