RTOLAP - Real Time OLAP
Whilst many OLAP Servers like Microsoft Analysis Services store pre-calculating consolidations and calculated elements to achieve rapid response times. A Real Time OLAP Server will calculate the values on the fly, when they are required. The essential characteristic of RTOLAP system is in holding all the data in RAM.
It is a protocol which analyzes fly values when required. It saves every bit of information in RAM. The calculations are executed in a “right-away” manner which reduces the setback linked with “information outburst” since it only saves information under the RAM size standard.
- Since precalculated values aren't stored, the size of a cube in an RTOLAP system is smaller than of an OLAP product which resorts to precalculation. RTOLAP often reduces the problem which may be associated with "Data explosion", by means of storing less data.
- RTOLAP essentially performs calculations "just-in-time" by only calculating values when they are needed space can be saved, since in a precalculated system, a great deal of calculations will be stored which may well never be called up.
- Incremental updates are available once they are loaded, and any modifications to data will flow through the system immediately. With RTOLAP when a change is made, everyone sees the result. This isn't a unique characteristic of RTOLAP, since other OLAP systems (e.g. SAS Institute, Microsoft Analysis Services, MicroStrategy) behave the same way.
- Since RTOLAP stores the entire cube in RAM, it doesn't scale to the data volumes larger than the RAM size
- Performance of queries can be slower since the values need to be calculated on the fly instead of being accessed from the precalculated storage
- Modern big data solutions, like those provided by Perceivant, have removed these disadvantages, while retaining the advantages.