wesley tanaka


Installing Ubuntu onto a NFS mount


It is possible to install Ubuntu onto a NFS networked filesystem directory, which you might want to do in order to boot that copy of Ubuntu over the network on a diskless thin client computer.

Set up the base filesystem:

Booting Ubuntu from a network filesystem


It is possible to boot Ubuntu over the network on a computer with no local hard drive. You need:

  • A DHCP server that supports PXE
  • A TFTP server with enough space to host the kernel binary
  • An NFS server with enough space to host the root filesystem

The steps involved are:

Building debian .deb packages from source


These are my notes on building debian packages from source.

The steps in preparing a binary debian packages are roughly:

  1. Upstream source code (e.g. .tar.gz)
  2. Unpacked source code directory
  3. Unpacked source code directory containing a debian/ subdirectory
  4. Debian prepared source code (.orig.tar.gz) containing the debian/ directory
  5. Source package (.dsc)
  6. Binary package (.deb)

Steps without using ubuntu bzr:

LXC command cheat sheet / quick reference

lxc-create -t download -n mycontainer -- -d ubuntu -r trusty -a amd64
Create an (unstarted) container. Other architectures include: i386
lxc-start -n mycontainer -d
Boot up a container -- excluding -d will also run lxc-console
lxc-attach -n mycontainer
Execute a root shell inside of a started container
lxc-console -n mycontainer
Open a login screen inside of a started container
lxc-stop -n mycontainer
Shut down a started container

Progress Bar Project Management


To an outsider looking in, commercial software projects* should look like a well designed progress bar.

Dockerfile library


I've created some scripts and Dockerfiles to build docker images. There are images for:

Ways to download a file via HTTP from the command line


Current Ubuntu distributions do not come with either wget or curl installed, but they do include Python and Perl. Here are some alternative ways to download a file and print it on stdout without wget or curl:

Shaing your home directory between Mac OSX and a VMware Fusion 7 Linux Virtual machine


First, give your virtual machine a fixed IP address.  VMWare Fusion's dhcpd server will often change the IP address assigned to your virtual machine -- even while it's still running.  While you can give the entire IP address range block access to your home directory over NFS, that can disrupt some other tools like Vagrant.

% cat /etc/exports
/Users/wtanaka -maproot=root:wheel

Restart nfsd in Mac OS:

% sudo nfsd stop
% sudo nfsd start

Assigning a fixed IP address to a virtual machine in VMware Fusion 7


1. Find the dhcpd.conf corresponding to the network that your virtual machine is on.  By default the NAT network uses /Library/Preferences/VMware Fusion/vmnet8/dhcpd.conf. You can find the right one by running ifconfig (or ipconfig in Windows) inside your virtual machine to find its IP address, and look for a range containing that IP address in all of the dhcpd.conf files.

2. Edit the file to add the portions in green.  Be careful not to use a number in the dynamically allocated range, the host computer's IP, or the DNS IP:

Deleting a single URL from Chrome History

  1. Go to chrome://history/
  2. Search for the URL in the "Search history" text field
  3. Click and shift click to select all instances
  4. Click "Remove selected items"
Syndicate content
by Wesley Tanaka