Too many lines in chart

[#1]

Hi all!

i have a 10 000 000 lines in csv:

Header 1	Header 2	Header 3	Header 4	Header 5
Date	Domain	Connections	Bytes in	Bytes out

I need to display  in table top 20  domain by Connections, top 20 Domain by Bytes in and Bytes out.

There is 128 GB RAM on my virtual machine.

If i use Dimension limits - the largest 20 i have the one problem- QV is failed becouse of it try to select top 20 in 10 000 000 domains.

I tryied  to use set analisys  in my Expression, but he anyway try to select top in 10 000 000.

Could you tell me what shoul i do for resolving this problem?

[Anonymous]

Kseniya,

It is better to calculate the top domains in the script, it will help with the front end performance.

Regards,

Michael

[#3]

i thought about it, but i have some filters like service: streaming media, peer-to-peer, http, File Sharing, Remote Access, IP Protocols, Web Browsing i t.e (~ 200 filters) and if i calculate the top domains in the script,i couldn't to filter it.

[Anonymous]

OK, what about this:

Create three rank fields in the data - RankConnections, RankBytesIn, and RankBytesOut - integer value for each domain.  Now, if you don't have selections, your top by RankConnections will be 1 to 20.  If you make selections, it could be anything (e.g. 120,500,666,... 5215045) - but still sortable without dynamic rank calculations.

Hope it works...