Looking for up-to-date content? Check out IBM's WebSphere, Liberty, Java & DevOps Community, as well as @wasdevnet and @openlibertyio on Twitter!
ESQL shared variables are commonly used to cache database table contents and are usually implemented as an array
containing the result of a SELECT operation. Searching the cache consists of issuing a SELECT against the array but because of
the sequential nature of this search, as the cache grows, the search takes longer such that most users have found that at cache sizes of more than a few thousand rows, it is faster to access the database rather than the cache.
This session demonstrates a novel way of structuring the cache that scales well, and remains faster than DB access even when the cache is larger than 10K rows.
The new cache structure is transparent to applications, so existing caches can be safely converted, potentially resulting in significant reductions in response time and CPU usage and has been successfully demonstrated with two customers, both with large Message Broker installations.
The session also includes a comparison with the new Global Cache capability.
Gerardo Brenner:
Gerardo Brenner is a Senior Consultant with Prolifics and he specializes in Integration solutions using Message Broker and IBM Integration Bus.
He was born in Buenos Aires and after gaining a degree in Computer Science, joined IBM Argentina as a programmer in 1978.
Shortly afterwards, he moved to the UK where he worked for IBM in their Development Laboratories at Hursley initially in CICS Development and later on in Message Broker development, where he specialized in the design, implementation and testing of the ESQL and Database support in MB.
He took early retirement from IBM in 2011 and soon joined Prolifics where his MB expertise has enabled him to work with a number of customers on the design and implementation of solutions to their Integration requirements.
[image not found]
