Internet Speeds

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Internet Speeds

Daniel-2
Having tested my speeds before, it just occurred to me ..... "How can
'they' separate the download and upload speeds/times?? Isn't it just
one, single, signal, effectively??"

For example, if I want to do a test, I click a button on a website,
which initiates a signal from my computer, through various Internet
servers/links to the SpeedCheck site, which then sends a signal back
through various Internet (maybe different) servers/links to my computer.

Does the distant terminal 'look' at the timestamps from my computer
clock as contained in my message header and compare it to 'their'
computer/server clock and convert that into an effective upload speed??
And then my computer does likewise to work out my download speed??

What happens if my computers clock is fast or slow?? Can I have an
infinite speed one way or the other?? And, conversely, a dead slow speed
the other way??

--
Daniel

User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.1 Build identifier: 20171016030418
or
User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.5 Build identifier: 20190609032134

User agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.1 Build identifier: 20171015235623
_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

Don Spam's Reckless Son
Daniel wrote:

> Having tested my speeds before, it just occurred to me ..... "How can
> 'they' separate the download and upload speeds/times?? Isn't it just
> one, single, signal, effectively??"
>
> For example, if I want to do a test, I click a button on a website,
> which initiates a signal from my computer, through various Internet
> servers/links to the SpeedCheck site, which then sends a signal back
> through various Internet (maybe different) servers/links to my computer.
>
> Does the distant terminal 'look' at the timestamps from my computer
> clock as contained in my message header and compare it to 'their'
> computer/server clock and convert that into an effective upload speed??
> And then my computer does likewise to work out my download speed??
>
> What happens if my computers clock is fast or slow?? Can I have an
> infinite speed one way or the other?? And, conversely, a dead slow speed
> the other way??
>

Download a decent-sized file, upload the same file.  The times will be
different.
It has been a few years since I had a good look at the mechanics of
TCP/IP and some work has been done on optimisation in the meantime.
The transmitter can send a block and then wait for confirmation that the
block has been received, or the transmitter can send a number of blocks
asynchronously and process the confirmations the same way.
Your computer's clock can be set to a different timezone, in fact it
probably will be.  I think differences are measured.

--
spammus ergo sum, viruses courtesy of https://www.nsa.gov/malware/
_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

PietB-2
In reply to this post by Daniel-2
Daniel wrote:
> Having tested my speeds before, it just occurred to me ..... "How can
> 'they' separate the download and upload speeds/times?? Isn't it just
> one, single, signal, effectively??"

No. Whatever the medium it's transported over, it's a stream of packets,
each with a source and a destination address. So if you're the source,
it's "upstream", if you're the destination, it's "downstream".

-p

_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

PietB-2
In reply to this post by Don Spam's Reckless Son
Don Spam's Reckless Son wrote:
> Download a decent-sized file, upload the same file.  The times will
> be different.

Not necessarily. If you have e.g. a fiber 100/100 internet connection,
download and upload speed may be [nearly] the same.

-p

_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

Jeff Barnett
In reply to this post by Daniel-2
Daniel wrote on 7/23/2019 1:59 AM:

> Having tested my speeds before, it just occurred to me ..... "How can
> 'they' separate the download and upload speeds/times?? Isn't it just
> one, single, signal, effectively??"
>
> For example, if I want to do a test, I click a button on a website,
> which initiates a signal from my computer, through various Internet
> servers/links to the SpeedCheck site, which then sends a signal back
> through various Internet (maybe different) servers/links to my computer.
>
> Does the distant terminal 'look' at the timestamps from my computer
> clock as contained in my message header and compare it to 'their'
> computer/server clock and convert that into an effective upload speed??
> And then my computer does likewise to work out my download speed??
>
> What happens if my computers clock is fast or slow?? Can I have an
> infinite speed one way or the other?? And, conversely, a dead slow speed
> the other way??

It has to with how fast you ISP is willing to schedule and push data.
Comcast, for example has different speed tiers at related prices so they
can control speeds and prioritize traffic in either or both directions -
faster tiers are faster in both directions. If you purchase a dedicated
IP and run servers, I believe that speeds are equal. I believe that
there is far more data heading to rather than from consumers so it makes
sense (maybe) to prioritize by data directions.
--
Jeff Barnett


_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

Daniel-2
Jeff Barnett wrote on 24/07/2019 6:48 AM:

> Daniel wrote on 7/23/2019 1:59 AM:
>> Having tested my speeds before, it just occurred to me ..... "How can
>> 'they' separate the download and upload speeds/times?? Isn't it just
>> one, single, signal, effectively??"
>>
>> For example, if I want to do a test, I click a button on a website,
>> which initiates a signal from my computer, through various Internet
>> servers/links to the SpeedCheck site, which then sends a signal back
>> through various Internet (maybe different) servers/links to my computer.
>>
>> Does the distant terminal 'look' at the timestamps from my computer
>> clock as contained in my message header and compare it to 'their'
>> computer/server clock and convert that into an effective upload
>> speed?? And then my computer does likewise to work out my download
>> speed??
>>
>> What happens if my computers clock is fast or slow?? Can I have an
>> infinite speed one way or the other?? And, conversely, a dead slow
>> speed the other way??
>
> It has to with how fast you ISP is willing to schedule and push data.
> Comcast, for example has different speed tiers at related prices so they
> can control speeds and prioritize traffic in either or both directions -
> faster tiers are faster in both directions. If you purchase a dedicated
> IP and run servers, I believe that speeds are equal. I believe that
> there is far more data heading to rather than from consumers so it makes
> sense (maybe) to prioritize by data directions.

Thanks, Jeff. I don't think my ISP, a local guy who I used to live up
the road from (o.k., up several roads from), throttles my account, maybe
the Telco he buys his service from does though.

--
Daniel

User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.1 Build identifier: 20171016030418
or
User agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.5 Build identifier: 20190609032134

User agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
SeaMonkey/2.49.1 Build identifier: 20171015235623
_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general
Reply | Threaded
Open this post in threaded view
|

Re: Internet Speeds

NFN Smith
In reply to this post by Jeff Barnett
Jeff Barnett wrote:
>
> It has to with how fast you ISP is willing to schedule and push data.
> Comcast, for example has different speed tiers at related prices so they
> can control speeds and prioritize traffic in either or both directions -
> faster tiers are faster in both directions. If you purchase a dedicated
> IP and run servers, I believe that speeds are equal. I believe that
> there is far more data heading to rather than from consumers so it makes
> sense (maybe) to prioritize by data directions.


This is typical for virtually any ISP, not just Comcast, where upload
speeds are only a small fraction of available download speeds,
especially for consumer-grade accounts.  By my experience with multiple
providers, I'm used to seeing upload speeds that are below 1 Mb/second,
regardless of download speeds, whether 5 Mb, 40 Mb, or even higher.
Most of the time, even if you're paying for a higher tier of service,
the difference is only applied to downloads, and you frequently don't
get higher upload speeds -- or if you do, not significantly higher.

I haven't done a lot of looking, but even in data centers where there's
lots of servers, upload speeds may be higher than for consumer accounts,
but they're still going to be slower than download speeds.

For consumer stuff, a lot is based on the expectation that users
aren't/shouldn't be running servers, and as a result, there isn't the
need for the additional bandwidth, other than comparatively rare bursts.

Smith
_______________________________________________
general mailing list
[hidden email]
https://lists.mozilla.org/listinfo/general