I have a home server I use for all kinds of stuff when I’m away. I have set it up recently so that it can serve a folder mounted over the network from a Windows machine.

And this is where the fun started!

The folders and files appear correctly, I can navigate and I can download them. Only that the files, even if apparently having the same size are corrupted. I checked the FollowSymLinks and as expected was set as required. I tried wget in the command prompt to see if I could find some enlightment but I got the following errors:

# wget http://192.xxx.xxx.xxx/windowsshare/folder/file.jpg
 --09:18:00--  http://192.xxx.xxx.xxx/windowsshare/folder/file.jpg
 => `file.jpg'
 Connecting to 127.0.0.1:80... connected.
 HTTP request sent, awaiting response... 200 OK
 Length: 763,685 (746K) [image/jpeg]
 
 0%
 [                                                                  ]
 0             --.--K/s
 
 09:18:00 (0.00 B/s) - Connection closed at byte 0. Retrying.
 
 --09:18:01--  http://192.xxx.xxx.xxx/windowsshare/folder/file.jpg
 (try: 2) => `file.jpg'
 Connecting to 127.0.0.1:80... connected.
 HTTP request sent, awaiting response... 200 OK
 Length: 763,685 (746K) [image/jpeg]
 file.jpg has sprung into existence.
 Retrying.
 
 --09:18:03--  http://192.xxx.xxx.xxx/windowsshare/folder/file.jpg
 (try: 3) => `file.jpg.1'
 Connecting to 127.0.0.1:80... connected.
 HTTP request sent, awaiting response... 200 OK
 Length: 763,685 (746K) [image/jpeg]
 
 0%
 [                                                                  ]
 0             --.--K/s

and it will continue like this creating empty files.

Especial the has sprung into existence baffled me. Also note that symlinking any other folder, even mounted as smbfs instead of cifs worked perfectly. At this point changing the mount to smbfs would have solved my problem but curiosity was stronger.

As it turns either apache or (more likely) the cifs driver has a bug. You can work around it by setting the following inside your apache configuration.

EnableSendfile off

Further upgrades will probably eradicate this problem.

A backup NAS I set up for a client had some problems – i.e. from time to time the network card would hang forcing a reboot (the machine was headless). The disks in the nas are RAID 1 providing redundancy in case of failure. We wanted the same with the network. So, nothing easier than to install a second network card and set up bonding.

What is network bonding?

Bonding is the same as port trunking. It means that two (ore more) physical network connections are combine to form one virtual network connection. The added benefit is that bandwidth adds too.  So theoretically this could offer multi gigabit capabilities.

How is it configured

First you need a small utility that comes with the linux kernel: ifenslave. So we must compile it. After that it should be moved somewhere more apropriate.

cd /usr/src/linux/Documentation/networking
gcc -Wall -O -I/usr/src/linux/include ifenslave.c  -o ifenslave
cp ifenslave /sbin/ifenslave

Then we need some scripts that will set up bonding on boot. So we go to /etc/rc.d and create a new file called rc.bond

cd /etc/rc.d
touch rc.bond

Edit the file you just created to contain the following:

#!/bin/sh
 
        case "$1" in
          'start')
            echo "start bond0"
            modprobe bonding mode=balance-rr miimon=100
            modprobe tg3
            ifconfig bond0 up
            ifenslave bond0 eth0
            ifenslave bond0 eth1
            #You don't necesarily need to change the hardware address
            #It will take it from the first hardware interface
            #ifconfig bond0 hw ether 00:16:3e:aa:aa:aa
          ;;
          'stop')
            ifconfig bond0 down
            rmmod bonding
            rmmod tg3
          ;;
          *)
            echo "Usage: $0 {start|stop}"
          ;;
        esac

Save this file and make it executable

chmod +x /etc/rc.d/rc.bond

We need to start this on boot. So edit rc.M You need to find the line containing #Initialize the networking hardware and insert this after it

# If script rc.bond is executeable then start it
if [ -x /etc/rc.d/rc.bond ]; then
  . /etc/rc.d/rc.bond start
fi

Finally we must configure rc.inet1.conf to set our bond interface and to make sure the physical interfaces are not configured. Add the following:

IFNAME[4]="bond0"
IPADDR[4]="XXX.XX.XX.XX"
NETMASK[4]="255.255.255.0"
USE_DHCP[4]=""
DHCP_HOSTNAME[4]=""

Make sure that the physical interfaces, like 0, 1 and so on are NOT configured.
Reboot, and everything will be set.

For your reference those are the bonding modes available:

mode=0 (balance-rr)
Round-robin policy: Transmit packets in sequential order from the first available slave through the last. This mode provides load balancing and fault tolerance.

mode=1 (active-backup)
Active-backup policy: Only one slave in the bond is active. A different slave becomes active if, and only if, the active slave fails. The bond’s MAC address is externally visible on only one port (network adapter) to avoid confusing the switch. This mode provides fault tolerance. The primary option affects the behavior of this mode.

mode=2 (balance-xor)
XOR policy: Transmit based on [(source MAC address XOR'd with destination MAC address) modulo slave count]. This selects the same slave for each destination MAC address. This mode provides load balancing and fault tolerance.

mode=3 (broadcast)
Broadcast policy: transmits everything on all slave interfaces. This mode provides fault tolerance.

mode=4 (802.3ad)
IEEE 802.3ad Dynamic link aggregation. Creates aggregation groups that share the same speed and duplex settings. Utilizes all slaves in the active aggregator according to the 802.3ad specification.

	Pre-requisites:
	1. Ethtool support in the base drivers for retrieving
	the speed and duplex of each slave.
	2. A switch that supports IEEE 802.3ad Dynamic link
	aggregation.
	Most switches will require some type of configuration
	to enable 802.3ad mode.

mode=5 (balance-tlb)
Adaptive transmit load balancing: channel bonding that does not require any special switch support. The outgoing traffic is distributed according to the current load (computed relative to the speed) on each slave. Incoming traffic is received by the current slave. If the receiving slave fails, another slave takes over the MAC address of the failed receiving slave.

	Prerequisite:
	Ethtool support in the base drivers for retrieving the
	speed of each slave.

mode=6 (balance-alb)
Adaptive load balancing: includes balance-tlb plus receive load balancing (rlb) for IPV4 traffic, and does not require any special switch support. The receive load balancing is achieved by ARP negotiation. The bonding driver intercepts the ARP Replies sent by the local system on their way out and overwrites the source hardware address with the unique hardware address of one of the slaves in the bond such that different peers use different hardware addresses for the server.

[source: http://www.linuxhorizon.ro/bonding.html]

The most used are the first four mode types…

Recently I set-up a machine in my closet to act as a backup server. One of it’s jobs is to download via pop3 all the email from our Gmail accounts just in case the cloud goes puf . For this job I used fetchmail to fetch the emails to local users on the machine. All nice and simple … you’d think.

The mail didn’t get delivered because sendmail complained about the above error. Now this is weird – I just want to have the mails delivered locally. Well, since spam is such a big problem sendmail likes to check that the domain from which the message originated is real.

Well, since some of the old emails we have are from addresses that no longer exists sendmail refused to deliver them. So we just have to tell email to accept email coming from inexistant domains.

Note: that this is not safe or recommended in anyway if the sendmail is running on a live email server. It’s acceptable in this case since this sendmail will only get messages downloaded by fetchmail from a few google accounts.

So the setting to change in order to allow messages from nonexistent domains is found in

/usr/share/sendmail/cf/cf/sendmail-slackware.mc

and it’s this one:

dnl# Turn this feature on if you don't always have DNS, or enjoy junk mail:
dnl FEATURE(`accept_unresolvable_domains')dnl

and after changing it you must run:

cd /usr/share/sendmail/cf/cf
cp sendmail-slackware.mc config.mc
 m4 /usr/share/sendmail/cf/m4/cf.m4 config.mc > /etc/mail/sendmail.cf

As you know Linux does not use drive letters like Windows does. All files are organized under a big tree hierarchy. We use the mount command to mount partitions and this is the same command used to mount remote windows partition or windows shares.

Please note that you’ll need to know the following:

• Windows username and password (the ones needed to access the share)
• The machine and share name. Something like //xpdesktop/pictures
• Have root access on the linux machine

Login on your linux machine as root. Type the following command to mount your remote windows share:

mkdir /mnt/share_name
mount -t cifs //xpdesktop/pictures -o username=john, password=johnspass /mnt/share_name

Where:

• xpdesktop is the network name of the Windows machine
• john is the username on  the windows machine
• johnspass is the password used to access the windows share
• /mnt/share_name is the target folder where the partition will be mounted (the folder you just created above)

You can now cd and ls the remote windows share as any other folder on the linux machine:

cd /mnt/share_name
ls  -l

I’ll show you in a following article how you can automatically mount a remote Windows share at boot time.

I have a headless machine at home that I plan to use as a backup and development server. Even though I’m a Skackware guy I chose to install Debian on it. Of course I could have hooked a monitor and keyboard to it but for the sake of exercise I wanted to see if it’s possible to install debian using ssh from the very beginning.

The short answer: it’s possible … maybe. Your headless machine should already be set to boot of cd if one is present, otherwise it will not work.

Here are the steps needed to start you off:

1. Download the netinst cd image

wget http://mirrors.kernel.org/debian-cd/5.0.3/i386/iso-cd/debian-503-i386-netinst.iso

2. Mount the ISO to a folder, let’s call it isoorig

mkdir isoorig
mount -o loop -t iso9660 debian-503-i386-netinst.iso isoorig

3. Extract to new folder called isonew

mkdir isonew
rsync -a -H –exclude=TRANS.TBL isoorig/ isonew

4. Change the menu to load SSH on boot by default

/isonew# nano isolinux/txt.cfg

DELETE:
remove “menu default” from “label install”

ADD:

label netinstall
menu label ^Install Over SSH
menu default
kernel /install.386/vmlinuz
append auto=true vga=normal file=/cdrom/preseed.cfg initrd=/install.386/initrd.gz locale=en_US console-keymaps-at/keymap=us

CHANGE:
“default install” to “default netinstall”

EDIT: both files below and change “timeout 0″ to “timeout 4″ to make it auto select netinstall

nano isolinux/isolinux.cfg
nano isolinux/prompt.cfg

5. Create preseed.cfg file

nano isonew/preseed.cfg

6. PASTE this to the preseed file:

#### Contents of the preconfiguration file
### Localization
# Locale sets language and country.
d-i debian-installer/locale select en_US
# Keyboard selection.
d-i console-keymaps-at/keymap select us
### Network configuration
# netcfg will choose an interface that has link if possible. This makes it
# skip displaying a list if there is more than one interface.
d-i netcfg/choose_interface select auto

# Any hostname and domain names assigned from dhcp take precedence over
# values set here. However, setting the values still prevents the questions
# from being shown, even if values come from dhcp.
d-i netcfg/get_hostname string newdebian
d-i netcfg/get_domain string local
# Disable that annoying WEP key dialog.
d-i netcfg/wireless_wep string
# The wacky dhcp hostname that some ISPs use as a password of sorts.
#d-i netcfg/dhcp_hostname string radish
d-i preseed/early_command string anna-install network-console
# Setup ssh password
d-i network-console/password password install
d-i network-console/password-again password install

7. Recreate md5sum.txt file

md5sum `find -follow -type f` > md5sum.txt

8. Create your new iso image

mkisofs -o ../custom_install.iso -r -J -no-emul-boot -boot-load-size 4 -boot-info-table -b isolinux/isolinux.bin -c isolinux/boot.cat ../isonew

The image you just obtained is ready to burn. This loads everything automatically and goes to the SSH screen.

Thanks go to people on various debian forums, this information was collected via trial and error and by combining info found around. Have fun with your shiny new headless debian box.

Nokia 7610

As promised I follow with an article on how to breathe new life into an aging phone.

Well, the 7610 was never lightning fast and over the years it became more and more troublesome, to the point where my wife thought it was about to break. She also got a good deal on a 500 Mb/month data plan so new software was in order. Although the prospect of a new gadget in the house was appealing I decided to see what was to be done with this one.

Since this was the first smartphone in the family, when it was new, we tried all kinds of apps on it installing and uninstalling a lot of stuff. Since the original card was small some were installed in the phone’s memory. Even if today is good practice to install your apps on the phone’s memory so you can swap cards easily it’s not the same with a phone with just a few megs of space.

So what I did to get it up to speed:

1. I went out and searched the biggest RS-MMC card I could find. I managed to find a 1 Gb card for just a few bucks. It was the last in stock
2. I used Nokia PC Suite to back-up contacts and messages (I didn’t use the back-up tool – that one, in my opinion, also backs up a lot of junk, junk you try to get rid of in the first place).
3. I pressed Green * 3 and started the phone, I kept them pressed untill Formatting apeared on screen. This is for a complete software reformat.
4. I added the contacts, calendar, messages backed up at step 2.
5. I installed the following:
   1. slick – for messenging
   2. opera mobile – for faster browsing
   3. gmail – for full email access, it’s slow but works, for the quick inbox check I just configured the default email client
   4. Mobireader – for the occasional reading on the road
   5. Frozen bubble for S60
   6. Adobe pdf, quickoffice – she got some free versions with the phone, it’s enough for what she needs on the road.
   7. Dictionaries – she’s a professional translator – interpreter so those should come handy

Everything was installed on the 1 Gb mmc card leaving the phone memory free so now the phone is able to receive big files via bluetooth, and has enough space for contacts, calendar and so on.

Even if the phone is S60v2 I could still find a wealth of software available, many free, some payed.

The phone still occasionally locks up when answering a call but not as often as it did before, but again this is a problem that all Nokia 7610 phones I saw had. It’s still slow, especially when browsing or running the gmail java app, but it’s usable again, even for some light web, messenger, email tasks.

Ok, so I wanted a NAS on my home network. Having just aquired a shabby XPC SK43G that simply refused to work as a htpc (it’s original purpose) I decided to set it up as a NAS (Network Attached Storage). I went out and bought a Western Digital Green 1 Tb Sata2 drive. The xpc refused to detect the drive until I set it via jumper as Sata1. I had to search on the Western Digital support pages, as the jumper settings are not documented on the drive itself. I added a 512 MB DDRAM I had lying around, underclocked and undervolted the cpu to reduce heat and noise and so, the hardware part was done.

The next step was choosing the os and software. My first intent was Slackware set up to share over Samba and NFS. But a quick google search later I discovered the FreeNAS Project. I decided to give it a shot and useless to say I was hooked. It has cifs, nfs, upnp, ftp and http sharing built in, and more features that I could possibly list here. Just check their page. BTW if you are a flamboyant linux supporter you might want to note that FreeNAS is based off FreeBSD and not linux. It’s not a problem for me, as I am an adept of “chose by functionality, not by religion” but as always YMMV.

The system boots of a 2 Gb USB drive on which I installed the FreeNAS image. The setup was easy and straightforward.I also set up the hard drive to power down after 20 minutes of inactivity so the flash boot scheme works great.

I intend to add 2 more drives in the future so I can do RAID 5 on it but even now it’s a great machine to have on the network. It’s main purpose for now is backup of my other machines and my wife’s laptop.

Another open source project that proves itself a valuable asset to have !