Store data in memory completely

[dulanov]

Hello, can babudb store data in memory as redis can or it read them
from files every time (from linux disk cache in reality)? My
application read/write ratio is more then 1000 and data fits in to the
memory completely, so that feature is quite sensitive.

[Jan Stender]

Hi,

BabuDB keeps all data in the main memory until it exceeds a certain size
limit. At that time, it creates one or more checkpoint files, which are
typically memory-mapped and thus may have to be paged in from disk when
being accessed.

The Java release of BabuDB allows you to adjust the
"babudb.maxLogfileSize" configuration parameter to change the
checkpointing limit. If you know that your data will never exceed your
main memory, you can simply adjust it to some value larger than your
main memory in bytes. This will effectively prevent checkpoints from
being created and ensure that no data is read from disk anymore once the
system is up and running.

Hope this helps and best regards,
Jan