Linux packet mmap -- useful tool to squeeze performance

Linux packet mmap

In order to improve efficiency of packet raw transmission in Linux kernel.

useful Ubuntu kernel binary

on this link:
http://kernel.ubuntu.com/~kernel-ppa/mainline/

VirtualBox and serial consoles

1. Enable serial ports for your virtual machine
2. Select “Host Pipe”
3. Enter /tmp/vboxconsole as the filename
4. Use netcat to read the console: nc -U /tmp/vboxconsole

Install/Compile Linux Kernel 3.0 In Ubuntu

Install/Compile Linux Kernel 3.0 In Ubuntu

User space memory access from the Linux kernel

Although the byte may be the lowest addressable unit of memory within Linux®, it's the page that serves as the managed abstraction of memory. This article begins with a discussion of memory management within Linux, and then explores the methods for manipulation of user address space from the kernel.

Efficient data transfer through zero copy

Many Web applications serve a significant amount of static content, which amounts to reading data off of a disk and writing the exact same data back to the response socket. This activity might appear to require relatively little CPU activity, but it's somewhat inefficient: the kernel reads the data off of disk and pushes it across the kernel-user boundary to the application, and then the application pushes it back across the kernel-user boundary to be written out to the socket. In effect, the application serves as an inefficient intermediary that gets the data from the disk file to the socket.

Enhanced Export Chart Procedure

In Export Chart as Image File I described the VBA command to export a chart as an image file, and I presented a simple routine that exported the active chart...

Create a Bootable Ubuntu USB Flash Drive the Easy Way

We've already covered how to use an Ubuntu Live CD to backup files from your dead Windows computer, but using the boot cd can sometimes be a little slow. We can speed up the booting process by installing Ubuntu to a bootable USB flash drive instead.

Changing Priority on Linux Processes

What is Priority and Why Should I Care? When talking about processes priority is all about managing processor time. The Processor or CPU is like a human juggling multiple tasks at the same time. Sometimes we can have enough room to take on multiple projects. Sometimes we can only focus on one thing at a [...]

0x8086: Understanding TCP behavior in Linux 2.6

0x8086: Understanding TCP behavior in Linux 2.6: Lately I have been doing a lot of TCP performance analysis on different configuration settings. TCP is a very complex protocol and have plen...

Using and mounting a RAM disk in Ubuntu – The easy way

Using and mounting a RAM disk in Ubuntu – The easy way

# 2GB ramdisk
mkdir -p /ramdisk
mount -t tmpfs -o size=2048M tmpfs /ramdisk

5 Useful Unix DD Command Examples

5 Useful Unix DD Command Examples

Restrict DHCP response on Cisco 3560 layer 3 switch

DHCP snooping is a DHCP security feature that provides network security by filtering untrusted DHCP messages and by building and maintaining a DHCP snooping binding table.

DHCP snooping acts like a firewall between untrusted hosts and DHCP servers. It also gives you a way to differentiate between untrusted interfaces connected to the end user and trusted interfaces connected to the DHCP server or another switch. For DHCP snooping to function properly, all DHCP servers must be connected to the switch through trusted interfaces.

When a switch receives a packet on an untrusted interface and the interface belongs to a VLAN in which DHCP snooping is enabled, it verifies that the source MAC address and the DHCP client hardware address match (the default). If the addresses match, the switch forwards the packet. If the addresses do not match, the switch drops the packet.

DHCP snooping considers DHCP messages that originate from any user-facing port that is not a DHCP server port as untrusted. From a DHCP snooping perspective, these untrusted user-facing ports must not send DHCP server type responses, such as DHCPOFFER, DHCPACK, or DHCPNAK.

My story comes that there is a local network cluster connected to the department network. The department network has a DHCP server A which provides DHCP services for all ports that forwards DHCP requests to it. However, my local network cluster has a local DHCP server B too, which provides a local DHCP service for local nodes only. Therefore, a conflict happens when the local cluster node requests DHCP service. This request may get response from DHCP server A or get response from DHCP server B, and it depends on the fact that which is faster to take the necessary action and ring the bell. This conflict is totally caused by the nature of DHCP request broadcast that goes everywhere among the network.

My solution is to block the DHCP packets (DHCP response packets) from the un-necessary department DHCP server A. A sample of such commands on a Cisco 3560 layer 3 switch comes as follows:
	

cisco3560#configure terminal 
Enter configuration commands, one per line.  End with CNTL/Z.

cisco3560(config)#ip dhcp snooping 
cisco3560(config)#ip dhcp snooping vlan 99	/* my local network vlan number */
cisco3560(config)#ip dhcp snooping information option

cisco3560(config)#interface gigabitEthernet0/48	/* my local DHCP server B access */
cisco3560(config-if)#ip dhcp snooping trust 
cisco3560(config-if)#exit 
cisco3560(config)#ip dhcp snooping verify mac-address 
cisco3560(config)#end

cisco3560#configure terminal 
Enter configuration commands, one per line.  End with CNTL/Z.
cisco3560(config)#interface gigabitEthernet0/47	/* department DHCP server A access */
cisco3560(config-if)#no ip dhcp snooping trust
cisco3560(config-if)#end
cisco3560#copy running-config startup-config 


This example shows how to display the DHCP snooping configuration for a switch.
cisco3560# show ip dhcp snooping

This example shows how to display the DHCP snooping binding entries for a switch.
cisco3560# show ip dhcp snooping binding
	

Reference:
Configuring DHCP Features and IP Source Guard
How to configure a Cisco Layer 3 switch-InterVLAN Routing

A brief history of TCP and its variance

Two main congestion control variations are those offered by TCP Tahoe and Reno. (since not long before, 2005)


The two algorithms were retrospectively named after the 4.3BSD operating system in which each first appeared (which were themselves named after Lake Tahoe and the city of Reno, Nevada). The “Tahoe” algorithm first appeared in 4.3BSD-Tahoe (which was made to support the CCI Power 6/32 “Tahoe” minicomputer), and was made available to non-AT&T licensees as part of the “4.3BSD Networking Release 1”; this ensured its wide distribution and implementation. Improvements, described below, were made in 4.3BSD-Reno and subsequently released to the public as “Networking Release 2” and later 4.4BSD-Lite. The “TCP Foo” names for the algorithms appear to have originated in a 1996 paper by Kevin Fall and Sally Floyd.

Some useful command to save my time

In Linux Ubuntu, the command line ifconfig to configue network interface dynamically:
sudo ifconfig eth0 192.168.0.1 netmask 255.255.255.0 mtu 1500 up

In Linux Ubuntu, the command line ifconfig to configue network interface statically:
sudo vim /etc/network/interfaces, add info and restart network service 
(sudo /etc/init.d/networking restart), for example:
auto eth0
iface eth0 inet static
address 192.168.0.1
gateway 192.168.0.254
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
mtu 1500

In Linux Ubuntu, the command to update "locate db" is
sudo updatedb

In FreeBSD, the command to update "locate db" is
/usr/libexec/locate.updatedb

In Linux Ubuntu, the command line to watch clock time is
watch -n 1 -t date

In Linux Ubuntu, the command line to make a patch is
diff -Naur linux-<version>/net/ipv4/tcp_probe.c linux-<version>net/ipv4/tcp_probe_update.c > patch.txt



Reference:
Howto: Ubuntu Linux convert DHCP network configuration to static IP configuration
FreeBSD IP Alias: Setup 2 or More IP address on One NIC