RE: Ethernet frame format [7:5996]
Just getting started, there are probably some "easier" reads out there but that book will definitely give you the goods on TCP/IP... Regarding your question/statement, you are accurate that the "raw" Ethernet frame format has DA, SA, EtherType, Data, and FCS - to be a valid frame it just has to be between 64 and 1518 bytes (if we're including the 4 bytes of the FCS in our calculations) - notice that the top end number is not 1500 - the common "max" MTU size for Ethernet-attached devices talking IP. MTU is a function of L3, not L2. The IP header will indicate how many bytes of the Ethernet payload is consumed by IP "stuff" - add this to the 14 bytes consumed by the destination address (6), source address (6) and ethertype (2) to get the total frame size (+ plus the trailing 4 bytes for the FCS). This statement will be true unless adding all that up equals a number less than 60 (64 w/FCS) in which case the packet will be "padded" with 0's to make it a legal Ethernet packet. I think it is generally considered a good thing that packets aren't padded to the full Ethernet size (or MTU) - it that were the spec, I'm thinking that ATM would be a lot more popular as a LAN medium. Hope this helps. Ben -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of [EMAIL PROTECTED] Sent: Friday, May 25, 2001 6:59 PM To: [EMAIL PROTECTED] Subject: Ethernet frame format [7:5996] Dear Members List, I've just started the track for CCNA and, following all the repeated advices posted in this list, I started studing for Internetworking with TCP/IP, by Douglas Comer. The ethernet frame format stablishes as necessary information for the frame as DA, SA, Type, Data Area(variable from 46 to 1500 bytes) and a trailer FCS 4 bytes. I don't see how can we have different frame sizes correctly received, since there is no information about the specific lenght for every single frame, taking in account the asynchrounous nature of this communication. I thought that the layer 3 would pad till the MTU was reached, but I saw a trace on an ethernet network and I could see different frame sizes. Thanks in advance, Douglas Baltazar de Queiroz - Field Enginner --- UOL: o melhor da Internet. FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=6007&t=5996 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
Re: Ethernet frame format [7:5996]
Ethernet Version II does not have a length field, as you say, although IEEE 802.3 does. But Ethernet Version II still supports variable frame sizes. It just makes the NIC work a little harder. The NIC listens until clocked bits have stopped coming in and then it knows it's at the end of the frame. It could count the bits and still pass this info to the next layer up, but I don't think it does. The length field in IEEE 802.3 is an enhancement. The NIC can use the information for itself and it can easily pass the info to the next layer up. Priscilla At 07:59 PM 5/25/01, [EMAIL PROTECTED] wrote: >Dear Members List, > >I've just started the track for CCNA and, following all the repeated advices >posted in this list, I started studing for Internetworking with TCP/IP, by >Douglas Comer. > >The ethernet frame format stablishes as necessary information for the frame >as DA, SA, Type, Data Area(variable from 46 to 1500 bytes) and a trailer FCS >4 bytes. > >I don't see how can we have different frame sizes correctly received, since >there is no information about the specific lenght for every single frame, >taking in account the asynchrounous nature of this communication. > >I thought that the layer 3 would pad till the MTU was reached, but I saw a >trace on an ethernet network and I could see different frame sizes. > >Thanks in advance, > >Douglas Baltazar de Queiroz - Field Enginner > >--- >UOL: o melhor da Internet. >FAQ, list archives, and subscription info: >http://www.groupstudy.com/list/cisco.html >Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] Priscilla Oppenheimer http://www.priscilla.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=6008&t=5996 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
RE: Ethernet frame format [7:5996]
Great answer, but be careful with that "raw" term which to Novell people means the Ethernet raw format, aka ETHERNET_802.3, ask novell-ether, which is dest, src, length, immediately followed by the IPX header. The frame is only recognizable because an IPX header starts with an XNS checksum, which is . Novell didn't actually use the checksum when they borrowed XNS to create IPX. A "real" 802.3 frame has dest, src, length and an 802.2 header with DSAP and SSAP and maybe SNAP following. An Ethernet II frame has dest, src, type. Priscilla At 11:26 PM 5/25/01, R. Benjamin Kessler wrote: >Just getting started, there are probably some "easier" reads out there but >that book will definitely give you the goods on TCP/IP... > >Regarding your question/statement, you are accurate that the "raw" Ethernet >frame format has DA, SA, EtherType, Data, and FCS - to be a valid frame it >just has to be between 64 and 1518 bytes (if we're including the 4 bytes of >the FCS in our calculations) - notice that the top end number is not 1500 - >the common "max" MTU size for Ethernet-attached devices talking IP. MTU is >a function of L3, not L2. > >The IP header will indicate how many bytes of the Ethernet payload is >consumed by IP "stuff" - add this to the 14 bytes consumed by the >destination address (6), source address (6) and ethertype (2) to get the >total frame size (+ plus the trailing 4 bytes for the FCS). This statement >will be true unless adding all that up equals a number less than 60 (64 >w/FCS) in which case the packet will be "padded" with 0's to make it a legal >Ethernet packet. > >I think it is generally considered a good thing that packets aren't padded >to the full Ethernet size (or MTU) - it that were the spec, I'm thinking >that ATM would be a lot more popular as a LAN medium. > >Hope this helps. > >Ben > > > >-Original Message- >From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of >[EMAIL PROTECTED] >Sent: Friday, May 25, 2001 6:59 PM >To: [EMAIL PROTECTED] >Subject: Ethernet frame format [7:5996] > > >Dear Members List, > >I've just started the track for CCNA and, following all the repeated advices >posted in this list, I started studing for Internetworking with TCP/IP, by >Douglas Comer. > >The ethernet frame format stablishes as necessary information for the frame >as DA, SA, Type, Data Area(variable from 46 to 1500 bytes) and a trailer FCS >4 bytes. > >I don't see how can we have different frame sizes correctly received, since >there is no information about the specific lenght for every single frame, >taking in account the asynchrounous nature of this communication. > >I thought that the layer 3 would pad till the MTU was reached, but I saw a >trace on an ethernet network and I could see different frame sizes. > >Thanks in advance, > >Douglas Baltazar de Queiroz - Field Enginner > >--- >UOL: o melhor da Internet. >FAQ, list archives, and subscription info: >http://www.groupstudy.com/list/cisco.html >Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] >FAQ, list archives, and subscription info: >http://www.groupstudy.com/list/cisco.html >Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] Priscilla Oppenheimer http://www.priscilla.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=6009&t=5996 -- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
