Multiple-Instance Cluster

Situation: Two nodes, two nodes are active, two virtual
servers, two separate applications. One active and one
wait on each. Active on each fails over to wait on each
other server so that I should only lose one application at
a time.
Question: Two SQL instances on each node/virtual server?
Each SQL instance has its own IP address? The active and
wait have their own IP addresses?
Ok, so you want to run a 2 node cluster, with both nodes being active at the
same time. Two databases for two applications.
You will only need 1 instance of SQL on a node, unless the application
requires something different.
Each Node will have an IP addr, the cluster will have an addr, and each
instance (1 on each server) will have an IP addr.
With all the said, I never recommend an Active/Active cluster. SQL optimizes
memory. When you have a failure, both instances of SQL and its IP addr will
then be on one box. This will change the way SQL can use and optimize
memory. This process takes time, during that time, the system is much
slower. Performance on most Active/Active SQL Clusters suffer during a
failed state.
If at all possible, run both apps on the same node, at the same time. Making
sure of course, that each system has ample resources.
Hope this helps.
Cheers,
Rod
"Lucian" <anonymous@.discussions.microsoft.com> wrote in message
news:1d2b101c42329$d3d501c0$a401280a@.phx.gbl...
> Situation: Two nodes, two nodes are active, two virtual
> servers, two separate applications. One active and one
> wait on each. Active on each fails over to wait on each
> other server so that I should only lose one application at
> a time.
> Question: Two SQL instances on each node/virtual server?
> Each SQL instance has its own IP address? The active and
> wait have their own IP addresses?
|||With SQL Server 2000 you can install up to 16 instances (1 default and 15
named or 16 named). The number of nodes that can be a possible owner for
each instance depends on your edition/version of Windows. If you have a
typical 2-node cluster, each node can own 0,1, or many of the installed
instances.
Whether it makes sense to install multiple instances (something that used
to be called active/active when you could only have two instances) depends
totally on your environment - your hardware, your application, your
business needs, etc.
If you want to know more about clustering, check out BOL or the MSPress
book on High Availability.
Cindy Gross, MCDBA, MCSE
http://cindygross.tripod.com
This posting is provided "AS IS" with no warranties, and confers no rights.

Multiple updates and Identity fields

I have a table used by multiple applications. One column is an Identify field and is also used as a Primary key. What is\are the best practices to use get the identity value returned after an INSERT made by my code.. I'm worried that if someone does an INSERT into the same table a "zillionth" of a second later than I did, that I could get their Identity value.

TIA,

Barkingdog

Are you using SQL2k5 ? THen you can use the new OUTPUT clause. Otherwise you should have a look on SCOPE_IDENTITY() which should fit your needs. But the new OUTPUT function should be more straight forward in your case.

HTH, JEns K. Suessmeyer.

http://www.sqlserver2005.de|||

Jens,

I looked up OUTPUT in sql 2k5 BOL: Here's an example that I found:

>>>
Copy Code
USE AdventureWorks;
GO
DECLARE @.MyTableVar table( ScrapReasonID smallint,
Name varchar(50),
ModifiedDate datetime);
INSERT Production.ScrapReason
OUTPUT INSERTED.ScrapReasonID, INSERTED.Name, INSERTED.ModifiedDate
INTO @.MyTableVar
VALUES (N'Operator error', GETDATE());

--1. Display the result set of the table variable.
SELECT ScrapReasonID, Name, ModifiedDate FROM @.MyTableVar;

--2. Display the result set of the table.
SELECT ScrapReasonID, Name, ModifiedDate
FROM Production.ScrapReason;
GO
>>>>

So with OUTPUT the ScrapReasonID from @.MyTableVar will hold the identity of the row just inserted. Is this correct?

What is the difference in output between queries 1 and 2? Shouldn't they be the same?

TIA,

Barkingdog

Multiple Sort Order Specification in EDB database

Hi,

I am using the EDB as a database in smart phone applications.I can able to mount the database , create the tables , write the records into the tables and read from the table with single sort order specification.But if i am using more than one sort order specification seek database is throwing an error message "The drive cannot locate a specific area or track on the disk".

I created one table named as Icon with more than one sort order specification.In this table i am giving pageid and iconid as sort order specifications. While saving the data into the database i am opening the database with pageid sort order and storing into the table.while reading from the database i am opening the database with iconid as sortorder.So here while seeking the database the above error is displaying.And i tried to read the data from the table by using pagid as sortorder speicification.It is working fine.I am applying the 3 sort order specifications for another table named as contentTable. but it is working fine with 3 sort order specifications.I had done the samething for icon table .But no use.Please help me regarding this.

Thanks in Advance,

Thanks & Regards,

Prasanna Kumar

Moving to Sql Server compact edition forum where it has got better chances of being answered.

-Thanks,

Mohit