On Fri, Aug 2, 2013 at 9:29 PM, lars hofhansl <[email protected]> wrote:
> * HConnection: > public HTableInterface getTable(byte[] tableName) throws IOException > public HTableInterface getTable(byte[] tableName, ExecutorService > pool) throws IOException > public HTableInterface getTable(String tableName) throws IOException > Mind adding `public HTableInterface getTable(String tableName, ExecutorService pool) throws IOException` to the permutations? HTableInterfaces are retrieved from the HConnection. By default the > HConnection's ExecutorService is used, but optionally that can be > overridden for each HTable. > This sounds great! // HConnection.getTable(...) is lightweight. The table is really just a > convenient place to call table method and for a temporary batch cache. > Do the semantics of the batch cache change at all, with respect to accumulating mutations per Table/RegionServer/Region? // The HTableInterface returned is not thread safe as before. > I haven't looked at the patch -- can this be placed in big bold letters in the class and/or constructor javadoc? // It's fine to get 1000's of these. > Nice!
