VERY large!
Since 1998, we have been regularly invited by IBM Rochester to benchmark Sequel Data Warehouse (formerly RODIN) on each new generation of hardware. These benchmarks have shown that Sequel Data Warehouse scales linearly with the capacity of the system on which it is installed.
Sequel Data Warehouse ETL runs as a native workload on System i and takes advantage of all of the efficiencies and capabilities of the operating system and integrated database. Using our unique parallel loading technology, benchmarks have proven that Sequel Data Warehouse is capable of loading data at a rate in excess of 1 billion rows per hour. Granted, you need a fairly large system to achieve that - but the point is that If you have this much data to load, you probably DO have a very large system - and you want to be sure that the tool you choose can take advantage of it - as Sequel Data Warehouse has been proven to do.
You will note, that the system configuration used in the 2010 benchmark was a realistically configured large system at the time. Now, four years later the Power7+ systems are even larger, and an equivalent system would easily be twice as powerful for probably the same cost. Using the proven linear scalability of Sequel Data Warehouse (in comparison to system CPW), we can see that Sequel Data Warehouse will easily load data at rates in excess of 1 billion rows per hour in our next benchmark.
We have chosen not perform a Power7+ benchmark, and instead will wait for the POWER8 systems to ship later in 2014.
Benchmarks are all well and good - but what about a real customer example:
In Japan, one of the world's largest car manufacturers uses Sequel Data Warehouse to load a data warehouse on a large IBM i system with multiple cores. Data from over 1,000 dealers nationwide is loaded and merged into 270 tables; resulting in over 270,000 separate ETL steps being performed each night, loading hundreds of millions of rows of data. Sequel Data Warehouse has been performing this task flawlessly for many years.
