2 ms·
I don't really agree about Hadoop. MapReduce is interesting for some applications but it's not the only thing out there. We did extensive analysis of MapReduc
by jgrahamc 17y ago
I don't really agree about Hadoop. MapReduce is interesting for some applications but it's not the only thing out there. We did extensive analysis of MapReduce at my new company and it clearly wasn't suitable for our needs.
However, the idea of moving code near to the data is a good one since moving data tends to be the slow part. But that's hardly a new concept, just look at Teradata.
- nobody_nowhere 17y agoJust an example of something which benefits from inexpensive computing scale. Pretty narrow applicability. But it seems like there are plenty of other problems which also benefit.
- stingraycharles 17y agoI don't really agree about Hadoop. MapReduce is interesting for some applications but it's not the only thing out there. We did extensive analysis of MapReduce at my new company and it clearly wasn't suitable for our needs. You say that as if MapReduce is the only thing you can do with Hadoop. It has a key/value datastore too, and since recently a relational database: http://db.cs.yale.edu/hadoopdb/hadoopdb.html http://db.cs.yale.edu/hadoopdb/hadoopdb.html
- jgrahamc 17y agoYes, I was rather assuming Hadoop == MapReduce. You are correct that HadoopDB is an interesting development. http://db.cs.yale.edu/hadoopdb/hadoopdb.pdf http://db.cs.yale.edu/hadoopdb/hadoopdb.pdf