Hi,
   I have Index with size 1GB. Its each documents consist five Fields which
are use for search.For single result it take 30 to 40 milliseconds.I want to
reduce this time.How can I do this? is search performance depends on a Index
size? What is maximum capacity to add Documents  in Index?
  

--
View this message in context: http://lucene.472066.n3.nabble.com/Lucene-Index-Size-and-Performance-tp3970551.html
Sent from the Lucene - Java Developer mailing list archive at Nabble.com.

---------------------------------------------------------------------

what's the hardware configuration of your machines?
if you have enough RAM, you could use RAMDirectory.

On Tue, May 8, 2012 at 2:52 PM, parkhekishor <[EMAIL PROTECTED]> wrote:
> Hi,
>   I have Index with size 1GB. Its each documents consist five Fields which
> are use for search.For single result it take 30 to 40 milliseconds.I want to
> reduce this time.How can I do this? is search performance depends on a Index
> size? What is maximum capacity to add Documents  in Index?
>
>
> --
> View this message in context: http://lucene.472066.n3.nabble.com/Lucene-Index-Size-and-Performance-tp3970551.html
> Sent from the Lucene - Java Developer mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>

---------------------------------------------------------------------

Hi,
    My machine hardware configuration is 25G RAM and Intel Xeon 2.25GHz
processor.

--
View this message in context: http://lucene.472066.n3.nabble.com/Lucene-Index-Size-and-Performance-tp3970551p3983641.html
Sent from the Lucene - Java Developer mailing list archive at Nabble.com.

---------------------------------------------------------------------

Hi,
just for stats. Findfiles.net will be switched off by the end of this month.

FindFiles.net has  800.000.000 Documents (800 millions). 1KB per
Document. One Document has 11 mandatory and 10 optional fields.

Slave Indexes:
OS: Ubuntu 11.04
All 4 slave indexes 100 GB each are run on a 4 SSD 120GB in RAID 0, 24
GB RAM, Hexacore CPU  Intel i7. The 4 SSDs in RAID 0 reach about 600MB
per Second read performance.

Average Server load (top) 0.73, 0.89, 1.28
Daily average index response Time is: 0.73 seconds
Daily average requests about 12.000

Master Indexes:
OS: Ubuntu 11.04
2x 2TB SATA2, No RAID, 16GB RAM, Hexacore Intel CPU. Complete Reindex
time: 55 hours. Reindex based on SimplePostTool backed by a
multithreaded support which implements the producer-consumer pattern. 1
Producer Thread and 12 consumer Threads. Sing SimplePostTool

Hope that helps some of you for planning your own indexes.

Frontend is based on Wicket 1.4.19.
Backend is based on Solr 3.4.0

Kind Regards

Greg

On 05/14/2012 01:51 PM, parkhekishor wrote:
> Hi,
>      My machine hardware configuration is 25G RAM and Intel Xeon 2.25GHz
> processor.
>
> --
> View this message in context:http://lucene.472066.n3.nabble.com/Lucene-Index-Size-and-Performance-tp3970551p3983641.html
> Sent from the Lucene - Java Developer mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:[EMAIL PROTECTED]
> For additional commands, e-mail:[EMAIL PROTECTED]
>

---------------------------------------------------------------------