Problems on the networks. MPLS TE network issues
As a technical support engineer for a vendor and author of the “Network for the smallest” cycle and a podcast for telecom operators, I am often asked what I have to do at work and how my experience accumulates.
So, I deal with problem solving.
I would like to propose a new format for articles to networkers — puzzles. I describe the symptoms, provide the data you ask, you are looking for a solution.
So let's start with a simple one: we have this MEN network:
')
It runs MPLS TE. R1 is a central device providing connectivity to the Internet and base station controllers. Everything is organized through L2VPN. Up to 2G BS, of course, carried TDM.
From time to time a strange thing happens on the network: base stations lose synchronization with controllers. On a fixed network, losses begin, and the speed drops significantly.
All this relates to the access segment connected to R4.
It was empirically established that the problem is solved by turning off and turning on the interface R3 <-> R6.
Frequency - several times a day.
The main work and our freshly acquired hobby for solving other people's problems sometimes throws up such and not only such interesting tasks (Caution, spoiler) for troubleshooting.
PS Some logs and pictures I have left, so I will explain in words.
The task is to solve the problem, not to put forward an assumption. ;)
So, I deal with problem solving.
I would like to propose a new format for articles to networkers — puzzles. I describe the symptoms, provide the data you ask, you are looking for a solution.
So let's start with a simple one: we have this MEN network:
')
It runs MPLS TE. R1 is a central device providing connectivity to the Internet and base station controllers. Everything is organized through L2VPN. Up to 2G BS, of course, carried TDM.
From time to time a strange thing happens on the network: base stations lose synchronization with controllers. On a fixed network, losses begin, and the speed drops significantly.
All this relates to the access segment connected to R4.
It was empirically established that the problem is solved by turning off and turning on the interface R3 <-> R6.
Frequency - several times a day.
The first very talking tip
Graphs of traffic behavior on the day when the problem occurred, on link R2 <-> R4:
The second tip is an additional symptom.
After the fall of the interface between R2 and R4, the services do not crash and work for some time quite successfully, without loss. But after about a minute, services begin to suffer, until R3 <-> R6 is pulled.
The main work and our freshly acquired hobby for solving other people's problems sometimes throws up such and not only such interesting tasks (Caution, spoiler) for troubleshooting.
PS Some logs and pictures I have left, so I will explain in words.
The task is to solve the problem, not to put forward an assumption. ;)
Source: https://habr.com/ru/post/190982/
All Articles