When do we get this? [Option GlobeSurfer ICON USB Modem]

Apparently he's using a Virgin sim but roaming on Vodacom HSDPA, look at the post above yours.
 
Prometheus said:
Neither actually. I'm using Vodacom hsdpa at Virgin rates, paid R3.22 for 7.5MB. Just reconnected at 3.6Mbps. Don't know if it's a lapse in security or Cell C testing the Vodacom UMTS network but it rocks.

Getting a bit slow now, will do test after midnight.
Click to expand...

The 3.6Mb/s is merey a text string and does not indicate actual connection speed.

Run the Telkom speed test (http://www.telkom.co.za/minisites/dsl/speed1.html) and let's see what you're getting.
 
Well, Vodacom only supports up to 1.8Mbps under exceptional conditions. Don't know if 3.6Mb uses a different encoding scheme or just more slots. Modem is set on 921600bps in Windows so don't know if that will have an impact. Will do a test when the network is quiet as it seems to be overloaded.

The Telkom speed test is located at http://www.telkom.co.za/athome/products/dsl/speed.html but it's also not working atm.
 
Prometheus said:
Well, Vodacom only supports up to 1.8Mbps under exceptional conditions. Don't know if 3.6Mb uses a different encoding scheme or just more slots. Modem is set on 921600bps in Windows so don't know if that will have an impact. Will do a test when the network is quiet as it seems to be overloaded.

The Telkom speed test is located at http://www.telkom.co.za/athome/products/dsl/speed.html but it's also not working atm.
Click to expand...

Apologies, pasted an old link.

No 3.6Mb/s yet, sometime next year.
 
iBurst speedtest
download: 286kbps
upload: 27kbps
QoS: 38%
Round Trip: 155ms
Max Pause: 1558ms

Sentech speedtest
Download: 312.68kbps
Upload: ???

GamCo speedtest
Download: 214.90kbps
 
I/O Thingy

I got the globesurfer icon.

I'm very very impressed with the quick and easy installation, and speed.
Anyways, I was connected with Vodacoms 3G in 2005 without hsdpa, cool but not mind blowing. Now I sit here in namibia and enjoy much faster internet.
Just one thing, I realized that when i download a file from my browser I only get a maximum download speed of about 30-40KB/s. When I download 2 or 3 files at the same time I aslo get 30-40KB/s per file. When I use download accelerator with 8 segments downloading simulatneous I average around 140KB/s. (downloading MSN /16 MB, took me just 1:20s)
Peaking at about 160-170KB/s (1.3Mbs).

All these internet speed tests also don't nearly show the performance that I get. Iburts gives me 290Kbs, speedtest.net gives me 480Kbs. But when I download files with a download manager I normally don't see speeds under 1Mb/s.
It seems like something is limiting my speed per connection.

Any ideas?
 
I also get higher speeds when downloading. Downloads can only reach a certain speed on ANY connection. Your browser (or os rather) only requests the next bit of information once it receives the requested bit. To illustrate, let's say you have a 10k file and data is requested in 1k chunks. If your ping time is 100ms you can't download it faster than 1 second or 10k/s even on a 1GB/s connection. You can increase the packet size but that could cause more scattering of the frames making up the packet and so even longer ping times or a higher error rate causing more packets to be retransmitted. Sometimes decreasing the packet size can cause a shorter ping time but with a higher overhead. So it's a trade-off really.

With a segmented download multiple chunks are requested at the same time and they come in more or less one after the other while other chunks are requested again. I don't think there is a speed test that takes this into account.
 
Thanx for the feedback.

It makes allot of sense. Now I just need to find a Ping test site??
Tried to "ping" various sites from command prompt but all the requests were timed out.
Another funny thing that I don't understand. I don't think that my ping times are bad to the network, but I don't know of any tests that I can do to get a reference of what to expect when playing online games.
 
Well, since logging off last night Vodacom doesn't want to let me back onto their network. Message: "Logging on failed, Your connection is slowing down the network."

Only kidding about the last part :D Would have liked to do that segmented download to see what the network was capable of. Was good while it lasted though.
 
Prometheus said:
If your ping time is 100ms you can't download it faster than 1 second or 10k/s even on a 1GB/s connection.
Click to expand...

You've got it completely wrong :)

Latency (measure by ping times, for example) is completely independent of throughput.

Take a satellite link; high latencies, massive throughput. Or a 64K diginet line, good latency, crappy throughput.

Get a good book on TCP/IP (as suggested before ;) ) and read up on how IP and TCP works. Then read up on the effect of line clock plus processing latency on a specific packet size.
 
vodacom3g said:
You've got it completely wrong :)

Latency (measure by ping times, for example) is completely independent of throughput.

Take a satellite link; high latencies, massive throughput. Or a 64K diginet line, good latency, crappy throughput.

Get a good book on TCP/IP (as suggested before ;) ) and read up on how IP and TCP works. Then read up on the effect of line clock plus processing latency on a specific packet size.
Click to expand...
Line clock and processing latency... I thought we covered this already. :confused:
ic said:
Simple and common scenario that demonstrates the correlation between TCP/IP and packet loss and data transfer speed:

When transferring data [either up or down], for example using ftp|http [which is layered on top of TCP/IP], if a single frame within a specific packet is lost or delayed to the point where a timeout is triggered, what will happen is that TCP/IP will request the retransmission of the entire packet, which means more frames being transmitted and hopefully received and reassembled into their original packet, so far so good, but frames can and do arrive in any order [non-sequential] and all frames are required in order to assemble a packet from constituent frames, this means that TCP/IP has to wait until it has all the frames. If just one of the frames is missing then some time will have passed including the actual timeout before realising that a frame is not going to arrive. All of this adds latency to the actual data transfer itself, all of this will result in the overall data transfer taking longer than it should have under ideal conditions, and an increase in time results in a decrease in speed - regardless of the actual link speed...
Click to expand...
Can also read as an increase in latency results in a slower speed regardless of the actual link speed. As usual I have wasted my time (and money btw.) on terms which have very little - if anything - to do with the matter at hand. If it was such an important part of the equation there should be more info on it available or at least a wiki entry. ;) I guess next time I will simply ignore these subversive tactics and refer to this post to save time.


For everyone that doesn't know (and to tick off v3g some more):
The size of a frame is usually in the tens or hundreds of bytes depending on the protocol used where packets can be anything from a few hundred bytes to multiple kilobytes. If we assume that one frame gets lost for every 100 sent and we have packets consisting of one frame then statistically every 100th packet will be retransmitted and have a higher latency. If a packet consists of 10 frames then statistically every 10th packet will be retransmitted. If we now increase the packet size even further then statistically every packet will be retransmitted. Realistically some will make it through by some miracle. This applies not only to frames that don't arrive but also to frames that arrive late. The larger the packet size the more frames it has that are delayed and the longer the packet gets delayed.
 
Prometheus said:
If a packet consists of 10 frames then statistically every 10th packet will be retransmitted. ... The larger the packet size the more frames it has that are delayed and the longer the packet gets delayed.
Click to expand...
Your explanation is a bit confusing & misleading.

In the IP world 1 frame = 1 packet, athough they are not refering to the exact same amount of data. The term frame is usually used in context to layer 2 protocols, while packet refers to layer 3 protocols.

For example, an Ethernet frame of 1514 bytes includes a 14 byte header & 1500 bytes payload. The payload consists of a 1500 byte IP packet which consists of a 20 byte IP header & 1480 bytes of payload (e.g. TCP or UDP again with header & payload).

You dont get multiple packets within frames or visa versa.
 
Roman4604 said:
Your explanation is a bit confusing & misleading.

In the IP world 1 frame = 1 packet, athough they are not refering to the exact same amount of data. The term frame is usually used in context to layer 2 protocols, while packet refers to layer 3 protocols.

For example, an Ethernet frame of 1514 bytes includes a 14 byte header & 1500 bytes payload. The payload consists of a 1500 byte IP packet which consists of a 20 byte IP header & 1480 bytes of payload (e.g. TCP or UDP again with header & payload).

You dont get multiple packets within frames or visa versa.
Click to expand...
Go read what ic posted again which essentially says that packets consist of multiple frames. I would like to see how you manage to cram a 10k packet into only 1500. Would be some marvelous compression you have there. ;)
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X