VirtualBox port-forwarding

Today on my Munich-Florence train, I've waisted some time just trying some ways for "sshing" my puppet VirtualBox guest. Usually DHCPclient does everything and, once the ip address is got by the guest, I can easily "ssh" into it. This time, I had not a wifi connection and I've encoutered some annoying bounces on my ssh attempts. It looked like the guest was unreachable. I've instinctively tried every "nat", "bridged", "host-only" options.. and in the end .. it touched me to read the documentation :) (by the way, a very good paper) 

..and it was so i've discovered this official best practice to port-forward services on a VirtualBox guest

In this example (exactly extracted from the doc), we are going to portforwarding the SSH service, from our HOST 8888 port, to the GUEST (vm name: "Puppet Test Machine") on 22 port:

"sshService" is just a label.

In this way, our HOST will keep the forward on each interfaces. Anyway, it's possible to bind a specific interface though.

Now that we have our forward ready, we can connect the loopback on the given port
This rule will be permanet unless you explicitly remove it. See the rule properties:
Then delete the rule:

Python-stdeb

This is what i've done, to make a shiny python library of mine becomes a fully respectable deb package. Move in your python directory, where everything has been started and create your setup.py file, more or less like this sample:
So, in your directory you list:
Create the stdeb configuration file stdeb.cfg
This file will write parameters in the soon generated DEBIAN/control file. A complete list of the parameters translation is here: http://pypi.python.org/pypi/stdeb#stdeb-cfg-configuration-file
Given that, build your package with:
NOTE: It will compile the package with mypylib-"version"; the version is the one written in setup.py (version="")

enjoy your deb_dist directory.

Mikrotik (winbox): VLAN and bridge

The other day i needed to add a new wifi access point in a new place in our building. Our network, like many others, is partitioned in VLANs. We have almost a dedicated VLAN for any different network segment.

Our best practice, where permitted, is to keep access points as untouched as possible, working the VLANs on the management switches. This develop a better/centralized net control to us. However this time the location was linked to the rest of the infrastructure by a non-management switch.



The "Port 40" is just where the trunk from the unlucky zone, ends on the management switch.

So, since the NON-management switch could not collaborate with us, we are forced to tag the VLAN traffic on the AP itself.



In this article we will start from the default Mikrotik configuration up to achieve our new Bridge for the new VLAN. The only stuff we have already configured, is the switch port (in our case the number 40) marked as tagged with VID 6. Im obviously referring to the management switch.

note: I assume you already have a server for this VLAN (dhcp server and so on ..)

otherwise: http://www.cyberciti.biz/tips/howto-configure-linux-virtual-local-area-network-vlan.html

Let's start opening winbox which allow us to connect the mikrotik through the proper mac address, avoiding lost of connection playing around bridge, VLANs ..

~$ wine winbox.exe

Create the new VLAN interface which will be tied to the master-local port on the router



Why master-local port?

The master-local port on the routerboard is the collection of the slave ports, then i act on that to give the same address (vid 6) for both wireless and every wire connections

Once the VLAN interface is created, move to the "Bridge" tab and close the existing one (bridge-local), by clicking "disable" (the red "X")



.. i've forgot .. this VLAN is basically for guests, so the new bridge will be named "bridge-guest". Open "Bridge" tab and add that:



Move in the near tab "Ports". Double click on "master-local" and replace the "interface" field with the VLAN interface "guest" and the "bridge" field with the just bridge-guest



Configure "wlan1" interface replacing the field "bridge" from ''bridge_local'' to ''bridge_guest''; then go to "IP/DHCP Client". If everything goes ok, you will be able to get an ip address from your server on the other side. When on the tab, click on "+" and in the "Interface" field, choose "bridge-guest". Click ok and



thats all.

D-link switches and .. "martians packets" ??

Device: Dlink DGS-1224T
Firmware: 4.21.01
IP addr: 10.2.2.8

I have a VLAN interface on my linux router built on eth2 with VID 6

eth2 = 10.2.2
eth2.6 = 10.6.6 (VLAN)

This VLAN is configured on my Dlink switch as well with untagged VLAN ports. My Dlink management interface is set on 10.2.2

Since we use shorewall as iptables interface, we have found this on syslog:

Sep 10 10:03:56 lurch kernel: [2651431.667008] martian source 10.2.2.1 from 10.2.2.8, on dev eth2.6
Sep 10 10:03:56 lurch kernel: [2651431.667011] ll header: ff:ff:ff:ff:ff:ff:00:15:63:45:05:c5:08:04

Martians ??

A possible explanation:
Packets with 10.2.2 as source reach the Dlink management interface on 10.2.2.8 (for management or just a nagios ping); as reply, the Dlink perform an ARP request like "Who has 10.2.2.x" on every ports, even on the VLAN 10.6.6 ones. As result, the Linux kernel see those 10.2.2 source packets on its eth2.6 interface and complains consequently.

A possible fixaround:
Force Dlink management interface being reachable from one network only, in my case 10.2.2, on "Configuration/802.1Q Management VLAN" tab.

Nagios: check_http with --extra-opts

We have to check a https page which needs a login.

We want to store our login informations in a nagios only-readable file.

root@lurch:~# touch /etc/nagios/private/custom.ini
root@lurch:~# chown root.nagios /etc/nagios/private/custom.ini
root@lurch:~# chmod 640 /etc/nagios/private/custom.ini

In this example Im going to create an [https] section which includes the credentials for my login web page. This file, which will be passed to --extra-opts= plugin argument, could contains all the options listed by ./check_http -h command in the literally form. For instance:

./check_http -h
...
-a, --authorization=AUTH_PAIR
    Username:password on sites with basic authentication
...

Then:

root@lurch:~# su - nagios
nagios@lurch:~$ vim /etc/nagios/private/custom.ini

[https]
authorization = *user*:*password*

Let's write the 'commands' section. I want to call this command like 'check_https_auth'

nagios@lurch:~$ vim /etc/nagios/objects/commands.cfg

# zmo: check_https_auth 
define command{
	command_name    check_https_auth
	command_line    $USER1$/check_http -H $ARG1$ -I $HOSTADDRESS$ -s $ARG2$ --ssl -c $ARG3$ -u $ARG4$ --extra-opts=https@/etc/nagios/private/custom.ini
	}

Nagios: check remote hosts via SSH with command validation

Nagios server: Fedora release 16 (Verne)
Nagios remote node: Debian GNU/Linux testing (wheezy)


Basic SSH Configuration
Let's allow the access from the nagios server to the nagios node creating a ssh key passphrase-less:

zmo@nagiosSer:~$ cd .ssh/
zmo@nagiosSer:~/.ssh$ ssh-keygen -t dsa
Generating public/private dsa key pair.
Enter file in which to save the key (/home/zmo/.ssh/id_dsa): nagios     
Enter passphrase (empty for no passphrase): 
Enter same passphrase again: 
Your identification has been saved in nagios.
Your public key has been saved in nagios.pub.
The key fingerprint is:
....

.. so that you will have nagios and nagios.pub from the others

zmo@nagiosSer:~/.ssh$ ls
id_rsa id_rsa.pub nagios nagios.pub

On the remote node create the nagios user and the necessary file

root@nagiosNod:~# adduser nagios
Adding user `nagios' ...
Adding new group `nagios' (1006) ...
Adding new user `nagios' (1006) with group `nagios' ...
...
Enter new UNIX password: xxxxxxxx
Retype new UNIX password: xxxxxxxx
...
root@nagiosNod:~# su - nagios
nagios@nagiosNod:~$ mkdir .ssh
nagios@nagiosNod:~$ touch .ssh/authorized_keys

Paste your pub key, from the nagios server, into the remote nagios authorized_keys file

zmo@nagiosSer:~/.ssh$ cat nagios.pub | ssh nagios@nagiosNod.remote.com 'cat >> /home/nagios/.ssh/authorized_keys'
nagios@nagiosNod.remote.com's password: xxxxxxxx

Have a try

zmo@nagiosSer:~$ ssh -i .ssh/nagios nagios@nagiosNod 
Last login: Tue Aug 28 20:44:54 2012 from nagiosSer
nagios@nagiosNod:~$

On the nagios server

root@nagiosSer:~# yum install nagios-plugins-by_ssh.i686

On the nagios remote node
Install the nagios plugins

root@nagiosNod:~# apt-get install nagios-plugins-basic

NOTE: you can find where the plugins have been stored by

root@nagiosNod:~# dpkg -L nagios-plugins-basic
...
/usr/lib/nagios/plugins/check_apt
/usr/lib/nagios/plugins/check_tcp
/usr/lib/nagios/plugins/check_dummy
/usr/lib/nagios/plugins/check_ntp
/usr/lib/nagios/plugins/check_nwstat
/usr/lib/nagios/plugins/check_load
/usr/lib/nagios/plugins/check_procs
...

On the nagios server
Let's write a SSH config file

root@nagiosSer:~# cd /etc/nagios
root@nagiosSer:~/etc/nagios# vim check_by_ssh_config

Host nagiosNod 84.68.x.x
	User nagios
	Hostname nagiosNod.remote.com
	IdentityFile /home/zmo/.ssh/nagios

This file will be passed as ssh configuration to check_by_ssh. You can obviously add others hosts.

Configure the commands.cfg

root@nagiosSer:~/etc/nagios# vim objects/commands.cfg

# Check Remote Disk
define command {
        command_name    check_remote_disk
        command_line    $USER1$/check_by_ssh -H $HOSTADDRESS$ -F /etc/nagios/check_by_ssh_config -C "/usr/lib/nagios/plugins/check_disk -w $ARG1$ -c $ARG2$ -p $ARG3$"
}

A little explanation:
* $USER$ - the user which nagios run (nagios)
* $HOSTADDRESS$ - the numeric IP address definied in the server configuration file (/etc/nagios/objects/nagiosNod.cfg), mapped on "address" field
* -F /etc/nagios/check_by_ssh_config - The SSH configuration file previously written
* -C /usr/lib/nagios/.. - The command will execute on the remote node

Its better (not mandatory) to put that services in a different group. So, open /etc/nagios/objects/templates.cfg (or services.cfg) and append:

# Remote service
define service{
	name				remote-service
	use				generic-service
        max_check_attempts              4
        normal_check_interval           5
        retry_check_interval            1
        register                        0
	}

Now in /etc/nagios/objects/nagiosNod.cfg file we could append:

define service {
        use                             remote-service
	host_name			nagiosNod
        service_description             Home Partition
        check_command                   check_remote_disk!20%!10%!/home
	}

SSH validation
This step is for allow just the check_* commands execution from the nagios server on the nagios nod; not other commands, not a login or whatever else.

Create a validation commands file:

nagios@nagiosNod:~# touch /usr/local/bin/validate-nagios-check
nagios@nagiosNod:~# chown nagios.nagios /usr/local/bin/validate-nagios-check
nagios@nagiosNod:~# su - nagios
nagios@nagiosNod:~$ vim /usr/local/bin/validate-nagios-check
#!/bin/sh
# Ensure 
case "$SSH_ORIGINAL_COMMAND" in
	*\&*)
		echo "Rejected"
		;;
	*\;*)
		echo "Rejected"
		;;
	/usr/lib/nagios/plugins/check_*)
		exec $SSH_ORIGINAL_COMMAND
		;;
	*)
		echo "Rejected"
		;;
esac

A little explanation: $SSH_ORIGINAL_COMMAND is our /usr/lib/nagios/plugins/check_* . Using exec here, will prevent to open a shell for the command execution but just to exec the command.

Let's back to our authorized_keys. We have to append in the beginnin of the key definition (ssh-dss or ssh-rsa) the string to execute our validate-nagios-check script from the nagios server.

nagios@nagiosNod:~$ vim .ssh/authorized_keys
from="73.224.x.x",command="/usr/local/bin/validate-nagios-check" 
ssh-dss AAAAB3NzaC1kc3MAAACBAKK3jM9O+cmjPufrn9Ie7q+iJJu+1B0bHH6lhfYC8KandEIWP0gMGz4v...
...

You can test the validation script from the nagios server to the node.

root@nagiosSer:~# ssh -i /home/nagios/.ssh/nagios nagios@nagiosNod.remote.com
Rejected
Connection to nagiosNod.remote.com closed.
root@nagiosSer:~# ssh -i /home/nagios/.ssh/nagios nagios@nagiosNod.remote.com /bin/ls
Rejected
root@nagiosSer:~# ssh -i /home/nagios/.ssh/nagios nagios@nagiosNod.remote.com /usr/lib/nagios/plugins/check_http -hcheck_http v1.4.16 (nagios-plugins 1.4.16)
Copyright (c) 1999 Ethan Galstad 
Copyright (c) 1999-2011 Nagios Plugin Development Team
...

Dnsmasq and the network "pushing" on CentOS

This simple post shows you how tu push networks to your clients using the Dnsmasq DHCP server.

A possible scenario would be, your DHCP server (192.168.0.1) gives you 192.168.0.x address but you also need to reach 172.16.x.x and 10.0.x x networks, for access to (for instance) some VPN resources in your LAN. Those networks are NOT handled by your DHCP server. The quick solution is to make your DHCP server "pushing" that networks. And that is possibile thanks to "121" dhcp-option.

Let's assume our network id may be "vicinet" (you can also miss the network id, of course):

dhcp-option = vicinet, 121, 10.0.0.0/16, 192.16.0.1, 172.16.0.0/16, 192.168.0.1

The format will always be:

dhcp-option = 'id_net', 121, NETWORK, GATEWAY