GNS3: Graphical Network Simulator - Tell Me More, Please!
(Page 3 of 4 )
Now that we understand what the deal with CISCO IOS images is, it should also be said that the GNS3 always emulates the selected IOS images with high accuracy, and that if you struggle to do something, or some commands are lacking, or anything else for that matter, it’s all because of the IOS image. Those operating system images are varied and of all kinds; some lack few commands, while others work differently.
Moreover, it is almost guaranteed that the way an IOS image behaves in GNS3 is exactly the way an actual CISCO device having that same image would work in the real world. The emulation always works effectively and, thus, we are able to claim that the simulated networks are almost 100 percent identical to a real world situation. This is why GNS3 is so powerful—thanks to its reliability.
Anybody that is already experienced with virtual environment emulating operating systems such as Wine, VirtualPC, VMWare, and the like knows that there is a somewhat high performance hit compared to the real OS. This is understandable, because in terms of performance, a totally perfect emulator does not exist. The same happens in the case of GNS3. The networking devices in the real world can perform up to 100x times faster.
In the documentation of GNS3, it is stressed that this utility suite shouldn’t be used as a replacement for an actual router or other network device. Other usages than purely for testing or experimenting would be improper. For testing and simulating experimental networks and topologies, it is truly phenomenal.
It’s time to present the features and specifications of GNS3. First of all, check out the table below of the supported platforms.
We are going to skip the steps for creating a new project, adding the IOS image, and all that. This process is explained in great detail in the official documentation. The graphical environment is split in a three-pane fashion. The left pane contains the network devices which you can select to create the topology. The empty design pane is in the middle. You can place the devices there. On the right pane the topology summary (if enabled) can be found. On the bottom the Dynagen console is shown.
From the left pane you can select devices such as routers: c1700, c2600, c2691, c3600, c3700, c7200, etc.; PIX firewalls, Ethernet switches, ATM bridges and ATM switches, Frame Relay switches, Layer 2 switches, servers, cloud, etc. Surely, these depend on the chosen IOS and additional components can also be added. Once placed on the design pane, you can right click on the device to select the option to configure.
Creating multi-links, adding wires, zooming in/out, and all these actions can be taken for granted. Once you have created your envisioned topology and you’ve finished configuring the devices, it’s time to start them. After that, you can telnet (by default the Windows or Linux telnet client, but you can use PuTTy or the client of your choice) to the devices for a more advanced individual configuration.
The official GNS3 tutorial contains a walk through on how to find and set your idlepc value. This is important, because by doing so you can seriously reduce the CPU load usage during emulation. By default, as mentioned earlier, 100 percent is being used. This action puts the IOS into a state of sleep when it is not used.
Don’t neglect the above warning note. As soon as you launch GNS3 and create the simplest topology, find and set your idlepc value; after that point your PC becomes responsible once again for continuing your experimenting journey (this is critical in the case of slow computers). There are various video tutorials also available to aid you, if need be.
The real mind-blowing advantage of GNS3 is that you are able to configure your simulated network topology to communicate with your already existing “realworld” network as well. This way you can create connections with your routers. This is the part that helps network and system administrators a great deal. They can try out and experiment with various topologies without actually hooking everything up with cables. And if a simulated topology looks great, then they can take that for granted in reality, too.
Usually network simulators do not allow the kind of communication that GNS3 does. All in all, we could say that GNS3 is an amazing application because it makes good use of Dynamips, Dynagen and Pemu by creating an easy-to-use visual graphical environment (it was written in python in Qt library—just like KDE for Linux).
Next: Final Thoughts >>
More Software Articles
More By Barzan "Tony" Antal
