lucenenet-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Shad Storhaug <s...@shadstorhaug.com>
Subject RE: Hi,everyone!
Date Sat, 23 Jun 2018 08:50:05 GMT
Xiaokang,

First of all, make sure you are checking these numbers in Release mode (with optimizations
enabled).

Also, with that many returned records, the immediate thing that jumps to mind is whether there
is any kind of paging associated with the results as it would not be practical for a user
to view 30,000 documents at once based on a single action, anyway. So try to rule out if there
is an issue with optimization of the query first.

The best way to determine if this is a particular optimization problem with Lucene.Net or
if it is what is to be expected of Lucene's design is to port the example to Java in order
to compare the difference. If you could do it that would be great, otherwise please provide
a standalone example (unit test or console app, for example) that replicates the behavior
so we can make the comparison. If there is a large difference in performance, we can then
use the comparison to identify the bottleneck.

Thanks,
Shad Storhaug (NightOwl888)


-----Original Message-----
From: 小康 [mailto:xiaokang@cnblogs.com] 
Sent: Friday, June 22, 2018 11:05 AM
To: user@lucenenet.apache.org
Subject: Hi,everyone!

Hi, every one!
I have some doubts.
Is  my application's performance  good?
It's QPS is only 3~4.
Here is some information.

CPU : 2 processer Intel(R) Xeon(R) CPU E5-2682 v4 @ 2.50GHz

RAM : 4 G

IndexSize: 60G

I use booleanquery (or , high freq term ), single indexsearch and
FSDirectory.

The quest will return 30,000 documents.
Mime
View raw message