Free virtual networking software for linux

Virtual Ring Buffer (VRB) is an implementation of a character FIFO ring buffer. Virtual Ring Buffer provides direct access to the buffer so the calling program can construct output data in place, or parse input data in place, without the extra step of copying data to or from a calling program provided buffer area.
In order to facilitate this direct access, VRB makes sure that all references to buffer locations for either output or input is always a single contiguous block of memory so that the calling program does not have to deal with split buffer spaces every time the cycling of data reaches the end of the buffer and wraps around to the beginning.
Thus, the calling program is free to use any number of tools available which do not have to be aware that they are operating directly in a ring buffer. In the latest version, VRB also provides a buffer overflow protection mechanism so that buffer overflow bugs force the program to be trapped by the operating system if accessed pointers wander beyond the range of address space managed by VRB. This feature can be disabled if desired, since it occupies 2 extra pages of address space.
VRB does all this without adding any copying overhead of its own. This ability is achieved by placing a mirror image memory map immediately after the buffer that is a reference to exactly the same memory as the buffer itself. Thus a reference to the memory location immediately past the last location in the buffer accesses the same physical RAM as the beginning of the buffer.
Thus a pointer near the end of the buffer that has a length that would otherwise have to wrap around will instead extend into the mirror image space to get the same result, but without any special handling by the program.
Installation:
First extract the source where you wish to store it. One of these commands might be used within the desired directory after the tar files has been downloaded there:
bunzip2 < vrb-0.5.0.tar.bz2 | tar -xpf -
gunzip < vrb-0.5.0.tar.gz | tar -xpf -
A subdirectory called vrb-0.5.0 is created. Execute these command to change into that directory and compile the source (changing the prefix if desired):
cd vrb-0.5.0
./Configure --prefix=/usr/local
make clean
make install
Note that unlike most other source packages using a configuration system, the configure command name is capitalized. The above commands will not install the man pages at this time. To install the man pages, simply copy them to the desired location, such as:
cp -p vrb/man/man3/*.3 /usr/local/man/man3
Usage:
VRBs basic approach designed into its API is that the calling program asks for the address and length of either the empty space where output data can be placed, or the data space where input data is present (previously placed there). These steps are implemented by macros for maximum speed, and are separate for address and length. The calling program then accesses that data directly in the buffer via that address (and length).
Once the data handling is complete, the calling program decides how much of that data it has inserted into empty space, or taken from data space, and calls VRB functions to indicate the which, which updates pointers (but does not copy any data). There are also functions available to assist in doing read/write I/O. Once the man pages are installed, the command man vrb is the starting point.
Enhancements:
- A vrb_empty function was added.
- Dead code in vrb_take was cleaned up.
- A poll loop was implemented in vbuf. vbuf now uses getopt.
- Buffer size calculation and data type/format were fixed in vbuf.

Virtual server monitor (vsmon) is a monitoring tool for Linux-VServer that gives a system administrator a global view of all vservers running on his park.
For now, it does not offer control over the vservers life cycle (e.g. provisioning, start, stop, and restart operations). To use vsmon, you must deploy one backend on each vserver host and one frontend, preferably in its own vserver.
Enhancements:
- The frontend was reworked to display a summary of all hosts.
- An owner can be assigned to each vserver.
- The backend can be used on hosts without Linux-VServers, and it no longer obtains IP addresses.
- The libdnet dependency was removed.
- The parsing of /proc/"pid"/stat now works for kernel 2.6.18.
- A Nagios plugin was added, which enables the use of the vsmon backend to monitor a host disk, memory, and swap.
- The plugin monitors all mount points according to an "intelligent" algorithm, which determines warning and critical thresolds as a function of the total size of the partition.

Virtual Console init is an /sbin/init clone for virtual consoles. It is very easy to set up and use because it has no configuration files.
It is very small and has builtin getty, login, klogd, and slogd. Virtual Console init is suitable for embedded systems and home Linux boxes.
This is a new kind of "init" that only works for the virtual console, so dont try it to login from a serial line. The idea is having a very simple init for home Linux systems (although it could work on other systems as well, who knows...), allowing for a simpler and cleaner configuration. It is useful too in embedded systems or tiny boot disks (maybe as a complement for "busybox"). In other words: it works for me, I use it and Im releasing it in the hope of it being useful for others.
The binary itself doesnt accept parameters, so lets go directly to the behaviour... It must be run by "root" and have PID 1, so dont try to run it yourself, it wont work... You should install it under "/sbin/init", and the kernel will run it automatically at boot.
Take into account that this "init" has both the "getty" and "login" parts built-in, not as separate binaries. This means that if you have a server or another binary which needs "/bin/login", it probably WONT WORK. The sysklog system (kernel logger and system logger, "klogd" and "slogd") are built-in, too. You dont need sysklogd or equivalent, unless you want some feature of them... In that case, you can disable the builtin system.
Enhancements:
- vcinit is now licensed under the GPLv2 only, and not necessarily any later version of the GPL.

KNetworkManager is the KDE front end for NetworkManager. KNetworkManager project provides a sophisticated and intuitive user interface which enables users easily to switch their network environment.

The range of functions encompasses the features implemented by NetworkManager daemon. Up until now NetworkManager supports:

Wired Ethernet Devices (IEEE 802.3)
Wireless Ethernet Devices (IEEE 802.11): Unencrypted, WEP, WPA Personal, WPA Enterprise
Virtual Private Network (VPN): OpenVPN, VPNC
Dial-Up (PPP)

How does it work?

For both, Wireless LAN and Wired LAN, NetworkManager supports devices known to HAL. Unless working in offline mode, NetworkManager tries to keep the system connected at any time. For this, NetworkManager follows the following policy.

Once started, NetworkManager asks HAL about available network interfaces. If a wired network interface with a carrier is found, NetworkManager connects to this. Either by DHCP (default) or by setting up previously defined static configuration. Later on, when KNetworkManager starts up, NetworkManager will expose its information about network devices and wireless networks found by scanning to the applet.

At this point, if a user decides to unplug the wired connection, NetworkManager will not connect to an arbitrary wireless network. By default all available networks are untrusted. Only if a user decides to connect to a network manually once, this specific network n henceforth being marked as trusted.

n = { ESSID, Hardware address or addresses of the access point}

The trusted networks are stored individually for each user. KNetworkManager stores them and informs NetworkManager about the known, trusted networks. If a wired connection drops due to unplugging the cable NetworkManager will automatically connect to a trusted network. As this step was done without user interaction, NetworkManager will reconnect to the wired connection once it has a carrier.

If a user joins a wireless network by manual intervention (e.g. user clicks on a network listed in the context menu of KNetworkManager or connects to a hidden wireless network), NetworkManager will take down any previous connection. If one unplugs the wired connection after such a step, NetworkManager will not connect to the wired connection once it has a carrier.

The "Virtual Universe" is a 3D cyberspace which offers more possibilities than just chat: it is a combination of the Web, chat, and instant messaging within a realistic, three-dimensional cyberspace.
The "Virtual Universe" is a virtual reality environment which runs on top of the Internet.
Main features:
- chat with each other
- see the chat-partner as a virtual, three-dimensional person
- interact and communicate
- be creative in many ways
- create own worlds for any desired subject
Enhancements:
- A lighting bug in .COB/.SCN loader has been fixed.

Virtual Machine Manager (virt-manager for short package name) is a desktop application for managing virtual machines. It presents a summary view of running domains and their live performance & resource utilization statistics.

A detailed view presents graphs showing performance & utilization over time. Ultimately it will allow creation of new domains, and configuration & adjustment of a domains resource allocation & virtual hardware. Finally an embedded VNC client viewer presents a full graphical console to the guest domain.

The application logic is written in Python, while the UI is constructed with Glade and GTK+, based on mockups provided by UI interaction designers. The libvirt Python bindings are used to interacting with the underlying hypervisor.

This enables the application to be written independant of any particular hypervisor technology, although Xen is the current primary platform. When libvirt is ported to additional hypervisors minimal effort will be required to update the management UI.

The Linux Virtual Server is a highly scalable and highly available server built on a cluster of real servers, with the load balancer running on the Linux operating system.

The architecture of the server cluster is fully transparent to end users, and the users interact as if it were a single high-performance virtual server.

The basic goal of the Linux Virtual Server Project is to:
- Build a high-performance and highly available server for Linux using clustering technology, which provides good scalability, reliability and serviceability.