Wireless router, continued...

[Dan Sutter]

I need some advice on what wireless router to purchase.

Well... Maybe it is a good question after all.
At work, (my job), during the same time period, We have had
Netgear, Linksys and Cisco/Linksys. Today at work, we are using
a Cysco/Linksys 3200, for our WI-FI and the customers complain
almost every day that they cannot connect! So we reboot it.

Somethings fishy. How do they authenticate with your AP?
Whitelist of MAC addys? OR whut?

PS. I don't have any experience with DLINK equipment. Dan

Me neither. But I haint done heard no bad things about em neither
:)

§ñühwö£f Thank you §ñühwö£f.
I would like to pursue your "Somethings fishy" comment from above. Let me tell you the parameters: 1.) It is in the back kitchen of a
restaurant. 2.) It is on a dedicated circuit breaker. 3.) It is
plugged into an expensive, one socket surge protector. 4.) It is
installed 6 feet from the floor of an 8 foot space, in a corner on
top a small wire cornered shelf (for it's breathing space). 5.) The
corner is situated so that is as close as we can get to the center
of the building. 6.) It is an AT&T DSL 6MB connection.

If 'distance' were a problem, customers who sat nearest the kitchen
would find fewer problems than those near the most distant wall.
Does everyone have the same level of difficulty?

Exactly. Connect to the AP with a laptop and walk around while checking the signal strength. Steel beams and wiring can interfere with signal.

You don't mention configuration of the WAP, how does it get its IP
address? Is the WAP otherwise configured as pass-through, to some
controller/switch?

Bridge mode?

We have to be PCI compliant, so that means our WI-FI network is
isolated from our in house network,

I don't follow you on that one... What do you believe is the
implication of PCI compliance and WiFi?

yet they share the same internet connection.

How are those boxes 'sharing' the same connection? Does the other
box have Network Address Translation?

Should be sure to be using a different channel as well.

What do you think? Can routers be so sensitive to their
environment?

Yes, that is -possible-. Depends on the kind and density of objects
between the WAP and client, power level and antenna configuration on
the client machine, and so on. '...cannot connect...' suggests a
setup problem.

Or RF from some source. Again, gotta walk around and look for hotspots.
I have two AP's in my house. One has a tiny coverage area and the other gets all the way out to the garage. I set the preamble on the 2nd one to
"long" for better linkup. Gotta dig down into the advanced settings a bit & start tweakin em. Try a different channel, try setting the AP to "open"
then hide the BSSID and only give that to the clients. If that dont work try WPA. But I'd bet its RF from a source and signal dropouts in certain areas.

Thanks again §ñühwö£f and anyone,

But you see, once we isolated the WI-FI from the office, the interruptions for our WI-FI clientele continue
even though our service remains intact. (ever since we split access with a third router at the helm.)
I had once proposed to management that we should cut a hole through the wall and put the WI-FI router right in
the dining room where there is a regulated temperature and less cooking vapors.
They said, "let's just see how the new arrangement (1 modem, 1 main router and 2 dependent routers) works.
[not so good]
Again, I ask, Can routers be so sensitive to their environment?

TIA,
Dan Sutter
PS. I do not see my posts in comp.dcom.net-management










--- news://freenews.netfront.net/ - complaints: ***@netfront.net ---

[Wes Thom]

Post by Dan Sutter

I need some advice on what wireless router to purchase.

Well... Maybe it is a good question after all.
At work, (my job), during the same time period, We have had
Netgear, Linksys and Cisco/Linksys. Today at work, we are using
a Cysco/Linksys 3200, for our WI-FI and the customers complain
almost every day that they cannot connect! So we reboot it.

Somethings fishy. How do they authenticate with your AP?
Whitelist of MAC addys? OR whut?

PS. I don't have any experience with DLINK equipment. Dan

Me neither. But I haint done heard no bad things about em neither
:)

§ñühwö£f Thank you §ñühwö£f.
I would like to pursue your "Somethings fishy" comment from above. Let me tell you the parameters: 1.) It is in the back kitchen of a
restaurant. 2.) It is on a dedicated circuit breaker. 3.) It is
plugged into an expensive, one socket surge protector. 4.) It is
installed 6 feet from the floor of an 8 foot space, in a corner on
top a small wire cornered shelf (for it's breathing space). 5.) The
corner is situated so that is as close as we can get to the center
of the building. 6.) It is an AT&T DSL 6MB connection.

If 'distance' were a problem, customers who sat nearest the kitchen
would find fewer problems than those near the most distant wall.
Does everyone have the same level of difficulty?

Exactly. Connect to the AP with a laptop and walk around while checking the signal strength. Steel beams and wiring can interfere with signal.

You don't mention configuration of the WAP, how does it get its IP
address? Is the WAP otherwise configured as pass-through, to some
controller/switch?

Bridge mode?

We have to be PCI compliant, so that means our WI-FI network is
isolated from our in house network,

I don't follow you on that one... What do you believe is the
implication of PCI compliance and WiFi?

yet they share the same internet connection.

How are those boxes 'sharing' the same connection? Does the other
box have Network Address Translation?

Should be sure to be using a different channel as well.

What do you think? Can routers be so sensitive to their
environment?

Yes, that is -possible-. Depends on the kind and density of objects
between the WAP and client, power level and antenna configuration on
the client machine, and so on. '...cannot connect...' suggests a
setup problem.

Or RF from some source. Again, gotta walk around and look for hotspots.
I have two AP's in my house. One has a tiny coverage area and the other gets all the way out to the garage. I set the preamble on the 2nd one to
"long" for better linkup. Gotta dig down into the advanced settings a bit & start tweakin em. Try a different channel, try setting the AP to "open"
then hide the BSSID and only give that to the clients. If that dont work try WPA. But I'd bet its RF from a source and signal dropouts in certain areas.

Thanks again §ñühwö£f and anyone,
But you see, once we isolated the WI-FI from the office, the interruptions for our WI-FI clientele continue
even though our service remains intact. (ever since we split access with a third router at the helm.)
I had once proposed to management that we should cut a hole through the wall and put the WI-FI router right in
the dining room where there is a regulated temperature and less cooking vapors.
They said, "let's just see how the new arrangement (1 modem, 1 main router and 2 dependent routers) works.
[not so good]
Again, I ask, Can routers be so sensitive to their environment?

You are trying to solve a hardware and an IP problem simultaneously as if same. Simpler is to break the two into separate unknowns. And then solve each. For example, in similar situations, we would connect Wifi and hardwired networks to one POS server on separate Ethernet cards. Then known problems and other unforeseen problems mean one network always works no matter what happens to the other. It also makes security easier.

Second, electronics must work perfectly fine even in a 100 degree F room. Many will operate diagnostics overnight and call that stress testing. It is not. Stress testing is operating electronic hardware in all normal temperatures (including over 100 degree F and below 40 degree F) while running diagnostics (or normal operation). Use normal but extreme temperatures to identify defective hardware that works OK in a 70 degree room. Electronics that fail at those other normal temperatures will often fail months or years later at 70 degrees. Temperature is a toll to even eliminate defects that only cause failures months or a year later.

Third, a dedicated circuit is usually unnecessary. A simpler tool is an incandescent light bulb. Power the bulb from the same receptacle that powers hardware. Does the bulb remain constant intensity? Then power on that circuit is fine. Normal voltage for any electronics is even when that bulb dims to 50% intensity. That bulb is a superior tool for eliminating AC power problems from the suspect list.

Fourth, you have no idea what the Wifi is doing without a signal strength meter. For example, Dell provided one with Wifi software. Another package is Netstumbler. Five bars is useless. To establish WiFi integrity to all portable devices means software on a laptop (or a more expensive dedicated tool) to see the only useful numbers. dBs.

For example, in one building. the concrete wall was only on a lower floor. I could look across the building at the Wifi. And watch the signal diminish as I walked over that concrete wall. I look right at that Wifi and still could not get a useful signal until I walked back over that concrete wall (that was only downstairs). Radio waves at those frequencies have strange behavior.

Remember, leaking microwave ovens take out Wifi. Just another reason for a tool that provides numbers in dBs.

The tool must report tiniest changes in less than a second. Only then can walking around with the laptop report something useful. Again, a most useless tool is something with five bars.

Four examples of how to identify or solve hardware problems. IP problems are a completely different topic. Trying to solve both IP problems and hardware problems simultaneously only created multiple unknowns. As the unknowns (suspects) increase, then a problem's complexity increases 'exponentially'. Separate a hardwired network from the Wifi (using a router does not provide sufficient separation). Use above suggested hardware tools to first establish hardware integrity.

After hardware integrity is established, only then move on to possible IP problems. Above demonstrates how to break a problem into parts. Then solve one part at a time.

[Wes Thom]

If I didn't mention so before. The entire office network is hardwired
and doesn't rely on wireless.
The wireless is ONLY for our patrons.

If I understand those connections, well, I could walk into your store, put my WiFi in promiscuous mode, and collect all hardwired data transactions. Recommended not only to debug your problems. Also recommended for other unmentioned reasons. That Wifi network is best a separate network.

[Dan Sutter]

Post by Wes Thom

If I didn't mention so before. The entire office network is hardwired
and doesn't rely on wireless.
The wireless is ONLY for our patrons.

If I understand those connections, well, I could walk into your store, put my WiFi in promiscuous mode, and collect all hardwired data transactions.
Recommended not only to debug your problems. Also recommended for other unmentioned reasons. That Wifi network is best a separate network.

Thanks Wes,

Your posts gave me a lot of information and a practical approach.
Currently, My wife has gone back to school but had her laptop stolen.
So... She is using mine. It is tough to get a few minutes with it.
(she'll have her own soon, This project wiil begin then).

1st:
I do have MetaGeek's inSSIDer installed. I plan to use it to monitor the wireless
signal from my work's WIFI signal. I will find a standard early in the morning before anyone
is there. Run the microwave several times to check signal strength.
I thought I would have my laptop several feet away at that time.
To see what happens. inSSider has a time line graph without bars. The timeline starts at -100dBm
and goes up to 20dBm. (dBM?)
(To Check hardware issues 1st).

Then, I have already have downloaded "Wireshark" (but I haven't installed it yet).
(for "promiscuous mode" monitoring).
The download page claims:
"Each Windows package comes with the latest stable release of WinPcap, which is required for live packet capture.
If needed you can install the latest development release from the WinPcap download page."
I especially do not want that running while my wife is the primary user.
Even I am not sure I can keep the capture to a safe disk usage.
(I hope I find a hardware problem.)

AND, I still haven't researched the connection count(capacity) available with our current WI-FI router.
Could a router crap out if the demands on it surpassed its connection capacity?

Thanks for all your advise,
Dan Sutter





--- news://freenews.netfront.net/ - complaints: ***@netfront.net ---

[anyone]

Post by Dan Sutter

Post by Wes Thom

If I didn't mention so before. The entire office network is hardwired
and doesn't rely on wireless.
The wireless is ONLY for our patrons.

If I understand those connections, well, I could walk into your
store, put my WiFi in promiscuous mode, and collect all hardwired
data transactions.
Recommended not only to debug your problems. Also recommended for
other unmentioned reasons. That Wifi network is best a separate
network.

Thanks Wes,
Your posts gave me a lot of information and a practical approach.
Currently, My wife has gone back to school but had her laptop stolen.
So... She is using mine. It is tough to get a few minutes with it.
(she'll have her own soon, This project wiil begin then).
I do have MetaGeek's inSSIDer installed. I plan to use it to monitor the
wireless signal from my work's WIFI signal. I will find a standard early
in the morning before anyone is there. Run the microwave several times
to check signal strength.
I thought I would have my laptop several feet away at that time.
To see what happens. inSSider has a time line graph without bars. The
timeline starts at -100dBm and goes up to 20dBm. (dBM?)

"Power ratio in dB, relative to 1 milliwatt" -- look on-line, or:

Ratio, in dBm = 10*log10(Pout/1mW)

"Pout" is whatever power is delivered by a transmitter and measured by test equipment.

Post by Dan Sutter
(To Check hardware issues 1st).
Then, I have already have downloaded "Wireshark" (but I haven't
installed it yet). (for "promiscuous mode" monitoring).
"Each Windows package comes with the latest stable release of WinPcap,
which is required for live packet capture.
If needed you can install the latest development release from the WinPcap download page."
I especially do not want that running while my wife is the primary user.
Even I am not sure I can keep the capture to a safe disk usage.
(I hope I find a hardware problem.)

Wireshark will help you monitor transmitted traffic and response -- number of packets, kind of traffic (ACKs, RSTs,
etc.) -- and is fine for debugging some connectivity issues, but likely won't tell you much about hardware problems. It
usually does its work in an unobtrusive way.

Post by Dan Sutter
AND, I still haven't researched the connection count(capacity) available
with our current WI-FI router.
Could a router crap out if the demands on it surpassed its connection capacity?

Yes, in a sense. All of Ethernet has a feature known as 'back-off' for an event known as a packet 'collision' (not the
same as an address collision). Once a host senses a collision, internal hardware/firmware will cause its transmitter
to cease transmitting for the duration of a randomly-derived time count. Once the internal counter rolls-over, the
transmitter will try again. If there is another collision, it will back-off for another randomly-derived time count, and so
on, until it can successfully complete a transmission. If all of the hosts on a given network were to attempt
transmission at once, all could be forced into back-off, perhaps repeatedly, and successful packet transmission
could drop to a few percent. That's an extreme scenario, but it can occur. That process is a primary reason that
Ethernet protocol is described as "best attempted delivery".

[anyone]

On Fri, 29 Mar 2013 19:58:18 -0400, Dan Sutter wrote:

[...elected to trim this, because it has become difficult to parse question/response/attribution]

Post by Dan Sutter
But you see, once we isolated the WI-FI from the office, the
interruptions for our WI-FI clientele continue even though our service
remains intact. (ever since we split access with a third router at the
helm.)

Please clarify: precisely what do you mean by "...isolated the WiFi..." ?

I had earlier suggested you temporarily disconnect the 'office' network and test WiFi access. Can you do that?

Post by Dan Sutter
I had once proposed to management that we should cut a hole through
the wall and put the WI-FI router right in the dining room where there
is a regulated temperature and less cooking vapors.

Your description suggests that is not the problem.

Post by Dan Sutter
They said, "let's just see how the new arrangement (1 modem, 1 main
router and 2 dependent routers) works. [not so good]
Again, I ask, Can routers be so sensitive to their environment?

Technically yes, but you would need extreme conditions. If that were the case, it seems likely both wired and WiFi
access would suffer, not just one. There is a simple test for that -- try wired and WiFi access after hours, when the
remainder of the restaurant is shut down.

I still believe you should test for 'address collision' -- temporarily disconnect wired (office) connection, see how WiFi
behaves; temporarily disconnect WiFi, see how the wired connection behaves. If you find no difference, then you
have eliminated one possible source of trouble.

Post by Dan Sutter
Dan Sutter PS. I do not see my posts in comp.dcom.net-management

newsgroup added to this one...