By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
443,934 Members | 1,527 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 443,934 IT Pros & Developers. It's quick & easy.

Server configuration for MSSQL 2000 and MSSQL 2005

P: n/a
Does enabling/disabling Data Execution Prevention have a performance
impact on SQL 2000 or SQL 2005?
For SQL best performance - how should I configure for:
Processor Scheduling:
Programs or Background services

Memory Usage:
Programs or System Cache

Sep 5 '06 #1
Share this Question
Share on Google+
9 Replies


P: n/a
Hi

These setting will depend on what your system is doing. If you are solely
using it for SQL Server then there is little point in saving a large amount
of memory or processor for other tasks other than the OS, or if you have DTS
packages or other batch processes you would need to watch them. The exact
settings would need to be gained iteratively and monitored. Using the boost
priority for background services should have little effect if you are solely
using the system for SQL server, if it is a shared system then you could set
it, although it may have an impact when you need to access the system
quickly in an emergency. It would not only be SQL Server that gains from
setting background service priority. Depending on how many CPUs you have,
may allow you to set processor affinity, again this will depend on what your
system is doing and you could see an overall improvement in performance if
not all the CPUs are being used by SQL Server.

You may want to check out the articles and check lists on
http://www.sql-server-performance.co...cles_audit.asp

John

"dunleav1" <jm*@dunleavyenterprises.comwrote in message
news:11**********************@m79g2000cwm.googlegr oups.com...
Does enabling/disabling Data Execution Prevention have a performance
impact on SQL 2000 or SQL 2005?
For SQL best performance - how should I configure for:
Processor Scheduling:
Programs or Background services

Memory Usage:
Programs or System Cache

Sep 6 '06 #2

P: n/a
Thanks I'll check out the link.

The server sole use is sql server database.
There are no dts packages just sql statements executed via jdbc.

Sep 7 '06 #3

P: n/a
Hi

If that is the case you should be able to allocate most of the resources not
used by the OS to SQL Server.

John

"dunleav1" <jm*@dunleavyenterprises.comwrote in message
news:11*********************@h48g2000cwc.googlegro ups.com...
Thanks I'll check out the link.

The server sole use is sql server database.
There are no dts packages just sql statements executed via jdbc.

Sep 7 '06 #4

P: n/a
I have a process that I loads about 1000000 records into various tables
via jdbc.
With Sql server it takes about 35 minutes. If I use jtds or microsoft's
jdbc driver.
With Oracle it takes 5 minutes.

I have Oracle and Sql server configured on the same server. I stop the
other db vendor services when I run the process.Sql server is in simple
recovery mode and oracle in noarchivelog mode. They share the same raid
array configuration - raid 1/0.

Can Oracle's jdbc driver be that much better than Microsoft?

Sep 8 '06 #5

P: n/a
Hi

There can be many reasons for this to be slow not just the JDBC driver, you
may have contention for the discs with the different types of I/O and you
may have other poor configuration settings such as a poor choice of database
growth size, isolation level, or the code may not be the most optimal. You
don't say what the process is that loads this data, but look at using
something like BULK INSERT or possibly OPENROWSET instead of one insert
statement per row.

John

"dunleav1" <jm*@dunleavyenterprises.comwrote in message
news:11**********************@d34g2000cwd.googlegr oups.com...
>I have a process that I loads about 1000000 records into various tables
via jdbc.
With Sql server it takes about 35 minutes. If I use jtds or microsoft's
jdbc driver.
With Oracle it takes 5 minutes.

I have Oracle and Sql server configured on the same server. I stop the
other db vendor services when I run the process.Sql server is in simple
recovery mode and oracle in noarchivelog mode. They share the same raid
array configuration - raid 1/0.

Can Oracle's jdbc driver be that much better than Microsoft?

Sep 8 '06 #6

P: n/a

John Bell wrote:
Hi

There can be many reasons for this to be slow not just the JDBC driver, you
may have contention for the discs with the different types of I/O and you
may have other poor configuration settings such as a poor choice of database
growth size, isolation level, or the code may not be the most optimal. You
don't say what the process is that loads this data, but look at using
something like BULK INSERT or possibly OPENROWSET instead of one insert
statement per row.

John
I've tried four different vendor jdbc drivers - ms, jtds, jnetdirect,
jturbo.
jtds is the fastest.

I have a raid 1/0 disk array - the entire db is loaded on the array
along with the os. Identical configuration as Oracle. The average disk
queue is 30% for sqlserver regardless fo driver. Network bandwidth is
at 15% on the nic.
The database is presized so there is no growoth occuring on the log or
application db, or tempdb.

The process that loads the data is a process written in java - it
basically does inserts as batch - as a stored procedure.

>
"dunleav1" <jm*@dunleavyenterprises.comwrote in message
news:11**********************@d34g2000cwd.googlegr oups.com...
I have a process that I loads about 1000000 records into various tables
via jdbc.
With Sql server it takes about 35 minutes. If I use jtds or microsoft's
jdbc driver.
With Oracle it takes 5 minutes.

I have Oracle and Sql server configured on the same server. I stop the
other db vendor services when I run the process.Sql server is in simple
recovery mode and oracle in noarchivelog mode. They share the same raid
array configuration - raid 1/0.

Can Oracle's jdbc driver be that much better than Microsoft?
Sep 12 '06 #7

P: n/a
dunleav1 wrote:
I have a raid 1/0 disk array - the entire db is loaded on the array
along with the os. Identical configuration as Oracle.
Ideally, you want OS, db, logs, temp on separate physical drives, right?
Sep 13 '06 #8

P: n/a

Ed Murphy wrote:
dunleav1 wrote:
I have a raid 1/0 disk array - the entire db is loaded on the array
along with the os. Identical configuration as Oracle.

Ideally, you want OS, db, logs, temp on separate physical drives, right?
There is only one disk controller.
There are 6 disks configured within the array. The array is configured
RAID 1/0 (striped/mirror).

By sepearting everything on it's own disk you are "manually" striping
the OS, db, logs, temp from each other. Hardware striping is much
faster.

Sep 13 '06 #9

P: n/a

Ed Murphy wrote:
dunleav1 wrote:
I have a raid 1/0 disk array - the entire db is loaded on the array
along with the os. Identical configuration as Oracle.

Ideally, you want OS, db, logs, temp on separate physical drives, right?
There is only one disk controller.
There are 6 disks configured within the array. The array is configured
RAID 1/0 (striped/mirror).

By sepearting everything on it's own disk you are "manually" striping
the OS, db, logs, temp from each other. Hardware striping is much
faster.

Sep 13 '06 #10

This discussion thread is closed

Replies have been disabled for this discussion.