April 19th, 2006 06:04 AM
You mentioned that B= Bytes while b refers to bits, Now if you refer to the attachment in my thread the screenshot of that DU meter you could see its showing speed like XXX kB/sec so as per your reply it should XXX Kilo Bytes Per Second...... Now then this means this utility is not following the standards..... The reading it should be XXX kb/sec...
Now i Just checked the option in that software its saying clearly Kb/sec = Kilo Bits Per Second and KB/sec = Kile Bytes per Second.
But now the problem is I am getting normally 40 KB/sec and the router or DSL Modem i Have is an ADSL type..... So again 32 kB/sec has to be the maximum i can get on this link then how come I am getting more than that whenever huge data is being transfer in either ways....
You said that brust could be because of that router buffer..... than i have two queries here: -
1) Is the DU utility showing me the speed b/w this router and my system? If yes then how can i see the exact channel speed ? I mean the real DU speed of this Link i have.
2) If the DU utility is showing me the speed of the link than how come its possible that on link of 32 KBps i am getting 40 KB/sec normal and sometime reaching to 80 KB/sec for a min or more and sometimes 200+ KB/sec for even a second?
I am asking too much here..... But the problem is unless or until i get things clear in my brain i will keep on throwing such questions towards you folks.
One machine can do the work of fifty ordinary men. No machine can do the work of one extraordinary man!
April 19th, 2006 08:41 AM
Hi FanacooL and congratulations on your "addiction"
As I have mentioned, I am no technical expert in this area but I do have to explain the basic principles to non-technical audiences.
There are two concepts here, which are related, but not exactly the same.
1. Connection speed. This is generally measured in Kbps (bits), so in the example you mention that would be 256Kbps. Please be aware that this is an artificial limit imposed by the ISP. The actual spur that you are connected to is probably capable of almost 10Mbps in theory, but this will be reduced by distance, quality of the copper wire/connections and so on.
2. Data transfer rate. This is traditionally measured in KBps (bytes). The actual notation may vary but the convention is that if the "B" is uppercase, it means "bytes" and if it is lowercase it means "bits". So, as there are 8 bits in a byte your 256Kbps connection can theoretically transfer data at 32KBps. Naturally this is affected by inefficiencies in the cabling and hardware so an approximation of 30KBps is reasonable.
So, why do you get fluctuations in speed, in particular spikes of higher intensity?
1. Buffering. phishphreek80 is spot on there. Buffering data will produce spikes. You may recall the old dial-up modems (56.6Kbps) when you would start a download and watch the meter as the speed rapidly dropped to some sort of average value? This was the effect of buffering, and you did not even have to have a router for it to happen.
2. ISP/Local policy. If I have a certain amount of available resource I might have a fixed "cap" for each user, or I might decide to dynamically reallocate it. Obviously the dynamic reallocation makes better use of total available resource and helps to meet SLAs
3. My UK private connection approach. Here the ISP has "oversold" the resource............a bit like double booking hotel rooms and airline tickets. They promise me a connection speed of 2.2Mbps which means that they can only have 4 users on a spur, whereas I know that they have 50!
They are gambling on the fact that not everyone will be on at the same time doing large downloads. They probably only have about 175Kbps of resource per user if everyone was on at once. In fact, I strongly suspect that there are far less than 50 users on my link, given the speeds I experience, which are around 1.6Mbps. That compares with the average UK T1 of 1.51Mbps and ADSL of around 500Kbps.
If you cannot do someone any good: don't do them any harm....
As long as you did this to one of these, the least of my little ones............you did it unto Me.
What profiteth a man if he gains the entire World at the expense of his immortal soul?
April 19th, 2006 02:01 PM
I attempted to download your screenshot but since you had zipped it, instead of plainly posting it, I backed out of the download and instead followed your words for this discussion.
Now if you refer to the attachment in my thread the screenshot of that DU meter you could see
You think you should be getting no more than 30KB but you are getting 40KB, so what's the problem with this? Like the song says "Don't Worry...Be Happy"
But now the problem is I am getting normally 40 KB/sec and the router or DSL Modem i Have is an ADSL type
Have you called your DSL provider to ask them? They may have raised your bandwidth, they may have raised everyone's bandwidth, perhaps they bought better equipment, or they finally found that loopback problem and fixed it or... or... or...
You are not going to find out in this thread what is going on, but you might if you'd of called your ISP. You are getting a good education on DSL here though.
Anyways, since you haven't called (and probably won't) let me tell you this:
Some (I said SOME) but not all data speed measurement software measures not the true bit rate but rather the data rate. Meaning, data compression is applied to the source data which is then downloaded to your computer and then uncompressed and calculated. So... with this method (and it happens alot) more ACTUAL data is now in your computer beyond what the bit rate would suggest.
Some ISP's compress the incoming data at the ISP's source, shuttle it to your router which then uncompresses it and provides it to you using a 10/100 ethernet connection to your computer which is running DU meter. You notice more data than is possible on a slow link.
This is one way of getting more data downloaded on a slower link.
Did I explain that well enough?
Beta tester of "0"s and "1"s"
April 19th, 2006 02:28 PM
This may or may not be directly related to your question, but recently there has been some news lately on ISP's traffic shaping...
That's because there's different types of ISPs, some ISPs are what we call infrastructure based, and that means they build their own networks. But other ISPs may be just a brand, and are paying money for the fibre connectivity in that network, and maybe they can't even afford to buy a whole fibre and they're just leasing per megabit units of bandwidth. So you're going to be more inclined to try and shape your traffic to keep your cost base down.
Unfortunately, if you buy more bandwidth, somebody has to cover the cost of that, so what we're seeing is ISPs introducing tiered services. So for example, given the opportunity for light users to maybe have a metered package, who pay per unit at a time or per amount of data you download, so you can control your own costs as a user, and unlimited packages will inevitably cost a bit more.
Traffic shaping is an attempt to control computer network traffic in order to optimize or guarantee performance, low latency, and/or bandwidth. Traffic shaping deals with concepts of classification, queue disciplines, enforcing policies, congestion management, quality of service (QoS), and fairness.
Traffic shaping provides a mechanism to control the volume of traffic being sent into a network (bandwidth throttling), and the rate at which the traffic is being sent (rate limiting). For this reason, traffic shaping schemes need to be implemented at the network edges to control the traffic entering the network. It also may be necessary to identify traffic flows at the ingress point (the point at which traffic enters the network) with a granularity that allows the traffic-shaping control mechanism to separate traffic into individual flows and shape them differently 1 .
Two pre-dominate methods for shaping traffic exist: a leaky bucket implementation and a token bucket implementation. Both these schemes have distinctly different properties and are used for distinctly different purposes.
In computer networking, traffic shaping works by debursting traffic flows, i.e. smoothing the peaks and troughs of data transmission.
A before-and-after example of how traffic shaping works is as follows.
* Before traffic shaping: 10 packets in one second, 0 packets in the next second, 10 packets in the next second, 0 packets the next second.
* After traffic shaping: 1 packet per 0.2 seconds.
When lots of traffic flows past a packet bottleneck (logical or physical) the benefits of traffic shaping are:
* Less jitter.
* Reduced packet loss.
* Lower latency.
So a lot of this will show a variance on whatever speed tests you perform......
PC Registered user # 2,336,789,457...
"When the water reaches the upper level, follow the rats."