LC Tests - the big day

loosecannon

Senior Member
Joined
Jul 27, 2004
Messages
731
Reaction score
0
Location
.
preliminary multithread from www.kernel.org ...

linux-2.6.8.1.tar.bz2
patch-2.6.9.bz2

that really warmed the modem up a whoping

Current In:
4662.0 B/s (28.5%)

my peak was wait for it ...
Max In:
6306.0 B/s (38.5%)

The figures that intrest me the most ... these i want to see in the 75%+

Average max 5 min values for `Daily' Graph (5 Minute interval): In: 2612.0 B/s (15.9%)/ Out: 1092.0 B/s (6.7%)

i plan on using this tread as a place for recording simplfied information regarding the fantastic "big fix" and a place to show my grand children what a great company sentech was
 
I have noticed that http://www.download.windowsupdate.com (67.29.170.61) is an exclusion on the international cap. This makes sense - I doubt that there is a local mirror.

see for yourself:
wget http://www.download.windowsupdate.com/msdownload/update/v3-19990518/cabpool/msjavwu_8073687b82d41db93f4c2a04af2b34d.exe

<post your results of this particular file please>

downloading that file + an international torrent I got FULL speed as in maxed out 15.5 and above. Which is nice - I was capped by my package. A sign of things to come? Doubtful. Just a loophole to utilise some of the idle international bandwidth.

LC, those speeds seem to be hitting the cap. I have also noticed that one international thread can consistently sit at 5KB/sec (current cap). See "proof".

<font size="1"><center>**Official Sentech MyWireless <b>56kbps</b> customer using patented carrier pigeon clipped wing technology** <font color="green"> proof </font id="green"></center></font id="size1">
 
yeah part of what ive been led to belive is this cap is disapearing so lets see what happens ...

im getting pathetic results on this file maybe pretoria is "fixed"
 
like this

Resolving www.download.windowsupdate.com... done.
Connecting to www.download.windowsupdate.com[64.4.20.61]:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 5,319,000 [application/octet-stream]

2% [&gt; ] 147,194 4.75K/s ETA 17:42
 
hey how you cut and paste from my console ...

well still 20 minutes before the fun realy starts and sentech starts er carries on having to lie to us ...
 
the file seems to be cleared from the cache now. I cant get the same speed again. Ok but its confusing since Im running a TP on my gateway as well..

with that switch does it force the remote TP to re-get the file? surely not..

OR <i>they</i> are reading here.. yeah thats perhaps pushing it a bit.

was probably cached. anyways cant wait for cap to fck off. so much free bandwidth waiting to be used.

Guys I bet it WONT be at midnight

<font size="1"><center>**Official Sentech MyWireless <b>56kbps</b> customer using patented carrier pigeon clipped wing technology** <font color="green"> proof </font id="green"></center></font id="size1">
 
if you set the -C [correct me if im wrong] it puts pragma-no-cache in the request header and the TP will ignore it [RFC compliance] and therefore allow you to get a clean version
 
<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"> so much free bandwidth waiting to be used.
<hr height="1" noshade id="quote"></blockquote id="quote"></font id="quote">

You're joking, right?
 
sentech to sentech transfer

23:51:35 (6.18 KB/s) - `mplayerc.exe' saved [1340416/1340416]
 
malec what i meant is that they are capping the ppl downloading so tightly that the *normal* leftover BW is just sitting around. We are all hitting our 5Ks quite nicely. There must be some left to spare.

<font size="1"><center>**Official Sentech MyWireless <b>56kbps</b> customer using patented carrier pigeon clipped wing technology** <font color="green"> proof </font id="green"></center></font id="size1">
 
I'm not a Sentech user (thankfully) but I can see this happening :

(Uninformed) User : "I'm only getting 4 KB/s on my downloads all the time, I'm going to cancel!"

Orc : "Sir, please try the speed test on our site."

User : "The test says 30KB/s."

Orc : "Well sir, we are providing you with the speeds stated in the package and have no control over the rest of the internet."

User : "Oh ok" *click*
 
--- www.sentech.co.za ping statistics ---
10 packets transmitted, 10 packets received, 0% packet loss
round-trip min/avg/max = 524.8/727.2/1105.6 ms

packet size 1428 ...

this is what i like ...
 
remember that windowsupdate is multihomed via akamai - that's why the IP's are all over the place ... not to mention that round-robin dns also comes into play

<hr noshade size="1"><font size="1"><div align="right">Why do I even bother .....
Will they make the 128/512K deadline ... hope so for their sake
<s> 16 17 18 19 20 <font color="green">21 22</font id="green"> 23 24 25<font color="red"><b> 26</b></font id="red"> 27<font color="green"><b> 28 29</b></font id="green"> 30</s> <font color="red"><b> 31</b></font id="red">
</div id="right"></font id="size1">
 
greedy, they only bandwidth left is used up by winston smith browsing www.spur.co.za at full speed china
 
i bitch like a banshee when i dont get 128kbs on there speed test ...as that is in the core network and should not be part of the "contention ratio"

just called the support desk they will be updating me on current status ... also noticed a glitch almost spot on 12 ...

lets see what happens ...
 
loosecannon ... just a quick one ... mind sending a few ping packets to www.bbc.co.uk ...

My latency to that site has sky rocketed but need some testing doine ;-)

The fix is great .. except it doesn't seem to be implemented yet ;-)

<hr noshade size="1"><font size="1"><div align="right">Why do I even bother .....
Will they make the 128/512K deadline ... hope so for their sake
<s> 16 17 18 19 20 <font color="green">21 22</font id="green"> 23 24 25<font color="red"><b> 26</b></font id="red"> 27<font color="green"><b> 28 29</b></font id="green"> 30</s> <font color="red"><b> 31</b></font id="red">
</div id="right"></font id="size1">
 
Top
Sign up to the MyBroadband newsletter
X