I was very hopeful when I started looking at this article: Why the days are numbered for Hadoop as we know it, but the article turns out to promise much more than it delivers.
It starts out promising to tell us "what's next" after Google File System and Google Map/Reduce.
But in the end, what it mostly tells us is that it's getting harder and harder to choose a code-word for your project.
So if you want to know about Dremel, Pregel, and Giraph, read the article for some fairly interesting summaries of those projects.
Meanwhile, for something a bit more relevant, and a bit more technical, DBMS industry analyst Curt Monash has updated his top-level survey of database diversity: Database diversity revisited.
Monash's article is filled with links; to follow all of his analysis, you'll spend a lot of time clicking. But it's very much worth the trouble, because he's got lots of thorough summaries of the incredibly rich DBMS ecosystem.
Lastly, can't post about databases without some Postgres goodness, so:
- Some examples of manipulating date/time data and timezones
- Some info about Foreign Data Wrappers. Foreign Data Wrappers allow external data sources to be integrated into Postgres databases and behave somewhat like normal tables.
- A fairly detailed example/tutorial of setting up Slony replication
- The latest from Robert Haas: Absurd Shared Memory Limits. Postgres 9.3 will contain a revised implementation of shared memory. By using anonymous inherited memory segments, most kernel re-configuration is now avoided, and Postgres 9.3 will be one step easier to install, and require one fewer reboot, yay!