著者: Ralph Ronnquist 日付: To: dng 題目: Re: [DNG] network measurement
On Fri, 7 May 2021 20:02:13 -0400
Hendrik Boom <hendrik@???> wrote:
> I've been having trouble using zoom recently.
> If there's a large meeting, no questions I ask can heard clearly,
> even if the meeting uses the protocol of muting everyone else.
>
> I also have problems with short browser delays that may last a few
> seconds to a minute or so.
>
> However, the pppoe connection stays up when this happens.
>
> I've been complaining to the ISP, who does what it can and raises
> tickets with the phone company.
>
> There has been improvement since two weeks ago, when the ppp
> connectino would drop even more frequenty than the delays I'm
> encountering now.
>
> The problem usually occurs in the daytime, but occasionally in the
> evening.
>
> The phone company now opines that the data rate on the connection is
> barely sufficient for a multiparty zoom connection, and this is
> because of the distance to the exchange.
>
>
> ****
> Now I'd like to measure what's going on in my system.
> ****
>
>
> My home network accesses the rest of the world using a Linux box,
> whch also hosts a website, does SMTP, does firewalling/masquerading,
> and has a file server for the LAN. That Linux box is the only thing
> connected directly to the DSL modem, which operates in bridge mode.
>
> What tools does Linux have to measure this, so I can find out
> what's going on in my network and why. And maybe even correlate
> measurable activity with the perceived hiatuses.
>
> I'd be surprisd if there were none, Linux being born on the net.
> (I'd even be more surprised if they were easy to use.)
>
> Maybe I do ned to replace my DSL with a higher-capacity connection of
> some kind. But I'd like to see the numbers.
Often enough that kind of problem is due to DNS being slow rather than
actual networking; is it the same with a local, caching DNS ?