CSMA/CD a quick quesion

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

Just so i know i do understand this and i have read some topics i found on here with some good describtions by the serious guy in dark glasses (sorry cant remember username yet, damn newbies lol)

The NIC listens on the network for any data already being transmitted, if there is some it waits, if it hears nothing it sends its data out. If at that time though another NIC does the same thing a collison occurs so both NIC's stop sending and wait a while. Now am i right in guessing they wait a random time? As if a collison occurs and the NIC wait a predetermed length of time surely the same collison will happen again , right?

Also how exactly is a collison detected? Do the sent data packets get a recieved message come back so if there is a collison its not recieved and so the NIC knows , i doubt this as it would take ages surely, or do they bounce back, actually ill stop guessing and wait for the explaination )

Many thanks!

 

Oh man! Ad-blocking software has been detected! :'(

This website is run by the community, for the community... and it needs advertisements in order to keep running. Blocking our ads means your killing our stats!
Please disable your ad-block, or become a premium member to hide all advertisements and this notice.

Correct.

Spot on. Which is why a random time is used.

The NIC listens at the same time it is transmitting. But this depends on the media in question.

Quite a nice article here.

Harry.

 

Briiliant thankyou!

See im still not fully thinking in what im going to call with out being offensive, IT Nerdy terms, collison in my mind as data hitting another packet, you know as in car collison and stopping, not how that article explains they get mixed up so the sent data isn't the same as the recieved data,

How does the reciever though now though whats been sent it?
I know its different but i understand in the differential ended SCSI way it detects noise by having the resevrse of the signal sent also but i dont fully get how something knows what it should of recieved with out having something like this sent also?

 

If I understood your question correctly you would like to know how the receiver would know if the data has been sent even though it has not been received yet?

Well when ever you send data the nic that receives the packets usually sends a packet back stating that it received the data. If the the reply did not come through then the nic automatically assumes that the data did not get to the destination and resends the data, or for example when you try to ping another computer what happens is that you send 4 ICMP packets to the destination. If received you get a message back saying 4 ICMP packets been successfully received with times it took to get there in ms.

Although guys please correct me if I forgot to add something or made a mistake.

 

Hey mate thanks for the answer but i was wonder how does the reciever know its recieved data thats been in a collision? Is it just simply that it's so messed up it cant make head nor tail of what it means so it sends out a jam message to stop all transmissions?

 

Think of a 'phone call. While you talk you listen as well. What you hear is called 'sidetone' - a portion of your signal is fed back by your handset to the earphone.

We are so used to this now that if this doesn't happen we think the 'phone is dead!

Now imagine that while you are talking the other party rudely interrupts. You now hear sound in your earphone that doesn't match what you are saying.

You have a collision.

And the SCSI differential thing is not quite as you are saying. The signal is effectively a push-pull thing - you push on one side while pulling on the other. Interference on the line will do the *same* thing on both lines, i.e. push-push or pull-pull, so at the other end the interference signal on one line will find itself cancelling the same signal on the other line.

Harry

 

This doesn't cover the case of a collision though. It is just the standard method of verifying a message has been received at the far end.

Harry.

 

so in a kind of a way they have a conversation and roughly know what the other is going to say back? and when that suddenly chances it knows its had a collision?

The only the correct signal is left, aha i gets ya! thanks )

 

Er - No.

You know what *you* are saying. So if what you are hearing doesn't match this it must be somebody else. It doesn't matter what the other guy says - he could even whistle a tune or something. What you hear doesn't match what you are saying, so you know you have a collision.

Harry.

 

yep sorry i got it now slighty read what you first said wrongly, so its typically the transmitting NIC that senses the collision and sense the jam signal to hault transmitions , right now?

 

Thanks Harry for the correction. I misunderstood Dhughes question about the recieving end. I thought he meant what happens if the data did not get to the destination even though it was sent.

 

this is the acknowledgement protocol (ACK and NAK) which your referencing here isn't it? (Just double checking myself mainly)

 

Yeah pretty much.

 

Yup - that's right!

Harry.

 

you guys are the Bee knee's , really appreachiate your time )