[sf-lug] traffic shaping, ...: Re: [on-list] site up, http[s] down: Re: Wierd problems trying to access linuxmafia.com

Rick Moen rick at linuxmafia.com
Thu Jan 3 23:49:29 PST 2019


Quoting Michael Paoli (Michael.Paoli at cal.berkeley.edu):

> As a Raw Bandwidth customer, with most or all of their services,
> and I believe also the "package" you have (and probably same
> or similar to what I have), there are probably some additional
> tests one can do to rather effectively test {up,down}load bandwidth,
> and could of course also look at latency while running such tests,
> and may significantly reduce (at least some) other variables.

Granted, but -- seriously -- aren't the data I arrived at good enough?
If not, why not?

I've used the speedtest.net data before on a corporate context with
customers needing to test their uplinks from a variety of locations
around the world, and I came to have confidence in it.  Obviously your
point is well taken that the ideal remote endpoint for uplink tests
is right at Raw Bandwidth Communications's own infrastructure, but 
I am not aware of having that as part of my customer package (maybe;
I'd have to ask Mike Durkin) and good enough is good enough.  And 
yes, I could do timed downloads of materials of known size from Raw
Bandwidth's site, but that doesn't help even a tiny bit with measuring
outbound capacity that is obviously the main issue.

> In all probability, bottleneck would be the [A]DSL link, and not
> hosts or drive I/O at either end.

So close to certain as to not be worth time considering anything else.

> Also be sure, especially with http (or https, if used), to use a
> "dumb enough" client, to not be having the server compress and client
> decompress the data - one can probably control that (if it doesn't already
> not do compression by default) with a client such as wget or curl (and
> probably most or all ftp clients). 

Again, why are the data I already gathered not sufficient?  I could get
wrapped up in revisiting those metrics for days, but for what benefit?  
Seems to me, I already have a pretty reliable set of figures.

> Also, if one has shell access/account with Raw Bandwidth....

I don't.


Here is just about a perfect implementation of QoS / traffic shaping for my 
use-case, except it's evident that my kernel lacks one or more feature
necessary to support it:
https://www.tldp.org/HOWTO/ADSL-Bandwidth-Management-HOWTO/implementation.html



More information about the sf-lug mailing list