Windows 7 Ultimate SP1 x64
Cable modem: Motorola MB7220 V1.0 (2016)
Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)

Yes, the modem and router are separate devices. Is there a third-party
test application for me to judge the ongoing reliability of these devices?

No, I do not mean a speed test. Speed test Web sites give good results.
However, many times it seems that actual uploads and downloads are much
slower. I want to know if I have to replace either of these soon.

--
David E. Ross
<http://www.rossde.com/>

On Sat, 26 Aug 2023 18:45:16 -0700, "David E. Ross" <nobody@nowhere.invalid>
wrote:

>Windows 7 Ultimate SP1 x64
>Cable modem: Motorola MB7220 V1.0 (2016)
>Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>
>Yes, the modem and router are separate devices. Is there a third-party
>test application for me to judge the ongoing reliability of these devices?
>
>No, I do not mean a speed test. Speed test Web sites give good results.
> However, many times it seems that actual uploads and downloads are much
>slower. I want to know if I have to replace either of these soon.

Some random thoughts...

Speed test sites are generally configured to deliver as much speed, through
every link in the chain from them to you, as your system can handle. Other
download or upload sites may not have the same design priorities, so the problem
could be them and not you.

How are you measuring speed/throughput/bandwidth, whatever you want to call it?
I use a small app called Bandwidth Monitor on each of my PCs, but it's not free.
Someone posted a free open-source app that works in a similar way, but I don't
remember the name.
http://www.bwmonitor.com/ - not free, but has 30-day trial
Anyway, what's a normal speed, and what speed do you see instead when it seems
slow?

Is the slowness limited to a single PC? Do you have other PCs on the LAN and do
they all seem to be slower than you'd like?

Is your LAN all Gigabit, including the Ethernet switch (switches, if plural)?
Also, assuming your LAN is entirely Gigabit, have you checked whether you're
actually connected as Gigabit and not 100Mbps? A bad cable or a loose connector
can make that difference.

Are you using WiFi or hard-wired Ethernet? WiFi is subject to interference, so I
would definitely use Ethernet for testing purposes.

When you move data from one PC to another on your LAN, do you see any
performance issues? Things like network adapters, RAM, and hard drives can all
affect data transfer speed. LAN data transfers would not involve your cable
modem or the router portion of your Netgear router. Only the embedded Ethernet
switch would be involved, unless you're using WiFi, in which case the embedded
bridge and embedded Access Point would also be in the mix. Try to keep it
simple.

There's more, but I don't know what you've done already.

My LAN is all Gigabit, so intraLAN data transfers tend to run at about 990Mbps,
unless I'm writing data to a specific drive in a specific PC, in which case the
transfer speed drops to about 680Mbps. If the data transfer involves the WAN,
then all bets are off since there are too many variables not under my control.
As long as the LAN is solid, then I've done just about all that I can. Oh,
WAN-based speed tests, too, but you said you've done that and they're fine.

On 8/26/2023 9:45 PM, David E. Ross wrote:
> Windows 7 Ultimate SP1 x64
> Cable modem: Motorola MB7220 V1.0 (2016)
> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>
> Yes, the modem and router are separate devices. Is there a third-party
> test application for me to judge the ongoing reliability of these devices?
>
> No, I do not mean a speed test. Speed test Web sites give good results.
> However, many times it seems that actual uploads and downloads are much
> slower. I want to know if I have to replace either of these soon.
>

Motorola 8x4 DOCSIS 3.0 Cable Modem, 343 Mbps (MB7220)

The problem with the WNDR3400 v3, is it is 10/100 BT on all five Ethernet ports.
That means the connection to the DOCSIS Modem is at 100 Mbps, the modem
on the other hand, is 343 Mbps in some theoretical best-case ("channel bonding"???).

* WiFi Transmitters/Receivers (Tx/Rx) - 2x2 (2.4GHz) + 2x2 (5GHz)
* 530MHz powerful MIPS 32-bit processor
* Memory: 8 MB flash and 64 MB RAM

The WNDR3700 might have GbE on the Ethernet ports.

The Wifi, usually you count on no more than 50% of the box-label rate
as a best case. Then, the client end (your laptop), has to share some
of the claimed dual-band behavior. And so on. Neighborhood usage of
the same RF channel by your neighbors, degrades your thruput.

The "WAN to LAN routing rate", is what the MIPS processor manages
when handling packets one at a time. LAN to LAN Ethernet ports
use a hardware switch at wire rate for local traffic. The Wifi
might need MIPS processor handling for local Wifi traffic as well.

Firing up the old search engine, tells me I have to climb in a
time machine and go back ten years, when search engines actually worked.
It's virtually impossible to get the smallnetbuilder.com site to cough
up any data whatsoever on your router. It's become a Future Corporation
Advertising Hell Hole Filled With Ladies Underwear. And using the
selector buttons on charts, is a waste of time.

*******

If I had money burning a hole in my pocket, I'd look at new router,
but that's going to "take more than ten minutes to select",
because there are so many potential issues with them.
That is why you'd have to track down some reviews.

Oh yeah, that search engine issue again. Oh, well.

*******

A very very quick check, Wifi 7 doesn't seem to be here yet,
leaving Wifi 6. I would think MIMO would be worth more to you
than whizzy channel bandwidth (160MHz) because of potential
greenfield issues. (If the kit sniffs and finds N150 devices in
the neighbourhood, it may alter what channel width choices you
can use.) Again, it's the old thing about "take some fraction
of the exaggerated value on the tin". So while the spec is
pretty, maybe fitting this in your laptop, along with a new router,
only gives a small bump on thruput.

https://www.intel.com/content/www/us/en/products/sku/130293/intel-wifi-6-ax201-gig/specifications.html

MIMO is why some of the plastic routers for sale, have eight
antennas on them, and look like an "inverted dead bug" :-)
You can pay $500 for a "gamer router", and what do you get
for the money, other than black plastic gloss finish for your trouble.
And with the inability to track down reviews, you might never
figure out the good from the bad.

At least some of the new routers, can do WAN to LAN at Gbe rates.
Which is, roughly speaking, 50x better than my router :-)

Good cooling, requires that those router cases have some
thermal vents for convection airflow.

Visit a real computer store, and talk to a clerk I guess.
Not a Best Buy (as they're just going to box-shift on you
and not answer any of your questions). The staff at my
computer store are demoralized, as the owner is now bat shit crazy.
The other day, I actually had to shop at Staples instead :-/
And you know this is the wrong timeline, when you find Paul
shopping at Staples. Next I'll be caught buying tea cozies in
Nordstroms.

Paul

In message <lrolei9erg48ku660l3r18o8vf4f518191@4ax.com> at Sun, 27 Aug
2023 01:10:25, Char Jackson <none@none.invalid> writes
[]
>Speed test sites are generally configured to deliver as much speed, through
>every link in the chain from them to you, as your system can handle. Other
>download or upload sites may not have the same design priorities, so
>the problem
>could be them and not you.

Indeed; there may be nothing wrong at your end.
>
>How are you measuring speed/throughput/bandwidth, whatever you want to
>call it?
>I use a small app called Bandwidth Monitor on each of my PCs, but it's
>not free.
>Someone posted a free open-source app that works in a similar way, but I don't
>remember the name.
>http://www.bwmonitor.com/ - not free, but has 30-day trial
[]
https://codebox.net/pages/bitmeter2, though ancient, is free, and I find
quite usable and useful.
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

I'm the oldest woman on primetime not baking cakes.
- Anne Robinson, RT 2015/8/15-21

"Paul" <nospam@needed.invalid> wrote

|
| Visit a real computer store, and talk to a clerk I guess.
| Not a Best Buy (as they're just going to box-shift on you
| and not answer any of your questions). The staff at my
| computer store are demoralized, as the owner is now bat shit crazy.
| The other day, I actually had to shop at Staples instead :-/
| And you know this is the wrong timeline, when you find Paul
| shopping at Staples. Next I'll be caught buying tea cozies in
| Nordstroms.
|

I visited a Microcenter recently. It was full of clerks
desperate to help. Turns out they're having a hard time
against online competition. I hope they don't close. There's
really nothing else left.

I used to go to Staples for fun and see what was on sale.
Good deals were frequent. Now I think of it as a 24-hour
store: Limited stock, but handy if you don't mind paying
$35 for a USB cord that you're desperate to get. I suppose
they know that most people are buying online. So they may
as well gouge the impulse shoppers.

But Staples does, at least, still sell a selection of routers.
Best Buy has descended to the level of Target. You might
get lucky, but in general they don't really sell computer
equipment any more -- just a few odds and ends.

How about Newegg bought many items from them

KenW

On Sun, 27 Aug 2023 06:59:27 -0600, KenW <ken1943@invalid.net> wrote:

>How about Newegg bought many items from them

When it comes to computer-related things, I buy almost everything from
either Newegg or Amazon. Depending on what I want to buy, sometimes
one is cheaper, sometimes the other.

On 8/26/2023 6:45 PM, David E. Ross wrote:
> Windows 7 Ultimate SP1 x64
> Cable modem: Motorola MB7220 V1.0 (2016)
> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>
> Yes, the modem and router are separate devices. Is there a third-party
> test application for me to judge the ongoing reliability of these devices?
>
> No, I do not mean a speed test. Speed test Web sites give good results.
> However, many times it seems that actual uploads and downloads are much
> slower. I want to know if I have to replace either of these soon.
>

I guess I should have provided more details.

No, I do not use WiFi. I have disabled it on my router.

My LAN seems to have good speed.

My Internet connection is via Spectrum. The problem is surfing the Web,
uploading large files to my own Web site, and downloading application
updates. A recent multi-megabyte upload to my Web site ran at less than
4 Mb/sec and took well over 10 minutes. Spectrum claimed the speed
being slower than their ads was because my modem is too old. That claim
makes me want to test my modem to see if it is failing.

--
David E. Ross
<http://www.rossde.com/>

Beyond Meat, Impossible Meat, and other such meat substitutes
represent the ultimate in ultra-processed foods, which my doctor
told me to avoid. Real meat is natural. Beyond Meat and
Impossible Meat are definitely not natural.

On 8/27/2023 11:00 AM, David E. Ross wrote:
> On 8/26/2023 6:45 PM, David E. Ross wrote:
>> Windows 7 Ultimate SP1 x64
>> Cable modem: Motorola MB7220 V1.0 (2016)
>> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>>
>> Yes, the modem and router are separate devices. Is there a third-party
>> test application for me to judge the ongoing reliability of these devices?
>>
>> No, I do not mean a speed test. Speed test Web sites give good results.
>> However, many times it seems that actual uploads and downloads are much
>> slower. I want to know if I have to replace either of these soon.
>>
>
> I guess I should have provided more details.
>
> No, I do not use WiFi. I have disabled it on my router.
>
> My LAN seems to have good speed.
>
> My Internet connection is via Spectrum. The problem is surfing the Web,
> uploading large files to my own Web site, and downloading application
> updates. A recent multi-megabyte upload to my Web site ran at less than
> 4 Mb/sec and took well over 10 minutes. Spectrum claimed the speed
> being slower than their ads was because my modem is too old. That claim
> makes me want to test my modem to see if it is failing.
>

"My LAN seems to have good speed."

No, not if the product description says 10/100 BT. That's 12.5MB/sec.
How would you transfer a VM between your Daily Driver and your Test Machine,
at that speed ? That would feel like my floppy-speed point to point transfers over
Bluetooth with fsquirt.exe :-) (uh, don't do that)

You can buy a Netgear GbE switch, connect it to one LAN port, and
connect the room machines via GbE. That would at least improve the
LAN to LAN wirespeed interconnect.

*******

In the old days, something like this might not have worked due
to MAC addr issues. Maybe the cable setup is "used to" the MAC addr
of the Netgear router. In any case, you do not leave Windows
connected like this, for longer than is necessary. If the left-most
box was a combo modem/router and using IPV4 NAT, you'd feel "marginally safer".
IPV6 relies on security by obscurity and large numeric spaces for
its "security". A cable modem, could be hiding a one-port-router inside
and terminating the wire encapsulation method.

wired
Cable Modem --------- Test Machine [Browser to speedtest.net]

That setup will help verify the contracted rate, and the modem should
channel bond assuming channels are available on your segment. Speedtest.net
should not have capped connections like maybe a Microsoft server might.

I ran into one Microsoft server, capped at 80KB/sec, and using multiple connections
and aria2c.exe , that helped fill my link and cut the estimated download
time from the promised 24 hour value. It's strange when a server has a pissy
per-connection limit, yet the server also allows you to open a lot of
connections in parallel. When staff set up a server that way, the "messaging"
of the situation is unclear. Does that mean "Tech Bro only need apply" ?

If your contracted rate is 100/10 and capped by the hardware,
you expect the speedtest.net readout to reflect that.

"Speedtest.net" is hosted in a box at the ISP and does not
travel over the Internet. In the same sense as some Netflix
server capacity is located inside an ISP building as well.
Doing this, reduces their "transit internet" expenses. For
speedtest, ISP hosting improves the accuracy of the test.
It does not particularly wring out the entire ISP performance,
and is purely a modem test in your case. It's to prove the wire,
the bonded channels on the segment, are working to at least
the contracted level. You would be unlikely to hit 343Mbit/sec
that way. The output of the modem should be GbE Ethernet,
as it would be pretty silly to offer a 343Mbit/sec modem with
100Mbit/sec wiring.

*******

If we do this... the 100 Mbit/sec limit of the router is what we're measuring.

wired wired
Cable Modem --------- WNDR3400v3 --------- Test Machine [Browser to speedtest.net]
(343?) 1000 100 Measures 100 ?

*******

I have added a GbE switch to mine, to improve LAN to LAN performance.
My contracted rate is 15/1 so the wiring on Modem and Router
does not affect getting the full rate.

ADSL2 100 100 GbE 1000
Modem ------------ Router ----- Switch ----- Daily Driver
----- Test Machine
1000

*******

ISP DNS performance (the rate that www.sun.com gets translated to 1.2.3.4
address), affects interactive performance. Poor DNS may ruin the
web browsing experience. On your Daily Driver, changing the DNS
setting temporarily to 8.8.8.8 and clearing web cache folder,
may identify whether the ISP DNS is performant. There are several
public DNS, and 8.8.8.8 is simply the first example (provided by Google).

Windows used to have the problem, that if there were two [ISP] DNS
addresses, the first one was "down", Windows is too stupid to
reference the "up" address on the first try. It always tries
to access the "down" DNS first, then it tries the "up" DNS next.
At the time this was noted, Linux was properly configured so it
would continue to use the working DNS first, out of two provided
DNS addresses. Since my current ISP actually knows what they're doing,
I no longer see *any* DNS problems. Now, that's service :-) When
the ISP drops on failure, sometimes my VOIP works and nothing
else works, and there must have been enough DNS so the VOIP would
come up.

You can get 10GbE cards for desktops, for around $130, but
since the company making the chip was bought up, I don't
know how that has affected the market dynamics. RealTek has flooded
the market with 2.5GbE chips, and that's more likely to be
an option on a modern motherboard. Having 10GbE chippage on
a new motherboard, is possible but less likely. So there are options
for something more interesting than boring GbE chippage.

10/100BT junk room machines
GbE Current computer room standard
2.5GbE RealTek play
10GbE Money no object setup (could be done point to point and no router)

Paul

On 8/27/2023 11:13 AM, Paul wrote:
> On 8/27/2023 11:00 AM, David E. Ross wrote:
>> On 8/26/2023 6:45 PM, David E. Ross wrote:
>>> Windows 7 Ultimate SP1 x64
>>> Cable modem: Motorola MB7220 V1.0 (2016)
>>> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>>>
>>> Yes, the modem and router are separate devices. Is there a third-party
>>> test application for me to judge the ongoing reliability of these devices?
>>>
>>> No, I do not mean a speed test. Speed test Web sites give good results.
>>> However, many times it seems that actual uploads and downloads are much
>>> slower. I want to know if I have to replace either of these soon.
>>>
>>
>> I guess I should have provided more details.
>>
>> No, I do not use WiFi. I have disabled it on my router.
>>
>> My LAN seems to have good speed.
>>
>> My Internet connection is via Spectrum. The problem is surfing the Web,
>> uploading large files to my own Web site, and downloading application
>> updates. A recent multi-megabyte upload to my Web site ran at less than
>> 4 Mb/sec and took well over 10 minutes. Spectrum claimed the speed
>> being slower than their ads was because my modem is too old. That claim
>> makes me want to test my modem to see if it is failing.
>>
>
> "My LAN seems to have good speed."
>
> No, not if the product description says 10/100 BT. That's 12.5MB/sec.
> How would you transfer a VM between your Daily Driver and your Test Machine,
> at that speed ? That would feel like my floppy-speed point to point transfers over
> Bluetooth with fsquirt.exe :-) (uh, don't do that)
>
> You can buy a Netgear GbE switch, connect it to one LAN port, and
> connect the room machines via GbE. That would at least improve the
> LAN to LAN wirespeed interconnect.
>
> *******
>
> In the old days, something like this might not have worked due
> to MAC addr issues. Maybe the cable setup is "used to" the MAC addr
> of the Netgear router. In any case, you do not leave Windows
> connected like this, for longer than is necessary. If the left-most
> box was a combo modem/router and using IPV4 NAT, you'd feel "marginally safer".
> IPV6 relies on security by obscurity and large numeric spaces for
> its "security". A cable modem, could be hiding a one-port-router inside
> and terminating the wire encapsulation method.
>
> wired
> Cable Modem --------- Test Machine [Browser to speedtest.net]
>
> That setup will help verify the contracted rate, and the modem should
> channel bond assuming channels are available on your segment. Speedtest.net
> should not have capped connections like maybe a Microsoft server might.
>
> I ran into one Microsoft server, capped at 80KB/sec, and using multiple connections
> and aria2c.exe , that helped fill my link and cut the estimated download
> time from the promised 24 hour value. It's strange when a server has a pissy
> per-connection limit, yet the server also allows you to open a lot of
> connections in parallel. When staff set up a server that way, the "messaging"
> of the situation is unclear. Does that mean "Tech Bro only need apply" ?
>
> If your contracted rate is 100/10 and capped by the hardware,
> you expect the speedtest.net readout to reflect that.
>
> "Speedtest.net" is hosted in a box at the ISP and does not
> travel over the Internet. In the same sense as some Netflix
> server capacity is located inside an ISP building as well.
> Doing this, reduces their "transit internet" expenses. For
> speedtest, ISP hosting improves the accuracy of the test.
> It does not particularly wring out the entire ISP performance,
> and is purely a modem test in your case. It's to prove the wire,
> the bonded channels on the segment, are working to at least
> the contracted level. You would be unlikely to hit 343Mbit/sec
> that way. The output of the modem should be GbE Ethernet,
> as it would be pretty silly to offer a 343Mbit/sec modem with
> 100Mbit/sec wiring.
>
> *******
>
> If we do this... the 100 Mbit/sec limit of the router is what we're measuring.
>
> wired wired
> Cable Modem --------- WNDR3400v3 --------- Test Machine [Browser to speedtest.net]
> (343?) 1000 100 Measures 100 ?
>
> *******
>
> I have added a GbE switch to mine, to improve LAN to LAN performance.
> My contracted rate is 15/1 so the wiring on Modem and Router
> does not affect getting the full rate.
>
> ADSL2 100 100 GbE 1000
> Modem ------------ Router ----- Switch ----- Daily Driver
> ----- Test Machine
> 1000
>
> *******
>
> ISP DNS performance (the rate that www.sun.com gets translated to 1.2.3.4
> address), affects interactive performance. Poor DNS may ruin the
> web browsing experience. On your Daily Driver, changing the DNS
> setting temporarily to 8.8.8.8 and clearing web cache folder,
> may identify whether the ISP DNS is performant. There are several
> public DNS, and 8.8.8.8 is simply the first example (provided by Google).
>
> Windows used to have the problem, that if there were two [ISP] DNS
> addresses, the first one was "down", Windows is too stupid to
> reference the "up" address on the first try. It always tries
> to access the "down" DNS first, then it tries the "up" DNS next.
> At the time this was noted, Linux was properly configured so it
> would continue to use the working DNS first, out of two provided
> DNS addresses. Since my current ISP actually knows what they're doing,
> I no longer see *any* DNS problems. Now, that's service :-) When
> the ISP drops on failure, sometimes my VOIP works and nothing
> else works, and there must have been enough DNS so the VOIP would
> come up.
>
> You can get 10GbE cards for desktops, for around $130, but
> since the company making the chip was bought up, I don't
> know how that has affected the market dynamics. RealTek has flooded
> the market with 2.5GbE chips, and that's more likely to be
> an option on a modern motherboard. Having 10GbE chippage on
> a new motherboard, is possible but less likely. So there are options
> for something more interesting than boring GbE chippage.
>
> 10/100BT junk room machines
> GbE Current computer room standard
> 2.5GbE RealTek play
> 10GbE Money no object setup (could be done point to point and no router)
>
> Paul
>

I have no "test machine". The LAN consists of my PC and my wife's PC,
connected via Ethernet cables through the router. This is all described
at <http://www.rossde.com/computer/LAN.html>. My wife has made much of
her hard drive visible to me via the LAN. I reciprocated.

My wife's PC is OLD. It does not have a USB3 port. The portable hard
drive on which I archive backups uses USB3. (See
<http://www.rossde.com/computer/computer_backup.html>.) I have no
problem and little delay plugging the drive into my PC (which has USB3
ports) and copying my wife's backups though our router. I do the
copying from my PC (fetching data) and not from my wife's (sending data).

--
David E. Ross
<http://www.rossde.com/>

Beyond Meat, Impossible Meat, and other such meat substitutes
represent the ultimate in ultra-processed foods, which my doctor
told me to avoid. Real meat is natural. Beyond Meat and
Impossible Meat are definitely not natural.

"KenW" <ken1943@invalid.net> wrote

| How about Newegg bought many items from them
|

I've never tried them. I like to buy local if I can, so I
can see the product and return it easily if necessary. And
I don't want to support Amazon. Years ago I used to build
my own and went to TigerDirect for that, after all the
local suppliers closed. But then they sold out and ended
up just being a middleman front for various suppliers.

These days I'm not really interested in building myself.
Processors are costing more than the whole kit used to cost
and I'm not running Win10 very often. If I need a new
computer it will probably make the most sense to just buy one
of those $350 bottom end boxes from Staples. For more
involved things I trek to Microcenter. (500 GB Samsung SSD
this week for $40. You can't afford not to buy it!)

On Sun, 27 Aug 2023 09:12:41 +0100, "J. P. Gilliver" <G6JPG@255soft.uk> wrote:

>In message <lrolei9erg48ku660l3r18o8vf4f518191@4ax.com> at Sun, 27 Aug
>2023 01:10:25, Char Jackson <none@none.invalid> writes
>[]
>>Speed test sites are generally configured to deliver as much speed, through
>>every link in the chain from them to you, as your system can handle. Other
>>download or upload sites may not have the same design priorities, so
>>the problem
>>could be them and not you.
>
>Indeed; there may be nothing wrong at your end.
>>
>>How are you measuring speed/throughput/bandwidth, whatever you want to
>>call it?
>>I use a small app called Bandwidth Monitor on each of my PCs, but it's
>>not free.
>>Someone posted a free open-source app that works in a similar way, but I don't
>>remember the name.
>>http://www.bwmonitor.com/ - not free, but has 30-day trial
>[]
>https://codebox.net/pages/bitmeter2, though ancient, is free, and I find
>quite usable and useful.

Thanks, John. That's the one I was trying to think of. If I wasn't happy with
what I have, I'd check out that one.

On Sun, 27 Aug 2023 07:45:58 -0700, Ken Blake <Ken@invalid.news.com> wrote:

>On Sun, 27 Aug 2023 06:59:27 -0600, KenW <ken1943@invalid.net> wrote:
>
>>How about Newegg bought many items from them
>
>When it comes to computer-related things, I buy almost everything from
>either Newegg or Amazon. Depending on what I want to buy, sometimes
>one is cheaper, sometimes the other.

Exactly the same here, for me. Newegg is much easier to shop on than Amazon,
where search results for 'router' will eventually include household napkins and
gardening shears if you scroll far enough, but before I pull the trigger I check
both sites. I also use https://camelcamelcamel.com to see if the current price
is a good price or not. I use two features there, price history and 'alert me if
the price drops'.

David E. Ross wrote:
> Windows 7 Ultimate SP1 x64
> Cable modem: Motorola MB7220 V1.0 (2016)
> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>
> Yes, the modem and router are separate devices. Is there a third-party
> test application for me to judge the ongoing reliability of these devices?
>
> No, I do not mean a speed test. Speed test Web sites give good results.
> However, many times it seems that actual uploads and downloads are much
> slower. I want to know if I have to replace either of these soon.

Your modem is 1 Gb but your router/switch is 100 Mb. :(

Some modems and routers have built in loopback and cable tests that you
can run. Log in to your modem/router/switch and see. Or you can just
run PING and examine the results for R/T time and packet loss. That
would be easiest.

IMO, most browser type online speed tests do not work very well and give
conflicting results. Speed test software that reside on the computer
tends to give better results. I like the one from Speedtest.net.
https://www.speedtest.net/apps/windows
It's 50 meg d/l and 200 meg installed.

All of my routers/switches/modems have lasted 3-5 years before getting
slower or flakey. Heat eats them up. Taking the covers off of the
devices and using a small fan can extend their lives. Along with gluing
heat sinks to the IC's.

On Sun, 27 Aug 2023 08:00:51 -0700, "David E. Ross" <nobody@nowhere.invalid>
wrote:

>On 8/26/2023 6:45 PM, David E. Ross wrote:
>> Windows 7 Ultimate SP1 x64
>> Cable modem: Motorola MB7220 V1.0 (2016)
>> Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
>>
>> Yes, the modem and router are separate devices. Is there a third-party
>> test application for me to judge the ongoing reliability of these devices?
>>
>> No, I do not mean a speed test. Speed test Web sites give good results.
>> However, many times it seems that actual uploads and downloads are much
>> slower. I want to know if I have to replace either of these soon.
>>
>
>I guess I should have provided more details.
>
>No, I do not use WiFi. I have disabled it on my router.

Good, I'd say.

>My LAN seems to have good speed.

You haven't said what 'good' means.

>My Internet connection is via Spectrum. The problem is surfing the Web,
>uploading large files to my own Web site, and downloading application
>updates. A recent multi-megabyte upload to my Web site ran at less than
>4 Mb/sec and took well over 10 minutes. Spectrum claimed the speed
>being slower than their ads was because my modem is too old. That claim
>makes me want to test my modem to see if it is failing.

I thought you've done that already, by visiting one or more speed test sites,
but you haven't mentioned those numbers, either. Check your Spectrum plan to see
what numbers they use, then compare that to what you see on various speed test
sites. Run those tests multiple times throughout the day, always when no other
traffic-sucking activities are going on.

Since your modem and router are separate devices, you can check your modem
separately by connecting your PC directly to the Ethernet port on the modem,
then visiting the speed test sites. You'll have to reboot the modem, then reboot
the PC, to establish the new network connection since the modem only binds to
the first MAC address that it sees after a reboot, and the PC will have to
acquire a new IP address from the modem. (Actually, from the CMTS which is
upstream of the modem, but pretend it's coming from the modem to keep things
simple.)

If a direct connection to the modem gets you close to the promised speed, while
a connection via the router & modem is much slower, then your cable modem is
fine and your router is the limiting factor.

If a direct connection to the modem is still much slower than what you're paying
for from Spectrum, then your cable modem is likely the culprit. If it's rented,
consider swapping it for something better. If it's yours, decide if a newer
model is worth it.

BTW, if you haven't yet, you can browse to the modem's web interface by visiting
http://192.168.100.1. The most interesting page will likely be the page that you
land on, which should show you the upstream and downstream signal strengths,
SNR, number of bonded channels, etc. Downstream signal strength should be within
+-15dB, with the closer to 0dB the better, while upstream signal strength must
be less than about +55dB and hopefully somewhere between +35dB to maybe +45dB.

Signal strengths that are out of limits can result in stalled connections,
dropped connections, poor or inconsistent throughput, among other things.

In message <69dneit45iivldqn5o49u3qir6ls3fa0hq@4ax.com> at Sun, 27 Aug
2023 16:02:58, Char Jackson <none@none.invalid> writes
[]
>BTW, if you haven't yet, you can browse to the modem's web interface by
>visiting
>http://192.168.100.1. The most interesting page will likely be the page
[]
Although most such web interfaces I've come across are of the
192.168.x.y variety (I think I've seen x.y be 0.0, 0.1, 1.0, and 1.1,
but I might be wrong about that), I have also occasionally encountered
one that's 10.0.0.2 .
--
J. P. Gilliver. UMRA: 1960/<1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Security is the perfect excuse to lock you out of your own computer.
- Mayayana in alt.windows7.general, 2015-12-4

On 8/27/2023 10:07 PM, J. P. Gilliver wrote:
> In message <69dneit45iivldqn5o49u3qir6ls3fa0hq@4ax.com> at Sun, 27 Aug 2023 16:02:58, Char Jackson <none@none.invalid> writes
> []
>> BTW, if you haven't yet, you can browse to the modem's web interface by visiting
>> http://192.168.100.1. The most interesting page will likely be the page
> []
> Although most such web interfaces I've come across are of the 192.168.x.y variety (I think I've seen x.y be 0.0, 0.1, 1.0, and 1.1, but I might be wrong about that), I have also occasionally encountered one that's 10.0.0.2 .

There are meaning to those numbers.

192.168.x.y
10.x.y.z
172.

https://en.wikipedia.org/wiki/Private_network

You can widen the netmask while the OS is running. Powershell command.
But the results are not saved for the next time, so after
a reboot, it's all back to normal.

Paul

On 8/27/2023 4:02 PM, Newyana2 wrote:
> "KenW" <ken1943@invalid.net> wrote
>
> | How about Newegg bought many items from them
> |
>
> I've never tried them. I like to buy local if I can, so I
> can see the product and return it easily if necessary. And
> I don't want to support Amazon. Years ago I used to build
> my own and went to TigerDirect for that, after all the
> local suppliers closed. But then they sold out and ended
> up just being a middleman front for various suppliers.
>
> These days I'm not really interested in building myself.
> Processors are costing more than the whole kit used to cost
> and I'm not running Win10 very often. If I need a new
> computer it will probably make the most sense to just buy one
> of those $350 bottom end boxes from Staples. For more
> involved things I trek to Microcenter. (500 GB Samsung SSD
> this week for $40. You can't afford not to buy it!)

You can make a very nice computer for yourself, with a 5600G.
And the price has been dropping on those, as time passes.
Eventually, they will be discontinued. That's a six core, that
sips power. The G means there is a GPU inside, so you don't
need a video card (a $200 savings). Memory type is DDR4.
There are two motherboard types, and you use the one that
sips power for that.

The drop in the price of the processor, is making the
thing almost attractive.

$129 (comes with cooler)

https://www.newegg.com/amd-ryzen-5-5600g-ryzen-5-5000-g-series/p/N82E16819113683

The high end processor on the other hand, the mains power
meter shows it draws 240W when I run 7ZIP to compress some
files. Even the power supply seems to be getting tired of this :-)
The 12V rail on the power supply, is running 11.6 volts.

[Picture]

https://i.postimg.cc/hhWtJ5zg/AMD-ryzen-master-control-software.gif

AMD says "they recommend water cooling for the processor", and,
they're right. You cannot get absolutely the highest speed using
air. The air cooler I have currently, does a damn fine job, and
impressed the hell out of me. I thought the manufacturer was
full of shit, but, it actually works. Shocked. It's not a
Noctua cooler.

But for the 5600G, you don't need anything (cooler is in CPU box).
You will never need that piece of software above :-) I wasn't even
aware the motherboard had closed loop cooling control, until the
high power processor was in there, and the fucking thing turned
into a bucking bronco. At first, there were temperature transients,
up to 90C. Putting the right cooler on it, you can use it now.
These would be transients, which were hard for the control loop
to handle. The right cooler, provides a bit of thermal inertia.

So you're right, good processors are a pain. The smaller processors
are a lot less work, and you don't have to know anything to run those.
Just boot and go. They have turned out to be "night and day different".

if you're a builder who does not want hassles, the small processor
is "recommended".

Paul

On Mon, 28 Aug 2023 03:07:43 +0100, "J. P. Gilliver" <G6JPG@255soft.uk> wrote:

>In message <69dneit45iivldqn5o49u3qir6ls3fa0hq@4ax.com> at Sun, 27 Aug
>2023 16:02:58, Char Jackson <none@none.invalid> writes
>[]
>>BTW, if you haven't yet, you can browse to the modem's web interface by
>>visiting
>>http://192.168.100.1. The most interesting page will likely be the page
>[]
>Although most such web interfaces I've come across are of the
>192.168.x.y variety (I think I've seen x.y be 0.0, 0.1, 1.0, and 1.1,
>but I might be wrong about that), I have also occasionally encountered
>one that's 10.0.0.2 .

Mr. Google says that modem's web server lives at 192.168.100.1. That's also the
most common address used with cable modems, so even if we didn't know what to
use I would try that one first.

"Paul" <nospam@needed.invalid> wrote

| You can make a very nice computer for yourself, with a 5600G.
| And the price has been dropping on those, as time passes.
| Eventually, they will be discontinued. That's a six core, that
| sips power. The G means there is a GPU inside, so you don't
| need a video card (a $200 savings). Memory type is DDR4.
| There are two motherboard types, and you use the one that
| sips power for that.
| | The drop in the price of the processor, is making the
| thing almost attractive.
| | $129 (comes with cooler)

Not bad. Last I looked it seemed like $300 was the bottom.
On the other hand, my current FX-8300 was $65, with another
$65 for the motherboard. The grand total was about $400.
(Also watercooled.)

With the gradual degradation of Windows it's hard to be
inspired. What I have works well. When it dies? I don't know,
but Win10/11 are not systems that I'd expect to have fun with.
Linux will never be ready for prime time...

I'll probably just become one of those old people who only
needs a computer that's "good enough for email and browsing".
As long as I can get email and log into my dentist and urologist
websites, that may be enough. I'll just switch to more gardening
and take a laptop out of the closet when I need Internet. :)

On Sun, 27 Aug 2023 15:42:01 -0500, Char Jackson <none@none.invalid>
wrote:

>On Sun, 27 Aug 2023 07:45:58 -0700, Ken Blake <Ken@invalid.news.com> wrote:
>
>>On Sun, 27 Aug 2023 06:59:27 -0600, KenW <ken1943@invalid.net> wrote:
>>
>>>How about Newegg bought many items from them
>>
>>When it comes to computer-related things, I buy almost everything from
>>either Newegg or Amazon. Depending on what I want to buy, sometimes
>>one is cheaper, sometimes the other.

By the way, an aside: it's for that reason that my current computer,
was built from a mixture of parts I bought from both sources.

>Exactly the same here, for me. Newegg is much easier to shop on than Amazon,

That's not my experience at all. I don't find a big difference between
them, but if I had to choose which one I thought was easier to shop
on, I'd choose Amazon.

On the other hand, perhaps that's because I have much more experience
shopping on Amazon, not just for computer-related things,

>where search results for 'router' will eventually include household napkins and
>gardening shears if you scroll far enough,

Do you change the category from "All" to "Computers"? I just tried it
here with that setting. I didn't scroll through all the Router pages,
but, I went through several and never saw anything but computer
routers,

> but before I pull the trigger I check
>both sites. I also use https://camelcamelcamel.com to see if the current price
>is a good price or not. I use two features there, price history and 'alert me if
>the price drops'.

More often than not, when I shop on the internet, it's for something I
want or need now, so "alert me if the price drops" is not usually a
good choice for me. Yes, I could perhaps save a few dollars if I were
more patient, but it's unlikely to be more than a very few.

On Sun, 27 Aug 2023 16:02:39 -0400, "Newyana2"
<Newyana2@invalid.nospam> wrote:

>"KenW" <ken1943@invalid.net> wrote
>
>| How about Newegg bought many items from them
>|
>
> I've never tried them. I like to buy local if I can,

Even if the price is considerably higher? Not me.

>so I
>can see the product

To me, it depends on the product. Some things I might want to see,
others I don't. Most often, especially with computer-related things,
I don't particularly want to see it.

>and return it easily if necessary.

I've never returned anything to Newegg, but my experience with Amazon
returns is that it's very easy--easier than with a local store . I can
return things to Amazon by taking them to a nearby Whole Foods store,
which is much closer than the only decent computer parts store here in
town.

And for me, having things quickly delivered instead of driving and
picking them up (especially if they're heavy) is a big benefit.

>And
>I don't want to support Amazon.

Your choice, of course. Not mine. Amazon is a great convenience for me
and saves me a lot of money. I buy almost everything except fresh food
from them.

>Years ago I used to build
>my own and went to TigerDirect for that, after all the
>local suppliers closed. But then they sold out and ended
>up just being a middleman front for various suppliers.

> These days I'm not really interested in building myself.
>Processors are costing more than the whole kit used to cost
>and I'm not running Win10 very often. If I need a new
>computer it will probably make the most sense to just buy one
>of those $350 bottom end boxes from Staples.

That of course depends on what level of performance you want and what
your budget is. I try to buy a new computer as seldom as possible, so
I always try to get something as close to state-of-the-art as I can
afford. My last computer--home-built two years ago, mostly by a friend
who is younger than I am (I'm 85) and more skilled at building--is one
that I hope will last me the rest of my life. It's unlikely that I'd
be happy with a $350 bottom-end box from Staples or that it would last
me as long.

>For more
>involved things I trek to Microcenter. (500 GB Samsung SSD
>this week for $40. You can't afford not to buy it!)

I'm not interested in 500GB drives. It's too small for me. My current
computer has a 1TB NVMe drive and two 2TB HDs.

On Mon, 28 Aug 2023 09:24:51 -0400, "Newyana2"
<Newyana2@invalid.nospam> wrote:

>"Paul" <nospam@needed.invalid> wrote
>
>| You can make a very nice computer for yourself, with a 5600G.
>| And the price has been dropping on those, as time passes.
>| Eventually, they will be discontinued. That's a six core, that
>| sips power. The G means there is a GPU inside, so you don't
>| need a video card (a $200 savings). Memory type is DDR4.
>| There are two motherboard types, and you use the one that
>| sips power for that.
>|
>| The drop in the price of the processor, is making the
>| thing almost attractive.
>|
>| $129 (comes with cooler)
>
> Not bad. Last I looked it seemed like $300 was the bottom.

$300 is nowhere near the bottom. I don't want to recommend any
particular CPU, but to see where the bottom was I just checked on
Amazon.com. They have an Intel i3 for $71. And lots of AMDs for much
less than $300.

>On the other hand, my current FX-8300 was $65, with another
>$65 for the motherboard. The grand total was about $400.
>(Also watercooled.)
>
> With the gradual degradation of Windows

Your opinion, not mine. Yes, like you I don't like many of the current
Windows defaults, but as far as I'm concerned, they are just
defaults--there are easy workarounds and third-party utilities to
greatly improve it. I'm happy with Windows 11.

> it's hard to be
>inspired. What I have works well.

As does my computer with Windows 11 here.

>When it dies? I don't know,
>but Win10/11 are not systems that I'd expect to have fun with.

Most people don't expect to "have fun" with something new and
different. But like most things, spending some time with Windows 10 or
11 and learning what workarounds and utilities are best for you, can
quickly change that viewpoint.

>Linux will never be ready for prime time...

No comment. I don't know enough about Linux to have an opinion.

> I'll probably just become one of those old people who only
>needs a computer that's "good enough for email and browsing".
>As long as I can get email and log into my dentist and urologist
>websites, that may be enough. I'll just switch to more gardening
>and take a laptop out of the closet when I need Internet. :)

"David E. Ross" <nobody@nowhere.invalid> wrote

| Windows 7 Ultimate SP1 x64
| Cable modem: Motorola MB7220 V1.0 (2016)
| Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
|

Why not just get a new router? I bought a TP-LINK Archer AX1500
gigabit router 3 years ago because it was middle-of-the-road and
highly rated. They're still selling at Microcenter for $60. You can pay
over $300, but I don't know of any reason you'd need to.

"David E. Ross" <nobody@nowhere.invalid> wrote

| Windows 7 Ultimate SP1 x64
| Cable modem: Motorola MB7220 V1.0 (2016)
| Router: Netgear N600Wireless Router Model WNDR3400v3 (2020)
|

Why not just get a new router? I bought a TP-LINK Archer AX1500
gigabit router 3 years ago because it was middle-of-the-road and
highly rated. They're still selling at Microcenter for $60. You can pay
over $300, but I don't know of any reason you'd need to.