topic Large Data Model and Performance. in QlikView https://community.qlik.com/t5/QlikView/Large-Data-Model-and-Performance/m-p/397432#M558358 <HTML><HEAD></HEAD><BODY><P>I have an application in development which is going to be accessing a fairly large number of transactional data records.&nbsp; 12 million per quarter and I hope to include a full year of data so that means about 48 Million.&nbsp;&nbsp; (I have already aggregated as much as I can).</P><P></P><P>It is possible to divide my Fact table into to discrete pieces about 2/3 in one and 1/3 in the other.&nbsp; Most of the charts do not use both so there could be some savings there however a link table would be needed so not sure how much.</P><P></P><P>My question is What is the point where performance will be an issue?&nbsp; Have I already exceeded this at 20+ million rows?&nbsp; Testing with 8 million row tables already has its challenges.&nbsp; Requires server strength to keep out of memory issues from appearing.&nbsp; I do expect that the application will be full featured with many charts and tables.&nbsp;&nbsp; </P><P></P><P>Is it worth the separating of the two groupings?&nbsp; Will such take advantage of multi-threading or multi-processor capabilities?&nbsp; </P></BODY></HTML> Wed, 22 May 2013 18:08:15 GMT 2013-05-22T18:08:15Z Large Data Model and Performance. https://community.qlik.com/t5/QlikView/Large-Data-Model-and-Performance/m-p/397432#M558358 <HTML><HEAD></HEAD><BODY><P>I have an application in development which is going to be accessing a fairly large number of transactional data records.&nbsp; 12 million per quarter and I hope to include a full year of data so that means about 48 Million.&nbsp;&nbsp; (I have already aggregated as much as I can).</P><P></P><P>It is possible to divide my Fact table into to discrete pieces about 2/3 in one and 1/3 in the other.&nbsp; Most of the charts do not use both so there could be some savings there however a link table would be needed so not sure how much.</P><P></P><P>My question is What is the point where performance will be an issue?&nbsp; Have I already exceeded this at 20+ million rows?&nbsp; Testing with 8 million row tables already has its challenges.&nbsp; Requires server strength to keep out of memory issues from appearing.&nbsp; I do expect that the application will be full featured with many charts and tables.&nbsp;&nbsp; </P><P></P><P>Is it worth the separating of the two groupings?&nbsp; Will such take advantage of multi-threading or multi-processor capabilities?&nbsp; </P></BODY></HTML> Wed, 22 May 2013 18:08:15 GMT https://community.qlik.com/t5/QlikView/Large-Data-Model-and-Performance/m-p/397432#M558358 2013-05-22T18:08:15Z