Sybase IQ Introduction

Sybase IQ is an ana­lyt­ics serv­er designed spe­cific­ally for ana­lyt­ics, data ware­hous­ing, and busi­ness intel­li­gence envir­on­ments. Dis­tin­guished from con­ven­tion­al data­bases by its column-ori­ented, grid-based archi­tec­ture, pat­en­ted data com­pres­sion, and advanced query optim­izer, Sybase IQ is deployed in a wide vari­ety of set­tings deliv­er­ing unmatched per­form­ance, flex­ib­il­ity, and eco­nomy in even the most chal­len­ging report­ing and ana­lyt­ics environments.

Sybase IQ provides sig­ni­fic­antly faster query per­form­ance than tra­di­tion­al data­base envir­on­ments which were designed primar­ily for pro­cessing trans­ac­tions. The improve­ment over such solu­tions can be 10 to 1000 times great­er, even as thou­sands of users and years worth of data are added to the envir­on­ment, and even in the face of the most com­plex quer­ies and intens­ive pre­dict­ive analyses.

In a con­ven­tion­al rela­tion­al data­base man­age­ment sys­tem (RDBMS), data is stored and retrieved hori­zont­ally. Such is the ideal struc­ture for log­ging and pro­cessing trans­ac­tions, but not for ana­lyz­ing the data. Without build­ing indexes and views in such an envir­on­ment, retriev­ing needed inform­a­tion via quer­ies cre­ates a massive work­load. Build­ing and main­tain­ing indexes and views requires a sig­ni­fic­ant resource invest­ment, and even then views and sum­mar­ies still do not give a com­plete pic­ture, so ad hoc quer­ies remain prob­lem­at­ic. Addi­tion­ally, the data­base stor­age foot­print must be dra­mat­ic­ally expan­ded to make the envir­on­ment effi­cient for com­plex quer­ies . In such an envir­on­ment, a single tera­byte of raw input data will often expand to six tera­bytes or more of sum­mar­ies, aggreg­ates, and views. Between the labor-intens­ive man­age­ment and data bloat, such an envir­on­ment can be very expens­ive to maintain.

With Sybase IQ, data is stored and retrieved ver­tic­ally, mak­ing inform­a­tion highly access­ible for ana­lys­is. Each column of the data­base is stored sep­ar­ately, and the data itself serves as the index required to per­form vir­tu­ally any query. Query­ing a Sybase IQ envir­on­ment means retriev­ing only the columns used in that spe­cif­ic query, keep­ing the work­load down to a man­age­able level. Since the data itself is the index, there is no need to cre­ate and main­tain elab­or­ate sum­mar­ies, aggreg­ates, and views in order to eke out an accept­able level of per­form­ance. Moreover, Sybase IQ’s pat­en­ted data com­pres­sion provides a data ware­house or ana­lyt­ics envir­on­ment with a green, lean and effi­cient foot­print — fre­quently 70% smal­ler than the ori­gin­al input data.

Sybase IQ provides highly effi­cient and flex­ible scalab­il­ity of ana­lyt­ics and data ware­hous­ing environments:

  • You can start with one serv­er and add CPUs and memory as needed
  • You can add serv­ers and CPUs with little or no loss in scalability
  • You can add tera­bytes of disk space and Sybase IQ will man­age it efficiently
  • You can sup­port mul­tiple users per CPU.
  • With dynam­ic­ally con­fig­ur­able read and write nodes, Sybase IQ makes it easy to scale load­ing or query­ing or both, depend­ing on the spe­cif­ic needs of the environment.

    SOURCE: Sybase IQ Data Sheet

    No Comments