Chris Roland » Data

Best database engine or access for F#?

Chris Roland — Sun, 07 Mar 2010 03:46:10 +0000

After going through a couple blog articles on F# access to different database engines, I wondered what’s the best database engine/access for F#?

I use NHibernate and SQL Server on a daily basis from C#, so I naturally want to use NHibernate from F#. After some quick googl…research, I found this great article on how to setup F# to use NHibernate:
http://strangelights.com/blog/archive/2009/12/20/1650.aspx

I also came across an article on how to access MongoDB from F#:
http://codebetter.com/blogs/matthew.podwysocki/archive/2010/02/09/exploring-mongodb-with-f.aspx

After going through both and doing a couple quick spikes, MongoDB seems like a natural fit for F# and the functional way of life. Keep in mind, I’m new to functional programming and F#, so take this with a grain of salt and do your own research.

With F#’s parallel features, NoSQL database engines (MongoDB, CouchDB, etc) seem like a perfect fit. Large distributed data stores, asynchronous parallel processing, data stored as key/value pairs/objects, you get the picture.

It would be great to hear from anyone who’s used either with F# and your experiences. For now, I’m going to do some prototyping down both paths to see where what comes out the other side.

LINQ to SQL data mapping

Chris Roland — Sat, 28 Nov 2009 02:50:28 +0000

I’m currently working on an ASP.NET MVC project and I’m using LINQ to SQL to manage my data. One of the fields I’m working with is stored as a MONEY data type in SQL Server. I wanted to find the best CLR data type to use when I mapped the field in my LINQ Data Context. After some researching I came across a great MSDN article showing what types to use when mapping back to the database with LINQ. There is even an awesome grid showing the best types to use and the expected data loss.

Here is the link:
http://msdn.microsoft.com/en-us/library/bb386947.aspx

Installing Hypertable on Debian

Chris Roland — Mon, 06 Jul 2009 20:15:28 +0000

To really understand data management solutions like Hypertable (http://www.hypertable.org/), I need to install and use it. So to start out, I decided to just get a single node installed using the local filesystem.

I use Debian Lenny, which is a Linux distribution and even with a bunch of libraries installed, I still had to do some dancing.

Getting Started

First step, read the README, no really, there are bunch of dependencies. I felt like I was installing GnuCash before apt/yum/yast. For the Debian distribution you need to apt-get the following:

apt-get install g++ make cmake libdb4.6++-dev libboost-*-dev liblog4cpp5-dev expat-devel libreadline5-dev zlib1g-dev libevent-dev

Installing Hyperic-Sigar

I followed the instructions in the README without any issues.

Installing Thrift

Several dependencies I ran into are not in the README, but these are thrift dependencies so I guess it’s ok. The bootstrap.sh script errored, stating it was looking for the autoscan command. Then I ran into several other errors when Thrift tried to compile. So I put the full list of the remaining dependencies below:

apt-get install autoconf libtool byacc flex ruby-dev

Compiling Hypertable

The final and last step, compile! Well… sort of. I downloaded hypertable from the website and for some odd reason (could be me), I ran into an issue trying to run cmake against the source. In the README they mention grabbing the source from a git repository. So, I grabbed the source from the repository and bam!, it compiled. Remember to read the README, setup the build directory and src directory. I don’t think it matters where they are, just make sure you have them in separate directories.

I hope this helps someone and if anyone has any changes or comments I would like to hear them.

Next step, put Hypertable to work.