So, I have to learn a lot of things on my own as I don't have any networking training, only what I get on the job.

Scenario: I was sent to check on an access point at a condo we provide wifi for. First thing I did was went to the switch that the access point was plugged into and switched ports. The problem, dropping off, followed the WAP. I cable mapped it and it passed. The lights on the wap, though, the ctl and pwr lights, were fluttering intermittently/back and forth/just weird. So, I re-terminated both ends. It passed. I plugged it back in and the lights on the wap were behaving normally. The girl who logged in to the Ruckus ZoneDirector remotely to check on it after I "got it working" said that it was only at 100 when it should be at 1000. (She didn't know that I don't have a whole lot of networking knowledge, just bits and pieces) Apparently this means that while copper to copper cable mapping comes back fine, the wire could still be compromised and will not perform the way it's supposed to. The guy who would be considered my "boss" for the day (well, actually he's a partner in the biz so he's always my boss, I was just working "for" him today) says that what it signifies is that the wire, the copper, may have been stretched thin while being pulled through the conduit. I can understand this concept the way he described it.

What can you tell me about what's going on here? Is it also possible that the access point has gone bad? Would that cause the problem as well, the "it's only at 100" problem? I may be back early next week to the same place (among others with the same problem) to pull new wire. Should I bring a replacement wap and try that first? There are 6 - 7 that are having this same problem. I can't say that all wires were pulled with appropriate amounts of lubricant, or with lubricant at all...(I've learned well over the last year that lubricant can be your best friend, especially in tight pipes...heh...)

Thanks!

Read up on 10/100 and 10/100/1000

She wanted (or was expecting) 1000 Mbit i.e. gigabit. You gave her 100 Mbit.

Many reasons that could be the case: cable spec not rated for it, cable distance past limit, switch port isn’t 10/100/1000, improperly configured switch or AP, and on and on. And sure maybe someone used some wire stretchers while pulling it.

Check for a bent pin on the WAPs Ethernet port.

Also, walk the WAP over to the switch and see what link speed is achieved with a fresh patch cable.

Great point when troubleshooting in general, start at the feed/ beginning and work your way out to end point !!

Is stretching of cable really an issue? Maybe I've been lucky and conduits gave been lroperly sized....

I know it's theoretically an issue. I'm asking if you've ever seen problens proven to arise from stretched cable.

Grasshopper,

Here is a quick cable review.

The max pull tension spec of 25-34 lbf (depending on category cable type) is there for a reason, so I would bet someone has seen it make a difference. That’s equivalent to hanging just two gallons of paint on standard cat5/6 and you’re at the max pull tension. I’ve seen assholes in the field literally hanging their body weight trying to pull a cable...They had the pleasure of re-pulling a new cable.

Look at the bandwidth requirements for the various speeds. At higher speeds the physical characteristics of the cable are very important. You need a specific number of twists per foot and the overall diameter of the cable is important. If Godzilla hangs off the end of the cable, these cable characteristics will be distorted and the top speed capability will tank.

On occasion I'm brought in after the "smart" carpenter has had his apprentice pull all the networking. While the wire seems to go from here to there and the wire map is OK, there can be wire nut splices and areas that are stripped bare, with conductors shorting to pipes and metal studs. I've also seen daisy chain networking.

What wire certification tools would you recommend (ideally without breaking the bank) to verify cable integrity?

Presuming we are speaking for UTP cables:

Ideal SignalTek
[Link: idealnetworks.net]

Or

Fluke Cable IQ
[Link: flukenetworks.com]

Or

A laptop with a 1Gig NIC, cat6 patch cable, Router with a built in 2port Gig switch, cat6 patch cord, NAS with a 1Gig NIC
Transfer data files back and forth. Check for speed, connection type, check sum(did all the info make it?)

Basically what I knew about 10/100 and 10/100/1000 is that the latter is gigabit speeds and having a 10/100 piece of gear in your network can slow it down. I can almost guarantee that the other things you mentioned were not the case. Almost...well, the improper configuration at least. I know the distance, rating, and switch are good.


*smacks forehead* the switch was only 30 feet away, I should have done that. If I'm the guy that goes back, I definitely will. I'll look for a bent pin as well. It's not likely, but, from what I read after reading the responses here that makes sense as well.


Thanks, Buzz! I know enough to the point that at least half of that made sense to me. I'm sure I'll be referencing that in the future. Lemme ask ya...if the wap was only running 100mbps, what speeds do you think the access point could achieve?


Very good to know, man. Seems to me like I'm now the only person who has been pulling wire who knows this. I don't know how many footpounds pulling *however hard* is, but I'm sure we've broken that rule a number of times...

The only question I have right now is, as I asked Buzz...no, I'll add to it as well. operating at a 10/100 speed, what "speeds" do you think could be achieved? I know what bandwidth is for sure in the audio world, but when referring to speed and bandwidth in the networking world, is it the same? Basically, bandwidth would be the number of lanes on a highway. Cars would be the bits. If all of the cars are traveling in one lane, obviously that's not efficient, but if you have ten lanes then you can get ten times the bits to pass a certain point in the same period of time. Is this correct?

You guys are great. Thanks as always.

Grasshopper,

Here is an article discussing speed. I’m sorry that it is somewhat academic.

Discussing performance with customers is tricky because the tests that they will discover tend to count the number of bytes received, then multiply by eight and give that number as “speed”. This sort of report will always be lower than the advertised rate. Since there is overhead between the ISP and the customer’s router and additional overhead for the WiFi, I don’t loose any sleep if the reported speed is about 70% of the advertised rate. For, example, I’m typing on an iPad that is about 20 feet from the access point, there is an Internet Radio stream, a DOT, several Sonos units, and some thermostats also running over the WiFi. My reported speed is 45Mb for a nominal 60Mb advertised ISP speed. I have absolutely no complaints about this speed, but a customer might gasp that it should be 60.

I always discuss the reality of these tests with the customer before we run the test. Also keep in mind that WiFi speed reduces with distance and interference from other WiFi within range.

Another aspect of the overall impression of speed is the Ping time. “Ping” is the time required to turnaround a request to another network node. Long Ping times to remote sites will result in a sluggish feel to a website. The connection that I’m using has very low Ping times, often in single digits. I’ve used connections that advertise and test at much higher bit rates, but have longer Ping times. These connections don’t seem any faster than this connection.