Sorry, I meant *TaskTracker. The main purpose of the HBase cluster it to run Map-Reduce jobs. The random read load is fairly low. Currently, I have allocates 5 G to the Region Server, and 500 MB to Task Tracker and 500 MB to DataNode.
On Fri, Oct 14, 2011 at 2:51 PM, Arun C Murthy <[email protected]> wrote: > You mean DataNode, _TaskTracker_ and RegionServer I assume. > > DN and TT should not need more than 1G each - even that is excessive. You > want much of your resources free for actual work (map/reduce tasks if you are > running Hadoop MR). > > I'll let HBase experts comment on RS. > > Arun > > On Oct 14, 2011, at 2:45 PM, Sam Seigal wrote: > >> Hi All, >> >> I have the Datanode, JobTracker and RegionServer daemons running on a >> fleet of machines. Each of these machines have 8 G of memory and are >> dedicated hardware for running HBase. How do you guys decided which % >> of memory to assign to each ? What should this number be dependent on >> ? >> >> Thank you, >> >> Sam > >
