Oracle and IBM databases: Disk-based vs In-memory
The
case for in-memory databases (IMDB) can be made in three simple points (1)
performance - data is kept in RAM so no disk I/O limitations (2) HA
with built in fail-over (3) support for relational schema and SQL.
Current disk based RDBMS can run out of steam when processing large
data. Can these problems be solved by migrating from a disk based
RDBMS to an IMDB? Any limitations? To find out, I tested
one of each from the two leading vendors who together hold 70% of
the market share - Oracle's 11g
and TimesTen 11g,
and IBM's DB2
v9.5 and solidDB
6.3.
Continue reading "Oracle and IBM databases: Disk-based vs In-memory" »



Subscribe