For quite a while, I have been thinking on blogging about Hadoop in general and Hadoop distributed file system (HDFS) in particular. Why, you may ask?
Firstly, I have been contacted by students from as far as Bangladesh and Fiji asking me questions about HDFS via email. This made me think that disseminating internal details about HDFS to the whole wide world would really benefit a lot of people. I like to interact with these budding engineers; and their questions, though elementary in nature, sometimes makes me really ruminate on why we adopted a particular design and not another. I will sprinkle a few of these examples next week.
Secondly, I visited a few Universities last month, among them Carnegie Mellon University and my alma-mater Univ of Wisconsin. On my flight, I was getting bored to death, because I really did not like the movie that was playing and I did not carry any material to read. (Usually I like to read and re-read Sherlock Holmes over and over again.) But like they say, " an idle mind is the devil's workhop".... I started to jot down some exotic design ideas about HDFS.... And lo behold, I have a list of ideas that I would like to share! I will post them next week as well.