Does clustering improve performance for single user
We have a large number of records ( 300 million) loaded into a Qlikview document on a server with 256 GB RAM and 32 cores.
The graphs are very simple and there is no scope of simplifying them further.
The file performs fairly fast on this server, but we want to know whether it can be made faster if we cluster multiple servers of the same configuration. I know that clustering would help balance multiple users accessing the file, but would it create any difference for a single user. Another way of asking this question is - can clustering share the CPU/RAM of servers for one open instance of a document ?
If it cannot, then is it possible to distribute the load of different sheets in same document across different servers in a cluster.
Re: Does clustering improve performance for single user
You can cluster QlikView Components.
Qlikview Server and Qlikview webserver on one machine and QlikView publisher on another.
By this you can distribute tasks. Server 2 will be responsible for reloads and schedules where as server 1 will just serve user request.
Generally in a clustered environment Qlikview will share one complete instance of document to respective servers. and if the load on server 1 is more then it will redirect another user in a queue to server 2.
But I don't think it can share resources for the single instance on multiple servers.