Applications Must Consider Metrics, Not Just Ideals #powercenter #bitcoin #powerappsportals #etl

After we designed and implemented the trajectory repository in 2010,
new problems quickly appeared.

Perhaps that year Hadoop had already begun to emerge,
but in reality “big data” was still mostly a concept.

Many people had never even encountered truly massive datasets.

Our resources were limited.
We only had a shared Oracle database.

The trajectory repository soon encountered performance bottlenecks.

Relational databases are generally better suited
for returning small result sets.
But trajectory queries often returned very large numbers of rows.

As the time axis extended,
problems with linked data began to appear.

Some records that had been deleted in source systems
became disconnected.

As we integrated more business systems,
we also discovered something unexpected:

Some systems updated data
by deleting records and inserting new ones.

This created even more broken links in the trajectory repository
and significantly affected the user experience.

Later, the industry abstracted these recurring problems
into another knowledge domain.

Within the framework of DAMA International,
they are typically categorized under

Data Integration & Interoperability
and
Data Warehousing & Analytics Support.

History does not offer a chance to redo those early systems.

But future project designs
can benefit from more careful consideration.

Видео Applications Must Consider Metrics, Not Just Ideals #powercenter #bitcoin #powerappsportals #etl канала Chardaway pirry