CompTIA Network+ N10-008 – Module 17 – Troubleshooting Networks Part 4

6. 17.5 Environmental Metrics and Sensors

If we have an equipment room, such as a data center, the equipment in that room can be very sensitive to fluctuations in environmental conditions. These might include temperature, humidity, power fluctuations, or a power outage, or maybe water if we have flooding. And in this short video, I just want you to be aware that there are different sensors that can alert us if we have one of these fluctuations in an environmental condition.

And some sensors connect to the network, and they may have an SNMP MIB. That’s a management information base that can alert an SNMP manager if a threshold is exceeded. For example, let’s say we had a threshold of 85 degrees Fahrenheit for an equipment room. If we went above that, if we had 86 degrees, then that could fire off an SNMP trap to an SNMP manager and alert somebody that there might be something going on with the HVAC system. And here we have a few examples of what those sensors might look like.

First, a temperature and humidity sensor. This can alert us if the humidity in a room becomes too great. And you should check your equipment’s documentation to determine what is an acceptable humidity range, what is an acceptable temperature range if there is a power outage in the room, hopefully we have some sort of a power backup system. This might include an uninterruptible power supply or a Ups maybe. In addition to this, we have a generator. Oftentimes we’ll have a Ups that will keep the equipment up and going until the generator has time to kick in.

That’s what we did at a university where I used to work. It would take about 45 seconds for the generator to get up and going if there were a power outage. And the Ups will keep the equipment running until the generator kicked on. And then as long as you have gasoline in the generator, you should be good in terms of power. And some of these Ups devices, they do have an SNMP trap, like we were talking about, that can alert us if there is a power outage.

We might also want to monitor any potential flooding conditions. A lot of data centers are in basements, and those might be subject to more flooding, as opposed to having our data center on a higher floor. So we might have these water level sensors which can tell us not only, hey, there’s water in the floor, somebody needs to go check that out. Some of these can even tell us the depth of the water. And that’s a quick look at some different environmental metrics we want to consider, and some sensors that can alert us if those metrics get into a really bad range.

7. 17.6 Common LAN Issues

In this video, we want to take a look at some common issues that might occur on a local area network. What are some troubleshooting targets we might have? One is attenuation that’s where the signal strength degrades over time, maybe by having a cable that is too long. By the way, that signal strength, it’s measured in decibels or in DB. We might also have too much latency from point A to point B on the network. Some latency sensitive applications like Voice and video, if they have too much delay, it can really degrade the quality. For example, there’s a recommendation for voice traffic over a network. Ideally, according to the G 114 recommendation, one way audio from the time I speak it to the time you hear it, that should not exceed 150 milliseconds. Something else that voice is very susceptible to is Jitter. That’s the variation in inter packet arrival times.

Let’s say that we receive a voice over IP packet, and behind that we’ve got a big massive data packet, and then we receive another voice over IP packet. The amount of time between the reception of those two voiceover IP packets, that’s called jitter. And for Voice, that should not exceed 30 milliseconds. Ideally, sometimes we have cables running too closely to one another, and that could cause crosstalk, because when current flows through a wire, it creates an electric field around that wire. And if there’s a nearby wire, that electric field could be induced onto that other wire. That’s the reason we have twisted pair cabling. The twists help prevent that.

And another reason we twist that wire is to protect ourselves from EMI electromagnetic interference. Maybe there’s a power generator and it kicks on and it creates an electrical spike that could be picked up on our network cabling. Or maybe there’s a nearby microwave. There are several different things that could cause EMI. That’s one of the benefits of fiber optics. Because it’s optical cabling, it’s not subject to electromagnetic interference. And with copper cables, we may have a break in that cable, and that would be called an open. However, if two conductors, which should be separated if they’re somehow connected because somebody layed a sharp metal object on top of a cable, that would be called a short. And that term short is often used when the condition is really an open. And open means current is not flowing, and a short means current is flowing where it should not be flowing. Sometimes we use a cable that doesn’t have the correct pin out.

The wrong pins are being used for transmission and reception. Or we might be using the incorrect cable type. Maybe we have a T one cable, for example, that we’re trying to use to interconnect a couple of Ethernet switches together. That’s not going to work. And on those Ethernet switches, as well as routers, we might have a bad port. And that’s an easy troubleshooting step we can take if we have a client connected to a switch and they’re not communicating on the network, we might just take their cable and move it to a different port or maybe move it to a different switch. And some of those Ethernet switches can accept different transceivers.

That’s going to give us another way to connect into the switch. It might be a twisted pair transceiver, it might be a fiber optic transceiver, and those transceivers can run at different speeds. But if the transceiver at each end of a cable don’t agree on the parameters to use, we might have a transceiver mismatch. And a specific case of having that incorrect pin out that we talked about is when we have our transmit and receive leads reversed, maybe we’re connecting a couple of PCs together. Well, they’re going to use the same pins for transmit and the same pins for receive. So we need a crossover cable. If we’re going to connect those like devices together and the devices on the ends of an Ethernet cable, they should agree on duplex and speed. Duplex means can we send and receive simultaneously? If so, that’s full duplex. Or can we only send or receive at any one time, but not both at the same time. That’s an example of half duplex.

And if you have a switch port, for example, that’s configured for full duplex, but maybe your PC is configured for half duplex, that could cause an issue. And it can be difficult to troubleshoot because some data is getting through. Things just appear to be really slow. In fact, that’s one of the first things I would check when dealing with a slow connection. I would check to make sure that the duplex on each end of a cable agreed. Now the speed it has to match or we’re not going to be able to have any data flowing through at all. The cables could be damaged. I dealt with the situation one time where a cable was running underneath one of those chair mats that have the plastic spikes that go down into the carpet. And some of those plastic spikes had actually penetrated and damaged the cable. And the person working at the desk, they had had their chair roll back and forth over that cable so many times underneath the chairman that it was very flat, it was very damaged.

But the challenge was it worked part of the time and that can make things a bit difficult to troubleshoot. Sometimes we plug in a connector, maybe like a DB nine connector, and we think everything is secured, everything looks configured correctly, but we may have one or more pins in that connector that are bent, keeping all the connections from being made. There may be a portion of the network that makes us experience a bottleneck. Maybe. I’m running at a gig speed on my land and I’m going to a remote office that also uses a gig speed.

But I’m having to go over a ten meg wide area network connection that’s going to slow things down. That’s going to be a potential congestion point, also known as a bottleneck. We might have mismatched VLANs where one end of a cable is plugged into a switch port belonging to one VLAN, but the other end of the cable is plugged into another switch whose port belongs to a different VLAN.

They need to match up. We could have issues and something that we can check when we’re doing troubleshooting of connections. Going into a computer’s network interface card or a switch port is sometimes those NYCs, those network interface cards or the switch ports, they’re going to have Led status indicators to indicate things such as speed and duplex and whether or not we have any activity on that link.

8. 17.7 Common Wireless Network Issues

In this video, we want to consider some troubleshooting issues we might have in a wireless network. First up is reflection. Remember that data flows through a wireless network using radio waves. And if those radio waves hit a metal object like a filing cabinet, they could reflect in an unpredictable direction, giving us an unpredictable coverage area. Similar to reflection is refraction. That’s the bending of a radio wave, possibly because we went through a different type of medium. Maybe we’re going through drywall. Or maybe we have an access point next to a big water cooler. As radio waves flow through that water, they could become refracted. And as we flow through any medium other than a vacuum, the radio waves over time are going to be absorbed by whatever medium they’re flowing through, which leads to a degradation in signal strength called attenuation.

And the measure of the energy transmitted from an axis point’s antenna is called an EIRP. That stands for effective Isotropic radiated Power. That’s the measure of how much power is sent by the access point’s antenna. And this is going to combine the transmit power of the access point itself. Then we subtract the loss of the signal as it goes through the cable from the access point to the antenna, and then we add on the antenna gain. But that’s a term I’d like you to know. EIRP effective isotropic irradiated power. And speaking of antennas, we might have an incorrect antenna type. Maybe we need a directional antenna to go between a couple of buildings.

And we’re using an omnidirectional antenna, which gives us approximately the same radio signal strength going out in all directions, or vice versa. Maybe we put an access point in the middle of an office, intending to cover that entire office space, but it was some sort of a patch antenna, or maybe a yogi antenna, something that’s directional, and it’s not going to give us the coverage we wanted. And because we are using radio waves, we’re subject to interference by other things that are using radio waves, such as a microwave. Or maybe you’re nearby an airport, and you might have interference from radar. And you also need to be careful in placing your antennas to give you the coverage area that you want. And from a security perspective, you probably don’t want to place an antenna such that you have a large coverage area out in your company’s parking lot.

You may not want people sitting in their car trying to get on your network. And you may want antennas to give you complete coverage area across an office floor. But when you’re doing that, you want to make sure that any overlapping coverage areas are using different channels. If I have two antennas in close proximity with one another, and maybe on the 2. 4 gig band, they’re both using channel six, that could be an issue. They could interfere with one another due to that channel overlap. And sometimes we’ll have a wireless client boot up and it does not seem to get IP address information. That could be because it takes too long to associate with the access point. The PC boots up, it sends out the DHCP Discover broadcast, but it doesn’t go anywhere because we’ve not yet associated with the access point. We may have an issue with a captive portal.

You may have seen these in a hotel, for example. You get on their WiFi and you’re sent to this Captive portal page where maybe you have to enter some information like your room number or your name. And maybe you also have to agree to an acceptable use policy and after you’ve done that you’re supposed to be able to go out to the Internet. Well, sometimes you may have issues with that. I’ve seen an issue where I cannot pull up the captive portal and a common reason that happens is the device we’re using has DNS hard coded rather than learning that from a DHCP server. Or we may have a client that does associate with an access point, but then it disassociates. Some common reasons for disassociating from an access point include timeouts, such as an idle timeout or a session timeout. Maybe somebody makes a change to the wireless LAN that could disassociate a client. Maybe the admin working on the wireless LAN controller manually does a disassociation.

From a security perspective, the person that disassociates may be the victim of an attack. An attack that is sometimes launched is for a malicious user to send management frames to an access point, telling it to disassociate from a client. And when that client tries to reassociate the attacker, maybe they have an access point with the same name, in other words, of the same SSID. And when that client tries to reassociate, it might associate with the attackers access point, allowing the attacker to eavesdrop in on their traffic. You also want to know how many clients that an access point can comfortably handle. You don’t want to have an overcapacity situation where there are simply too many clients associated with that access point. Some access points are designed to be high density, but others are not. Make sure you know what the capacity is of your access point and the access point range can oftentimes be influenced by adjusting the radial power.

But you may be in a situation where the radio power is simply not strong enough to give you the coverage area that you want and you run into distance limitations. Sometimes a user will try to connect to a network and the SSID is not visible from a list where they can simply select it and they have to type in the SSID, the Service Set Identifier, which is the name of the wireless network. They may type it in wrong, they may have the wrong SSID, or when they’re trying to authenticate, they may have the wrong password or passphrase. Sometimes the security type being used between the client and the access point may differ. Maybe the access point requires AES encryption while the client wants to do TKIP. That could prevent the association from happening.

And we may not have the power levels set correctly on the access points. If the power levels are too high, then we might have some channel overlap. If they’re too low, then we might not be getting the coverage area we thought we were. And also, let’s consider signal to noise ratio. We don’t want our signal strength to be cluttered with background noise. There’s going to be some background radio frequency noise. That’s called the noise floor. And let’s say that it is at negative 90 DB, and our signal strength is negative 65 DB. Well, the signal to noise ratio is the difference. And the difference between 90 and 65, that’s 25. So we’ve got a signal to noise ratio in that case of 25 DB. The bigger the signal to noise ratio you have, the better. And that’s a look at some common troubleshooting target targets for wireless networks.

• Category: Uncategorized