Hi Jayan,
Here are a small change to Browse.java that provided a noticable speed
improvement for >180,000 records in the browse pages.
1. At [EMAIL PROTECTED] we do not have any duplicate entries in the
ItemsByAuthor, ItemsBySubjects, ItemsByTitle, ItemsByDate and we can
change the sql code generated by createSqlInternal in Browse.java from
select * from (SELECT DISTINCT * from ItemsByTitle ) distinct_view
order by sort_title{2}, item_id{2} LIMIT {3}
to
select * from ItemsByTitle order by sort_title{2}, item_id{2} LIMIT {3}
This gives a significant performance increase to the browse pages.This
can only be used if you do not have duplicate entries. To check if you
have any duplicates in your browse tables perform the following
dspace=# select count(*) from (SELECT DISTINCT * from ItemsByTitle )
distinct_view
dspace-# ;
count
-------
10063
(1 row)
dspace=# select count(*) from ItemsByTitle;
count
-------
10063
(1 rows)
Repeat for ItemsByAuthor, ItemsBySubjects, ItemsByDate
If you are confident that no duplicates exist then change Browse.java
createSqlInternal(BrowseScope scope,String itemValue, boolean isCount)
from
sqlb.append(" from (SELECT DISTINCT * ");
sqlb.append(" from ");
sqlb.append(tablename);
sqlb.append(" ) distinct_view");
to
//sqlb.append(" from (SELECT DISTINCT * ");
sqlb.append(" from ");
sqlb.append(tablename);
//sqlb.append(" ) distinct_view");
Hope this helps
James
Jayan Chirayath Kurian wrote:
> Hi! James,
>
> Search and browse feature appears to be a bit slow. Indexing was fine by
> allocating 800m to the dsrun batch utility. Please suggest since we hope
> that the system will be used by almost 10,000 users (students + staff)
> at our campus.
>
> Thanks,
> Jayan
>
> -----Original Message-----
> From: James Dickson [mailto:[EMAIL PROTECTED]
> Sent: Tuesday, July 24, 2007 4:25 PM
> To: Jayan Chirayath Kurian
> Cc: DSpace Tech
> Subject: Re: [Dspace-tech] DSpace optimization
>
> Hi Jayan,
> What part of dspace are you having difficulty with? Browse, Search,
> Indexing.. For indexing we have implemented a batch indexing process
> that is not so memory intensive as the existing one. There are a few
> tweaks that can be performed too speed up the browsing. Unfortunately,
> throwing more memory will increase the number of concurrent user you can
> serve, but will not really have much effect on performance.
>
> James
>
> Jayan Chirayath Kurian wrote:
>> Hi!
>>
>>
>>
>> Can anyone suggest how to allocate more memory to Tomcat and
>> postgreSQL for a server with 1 GB ram, 300 GB hard disk and 170,000
>> records? Will allocating memory improve the client access speed?
>>
>>
>>
>> Thanks,
>>
>> Jayan
>>
>>
> ------------------------------------------------------------------------
>>
> ------------------------------------------------------------------------
> -
>> This SF.net email is sponsored by: Splunk Inc.
>> Still grepping through log files to find problems? Stop.
>> Now Search log events and configuration files using AJAX and a
> browser.
>> Download your FREE copy of Splunk now >> http://get.splunk.com/
>>
> ------------------------------------------------------------------------
>> _______________________________________________
>> DSpace-tech mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/dspace-tech
>>
>
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
DSpace-tech mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dspace-tech
