Troubleshooting Ethernet Networks Flashcards
Troubleshooting Methodology
- Identify Problem
- Establish Theory of Probably Cause
- Test theory to determin cause
- Establish Plan of action to resolve problem and identify potential effects
- Implement solution or escalate as necessary
- Verify full system functionality, and implement preventive measures
- Document findings, actions and outcomes
Idenfity Problem (Gather Informatiom)
- Define scope of problem
- Check system documentation
- Check recent logs or records of recent work
- Check vendor support sites
- Identify symptoms
- Duplicate problem, if possible
Identify Promblem (Question Users)
- Open ended questions incite explainations in their own words: What happens when…
- Closed call for yes/no or fixed response: What does it say… what do you see…
- Key question: Did it ever work/when did it stop working?
- What changed since it last worked?
Approach Multiple Problems Individually
- You may discover symptoms of multiple problems
- Treat them individually unless they seem related
- Split off to separate support ticket if necessary
Establish Theory of Probable Cause
- Use multiple approaches to determine cause
- Question the Obivous: step through waht should happen and determine point of failure
- Methodically prove function of each component (can be more time consuming)
Top-bottom vs Bottom-Top Approach
- You can test at each level of OSI model in sequence
- If it seems harward related, starting with bottom to top
- Divide and conquer: trying to approach layer most likely causing issue.
Test Theory of Probable Cause
- Once enough information is gathered, test to theory or escalate
- Escalate when it is beyond your knowledge or falls in a system outside of your control
- Escalation could also be to outside parties, like vendors
- Provide as much information as possible to escalation point
Establish Plan of Action
- Action plan is steps to take to solve issue
- RepaiR: determine whether cost of repair/time to reconfigure something is best option
- Replace: Often this is more expensive or time consuming if part is not available. Is also a possible opportunity to upgrade device or softward.
- Ignore: Sometimes it is not worth it to fix issue if workaround is possible. Document and move on.
Implement Solution
- Take actions to solve issue based on plan of action
- May need to escalate.
- If change does not solve problem, you may need to try something else.
- Record and test after every change
Verify functionalist and implement preventative measures
- Confirm that solution fixes issue.
- Make sure that it didn’t cause any other issues.
- Try to eliminate source of problem, if possible.
Document findings, actions and outcomes
- Record information and step taken on ticket or other documentation system
Baud Rate
- Physical layer transmission of information over channel
- Called symbols
- Number of symbols that can be transmitted per second, measure in hertz
Ethernet Speed
- The expected performance of a link
- Meaured in bps
Throughput
- Average data transfer rate achieved over a period of time
- Included erros and other loses incurred at physical or data link laters
- Can be affected by link distance or interferece.
- Typically mesaured at network or transport layer
Latency
- Speed at which packets are delivered over network
- Measured in ms
- Not usually a problem with local ethernets
Attenuation
- Loss of signal strength
- Measured in decibels (db)
- Expresses ration of signal strength from origin to destination
Noise
- Anything that gets transmitted close or within channel that is not intended signal
- Expressed in signal to noise ratio (SNR)
Channel Link
- Entire cable path between hose and switch port:
- Network transceiver in the host (end system)
- Patch able between host and wall port.
- Structured cable between wall port and patch panel (The permanent link)
- Patch cable between patch panel port and switch port
- Network transceiver in switch port (intermediate system)
Investigating Link Failure
- AKA, complete loss of connectivity
- Check patch cords are properly terminated and connected to network ports
- Sub in good cable if necessary
- Then test receivers, using loopback tool
- If you can rule out bad cables and NICs/network ports, use tools to test structured cabling
Loopback Adapter
- Specially wired RJ-45 plug with 6” stub of cable
- Wiring pinout is 1 (Tx) to pin 3 (Rx) and pin 2 (Tx) to pin 6 (Rx)
- Packet sent by NIc is receive by itself
- Tests bad network cards and ports
LED Status Indicators
- Check on NIC
- May need vendor documentation to interpret LEDs
- On switch port typically LED states are:
Solid Green: Connected but no traffic
Flickering Green: Operating normally with traffic
No Light: Link is not working or disconnected at other end
Blinking Amber: Fault has been detected (duplex mismatch or spanning tree blocking, eg.)
Cable Tester
- reports information on physical and electrical properties of the cable
- Including Conditions, crosstalk, attenuation, noise resistance, et cetera
TDR
Time Domain Reflectometer
- May be function on cable tester
- Measure length of cable
- can locate open or short circuits, kinds or other problems.
- Can give position of cable fault
Multimeter
- If cable tester or certifier is not available, multimeter can be used to verify connectivity
- You will need to know expected results of testing electrical circuit
Wire Map Tester
- Incorporated on some multimeters or standalone device
- Can identify following issues:
Continuity: A conductor does not form a circle because of damage or error
Short: Two conductors are joined at some point
Incorrect Pinout or termination: Incorrect wiring withing cable like reversed pair or crossed pair
Tone Generator
- Used to trace a cable from one end to another
- Fox and House or tone probe
Decible Loss
(db) Loss
- Represented in ration of received voltage to original
- Express logarithmically:
+3 dB means double, -3dB means halving
+6 dB means quadrupling, +6 dB means quarterin
Crosstalk
- Usually indicates bad wiring, bad connector or improper termination
- Check cable for kinks twisting or crush points
- Also measured in dB, but higher values represent less noise
Types include:
Near End (NEXT): Crosstalk on receive pairs at transmiter end, usually caused by excessive untwisting
Attenuation to Crosstalk Ratio, Near End (ACRN): Difference between insertion loss and NEXT
Attenuation to Crosstalk Ration, Far End(ACRF): Measured on receive pairs at recipient end.
Power Sum:
Straigh Though Cable
- Terminated with either T568A or B at both ends
- Used for uplink ( MDI port to MDIX port)
Crossover Cable
- Terminated in T568A at one end and B at other
- Used to connect host to host or hub to hub
- Typically no longer necessary because of auto-MDI-MDI-X
- All Gigabit Ethernet ports support this
Rollover/Console Cable
- Used to connect PC or laptop to command line terminal or switch or router
- Standar RJ-45, but wired differently than ethernet
- Modern cables use USB connector and port
- Typically paleblue
POE
- Cat 3 or better is required for POE
- Cat 5e or better is needed for POE+
- Can cause heating issues
- Shielded cabler is better at dispersing heat
OTDR
Optical time Domain Reflectometer
- Sends light pulses down fiber and times reflections from break
OSA
Optical Spectrum Analyzer
- used typically with WDM to ensure each channel has sufficient power
- At long distanced attenuation on wavelengths can vary
- Determines whether existing cable is suitable for reuse
Dirty Optical Cables
- Dirt, dust, or grease in transmission path will reduce signal strength or block entirely
- Most commonly on connectors
- Connectors should be covered with dust cap when removed
- Should be cleaned using solvent designed for fiber
- Can also occur during splace
- Ensure all equipment is properly cleanded
Incorrect transceivers
- Each optical interface is desnged with a specific transceiver
- Must check specifications make sure correct fiber type is used
- Including patch cords