No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Converged Networks: Finding Trouble

The push to converge can sometimes offer challenges to those troubleshooting and maintaining the network and infrastructure. Convergence should be something planned, and for those of you familiar with baking layer cakes, you should fit right in. Whatever methodology or tools you use and muster to nip problems, start with simple and work your way up.

Endpoints can eat away time in determining and solving issues, and with port statistics, chasing rabbits down those proverbial holes can be minimized. In the screen shot below, an end user HP network jack (formerly 3Com network jacks) hasn't stopped working, but the user complains of slowness in connectivity and less than ideal performance while using her workstation. Replacing the network jack resolved the problem.

So the network jack issue is resolved and the user is happy. But is the true root cause of the problem resolved? Of course you want it to be, but the question remains, why did the network jack fail? Networks are powered, and local power isn't friendly. Does the workstation have any local power protection devices?

In this situation, the customer installed AC-power SPDs (Surge Protection Devices, formerly known as TVSS: Transient Voltage Surge Suppression). What wasn't installed until recently was power conditioning in the closet hosting the switch stack and other gear. When customers do decide to act on recommendations of adding anything that attempts to minimize or prevent power disruption, they often think they automatically have a clean bill of health. This isn't the case, since the new power conditioning only protects the gear at the time of installation and moving forward, and cannot guarantee against prior disruptive power events that can cause future or slow-growing equipment failures.

In another issue, a customer's WiFi network is disrupted with loss of connectivity, and an access point appears to go out of service. The screen shot below captures the power consumption of the PoE devices and displays the type of PoE: IEEE Class 1, 2, 3 or 4. Note port 21 is a Class 3 PoE device and is using only 3.136 Watts. A quick call to the manufacturer revealed that these particular access points operate between 6-7 Watts. The access point was replaced and the WiFi connectivity issues resolved.

Now, with a new access point in place and the customer happy that WiFi is restored, did we fix the issue? Of course we did, but if you dig a little deeper you find out that only one section of the building was experiencing the "WiFi outages," and the users were all doctors who were prevented from logging in and out of EMR (Electronic Medical Records) used to see their patients. So then another question should arise as to why they didn't have coverage from another access point.

After hours we set up to test at the same work area used by the doctors to simulate the WiFi outages they were experiencing. The distance between access points was only 85 feet. We took the access point covering their work area out of service, and noted that we could still connect in the reception areas, but after walking another 20-40 feet into other parts of the suite (specifically, the exam area), our signal dropped to weak, then low, and then no connectivity. Next, we determined a new location for adding an access point that would cover the exam areas of the suite if the primary access point failed. The new access point is less than 60 feet away and is central to both sides of the octagon-shaped building.

While we replaced a defective AP, improving the WiFi coverage to prevent future issues was the more important task.

These are just two of many issues with endpoints, and instead of focusing only on restoring service, you may need to redirect your attention toward also improving service to prevent future outages or failures.

In getting back to baking layer cakes, converging infrastructures require more than just the finality of delivering icing on the cake. Both of these issues could have different outcomes based on the information available. When a failure occurs, addressing any possible improvements is an opportunity.