Hi HBase Developers, The previous email I sent seem to spur more conversion on the durability of HBase rather than its overall architecture; so I think i would send another email to plead comments on our architectural document ( http://pages.cs.wisc.edu/~suli/hbase.pdf) . We are doing some research on scheduling in storage systems, including HBase. We want to make sure that we are making reasonable assumption assumptions about the HBase architecture.
We did our best reading the code, and used some runtime tools to understand the internal structure of HBase , but it would be best if some developers of HBase could confirm that our understanding is correct (or, pointing out if it is wrong or inaccurate). We drew an document which describes HBase work-flow ( http://pages.cs.wisc.edu/~suli/hbase.pdf) . It emphasizes how different threads in HBase interact with each other, as that's what we are most interested in. We are wondering if you could take a look at this document and let us know your thoughts. We really appreciate your help. Thanks a lot! Suli -- Suli Yang Department of Physics University of Wisconsin Madison 4257 Chamberlin Hall Madison WI 53703
