Troubleshooting is an art that comes with experience, and to become a master in it, you are required to practice things practically on your own.

There are various error recovery features that are provided by the TCP protocol that help us to recover from loss of packets that might happen commonly in a production environment.

TCP retransmission and duplicate ACKs are some of those techniques that are used by the TCP protocol in order to make the life of network administrators a bit more comfortable.

Slow network is one of those common problems that you have to face on a daily basis. Before you start solving these latency issues, you should know the basic methodology that you can follow, that is, to categorize your scenario in one of the latency categories: a wire, client, or server.

Solving bottleneck issues, such as packets getting queued up inside the sender buffer area and causing trouble, is quite important. The best approach in solving a bottleneck issue would be to take the help of IO graphs that you learned about in the earlier chapter to visualize a situation and get hold over it.

Applications use protocols such as HTTP and DNS. This is very common, but you must be aware of error codes these can present, and without using Wireshark, you should be able to identify the situation. I do not know every error code, even I can not do that. But the most common ones that you might witness.

Creating a baseline is one of the most convenient ways of dealing with issues in your network. When you have a trace file containing an optimized traffic pattern, then, by comparing the normal pattern with the deviated pattern, you can solve the issue in less time with few resources. Collect the network traces for your baseline from various locations in your network at least 2-3 times.