1 Server multiple instances or 1 instance or even Multiple servers?

I am a new DBA working for a charity and I have project to consolidate the multiple databases scattered across the organization.
I have identified the main databases that are mission critical to the organisation. We also have a virtualised infrastructure so plenty of VM's. We currently run SQL Server 2008R2.
Fundraising Database (off the shelf package - 30GB in size with about 100 users)
Fundraising Data warehouse (off the shelf package - 30GB in size with about 5 users)
SQL Reporting Services Database (1GB - reports used across the whole organisation so about 400 users)
Case Load Services Database (Developed in house - 4GB in size with about 300 users)
Finance Database (off the shelf package - 2GB in size with about 150 users)
GFI Exchange archive databases - (off the shelf database - 30GB in size with about 300 users)
Then we have about 10 other smaller databases for system monitoring etc... that are important but not mission critical.
Now the question I have is (bearing in mind limited budget and resources) is whether I should put all of these databases on 1 single instance on 1 single Windows Server 2008 VM. Or would
it be better to have them on a single box with separate named instances? Or possibly even separate VM Servers splitting out certain databases each?
My thoughts was to have 2 VM's. SQL01 for the missions critical databases. SQL02 for the other databases that are non mission critical. I was even toying with the idea
of separate named instances on SQL01 for security and the ability to manage services separately without affecting other databases? For example if we need to restart a service during working hours we would not have to take out the whole organisation but
only users on 1 database.
I know there are many factors that affect this decision but from your experiences and knowledge what is the general rule of thumb when redesigning the architecture for a SQL database landscape
like we have? Obviously each of the system requirements of the off the self packages say a single database ideally not contending with other database but I take it that its not really what happens in the real world as licensing SQL and hardware resources
would be significantly wasted that way.

Hi marko,
According to your description, personally, the type of your issue is set to discussion, more forum members can involve into the thread and help you out. Thanks for your understanding.
When you choose multiple SQL Server instances or not, you should consider the factors of CPU resources, memory, network access and I/O which would be affected performance and so on.
If applications do not create logins, do not reboot the server, do not drop or create databases. I recommend you using separate databases on one server. If you need to consider security settings, logins, linked servers and so on, you need
to install separate instances. There are details about multiple database instances, and multiple databases, you can review the following article.
http://www.sqlsolutions.com/articles/articles/Multiple_Databases_or_Multiple_Instances.htm
http://www.toadworld.com/platforms/sql-server/w/wiki/10388.multiple-database-instances.aspx
Hope it can help you.
Sofiya Li
Sofiya Li
TechNet Community Support

Similar Messages

How to serve multiple weblogic instances with a single apche server

Can any body let me know how to make use of an apache server to serve 2 weblogic
instances.
Thanks in advance
damodar

Yes, the documentation explains how to do this:
http://e-docs.bea.com/wls/docs61/adminguide/apache.html
Regards,
Eric
"Damodar" <[email protected]> wrote in message
news:3c23ccb9$[email protected]..
>
Can any body let me know how to make use of an apache server to serve 2weblogic
instances.
Thanks in advance
damodar

How to restore a database backup taken from a SQL Server 2012 instance to SQL Server 2014

I am trying to restore a database backup taken from a SQL Server 2012 instance to my new SQL Server 2014 instance. When I add the .bak through the Mgmt studio UI, it does not seem to find a backup set in the file so I cannot proceed (does not give
me any errors though either. I tried restoring through TSQL command but only got a database restore terminating abnormally error with no other details. How can I restore this database to SQL 2014?

Hello,
Please examine SQL Server Error Log and share with us any error you can find. The following article may locate the log:
http://msdn.microsoft.com/en-us/library/ms187109.aspx
Regards,
Alberto Morillo
SQLCoffee.com

HELP!!: Two server application instances could work in the same port?

Hi everybody!!!
i have two server application instances over the same JVM and both of them need to get requests from the same port, how can i get this??
I`m not interested on using RMI, and i wouldn`t like to use sockets, because i have already using XML-RPC, so if you can tell me how can i get this without using sockets nor RMI, please help me...
I don�t if there any way using threads??
thanks a lot,
JCPARRAM

Two sockets can't be bound to the same port on the same machine at the same time. This means that one application can't have two sockets listening on the same port and that two applications can't each have a socket listening on the same port (no matter what language they are written in, and, if they are in Java, no matter whether they are on the same JVM or not).
Your only solution is to make one of the server applications listen on a different port.

Open Connection to SQL Server 2000 Instances

Dear All:
I always got following error when connecting to a SQL Server 2000 instance:
"Can't open a socket on NAMNGO\RNDSQLSVR:1433. Check host and port number and
make sure the
security manager allows this connection. You can also try running the SocketApplet
(java.net.UnknownHostException:
NAMNGO\RNDSQLSVR)"
I am using Wel Logic 6.1 and SQL Server 2000. SQL Server 2000 has an instance
named RNDSQLSVR with default TCP port is 1433.
Is there anyone hit this problem, please let me know.
Please check the attached file for sample code.
Thanks,
Nam Ngo
[att1.html]

Hi Sree,
Thank you for your help. It worked.
Nam Ngo
"Sree Bodapati" <[email protected]> wrote:
Hi Nam,
The way you do this is by finding out on what port the instance of SQL
Server you want to connect is listening on. (Use the server network utility
that comes with SQL Server 2000 for this). Then just mention the
server=<machine name/ip> of the server this instance is running on. and
the
port as the port that is shown by the server network utility as port=<port
number>. Microsoft specifies NAMNGO\RNDSQLSVR:1433 as the convention
for
ODBC but , this driver is a type 4 driver and it needs only the machine
name
and port, it doesnt need any instance name.
(Each instance of SQL Server 2000 listens at a different port)
HTH
sree
"Nam Ngo" <[email protected]> wrote in message
news:[email protected]..
Dear All:
I always got following error when connecting to a SQL Server 2000instance:
"Can't open a socket on NAMNGO\RNDSQLSVR:1433. Check host and portnumber
and
make sure the
security manager allows this connection. You can also try running theSocketApplet
(java.net.UnknownHostException:
NAMNGO\RNDSQLSVR)"
I am using Wel Logic 6.1 and SQL Server 2000. SQL Server 2000 has aninstance
named RNDSQLSVR with default TCP port is 1433.
Is there anyone hit this problem, please let me know.
Please check the attached file for sample code.
Thanks,
Nam Ngo

Using a SQL Server database instance for repository

Hi,
My customer is using a non-default instance on SQL server as their database and want to place the repository here. There are no problems creating and connecting to the repository from the Designer. I simply define the server as "SERVER\INSTANCE" and I can connect OK.
The problem occurs in the Management Console, when I try to create an entry for this new repository. The Add process returns a Java error, telling me that the connection is being refused. I have checked, and re-checked, that all of the inputs are correct. I believe that this has something to do with the fact that the repository is an instance, rather than just the default database instance.
I have tried various differing ways of entering the server and database strings, with no positive result. I tried entering the following in the Server and Database entry fields:
SERVER\INSTANCE with DATABASENAME
SERVER
INSTANCE with DATABASENAME
SERVER/INSTANCE with DATABASENAME
INSTANCE@SERVER with DATABASENAME
SERVER.INSTANCE with DATABASENAME
SERVER with INSTANCE\DATABASENAME
SERVER with INSTANCE.DATABASENAME
All give an error - some return a "connection refused", some return a "server not found" error.
Any recommendation short of re-installing the database as a default instance would be appreciated.
DI and SQL Server reside on the same server running Windows 2003 Server R2 SE with SP2. Management Console runs in Apache Tomcat 5.5 on the same box.
Regards,
Derek.
Edited by: Derek Stobbart on Oct 1, 2008 5:14 PM

Hi Derek,
The Management Console does not support the \INSTANCE syntax -- the reason is that it uses JDBC to connect to repositories. When you add a SQL Server repository to the Management Console, note that it is asking you for hostname, port, and database name. You will need to provide it the port that the instance is running on (e.g. default=1433). You can get this from the person who administers the SQL Server instance.
Thanks,
~Scott

Universe on SQL Server names instance

Hi All,
I am very new to BO and like to get some information regarding the data sources that BO support.
I am using Business Obejcts Enterprise XI 2 release. And can I use SQL Server named instance as the data source. If so, how do I set the universe connection parameters. I select the DSN name and pass username and password. Ane when i create DSN, at the server name select SQLSERVER\INSTANCE NAME.
I appreciate if anyone can help me
palam01

Hi,
You can create connection using generic ODBC and odbc for sql server also.
For that first you have to create a DSN from the control panel-->data sources.
Create a DataSource using the database you want to use.
Now go to Universe>New>type the universe name
Click on new connection >Next>Select Microsoft>MS SQL Server 2000> ODBC Drivers>Next>Type(secured), name the connection.
Type in the user name and password and select the DSN name you created for sql server.
Click on test connection.
Click on next-->next and finish and OK.
Now the universe that you are creating will use this connection and you can also use this connection for creating other universes.
Hope this helps.
Regards,
gaurav

Can SQL server 2014 instance be introduced into SQL server 2012 clustering

Hello, everyone,
We have SQL server clustering set up in place (it is actually windows clustering), I am wondering if it is possible to introduce SQL server 2014 instance
into this clustering. Please share us your thoughts and experiences.
Thanks

if you are talking about SQL clustering -(FAILOVER CLUSTERED INSTANCES) , yes, you can do that.
you have windows cluster with nodea,nodeb....and you can install two different instances of failover clustered sql - sql 2012 and sql 2014.
Hope it Helps!!

Server and Instance

What is mean by 'server','instance' in MSSQLServer 2005
What is mean by 'server ','instance ' in mssqlserver 2008R2 ?
could anyone explain in simple way?
bhagee

Each SQL Server instance is an installed program which can have 0 to many databases in addition to the system databases.
SQL Server default instance (it can be only one) has the same name as the computer(server), example:
LONDONPROD
There can be 0 to many named instances on the same computer(server), examples:
LONDONPROD\SQL12
LONDONPROD\INVENTORY
LONDONPROD\SS14
LONDONPROD\BI
The versions of the instances maybe the same or different.
See computer(server) with two SQL Server instances (different versions):
Kalman Toth Database & OLAP Architect
SQL Server 2014 Database Design
New Book / Kindle: Beginner Database Design & SQL Programming Using Microsoft SQL Server 2014

Select data from multiple servers(instances)

Hi All,
Is possible that we can login to SQLPLUS of one server and be able
to select data from other servers. For MSSQL server we can use
opendatasource but I don't know if in ORACLE we can
get the data from other servers by login to SQLPLUS of one server.
Your help is greatly appreciated.
Thanks in advance for your response.
JP

Hi,
Please try this ...
DATA: BEGIN OF ITAB1 OCCURS 0.
        INCLUDE STRUCTURE A_TABLE.
DATA: END OF ITAB1.
DATA: BEGIN OF ITAB2 OCCURS 0.
        INCLUDE STRUCTURE B_TABLE.
DATA: END OF ITAB1.
DATA: BEGIN OF ITAB3 OCCURS 0.
        INCLUDE STRUCTURE C_TABLE.
DATA: END OF ITAB3.
SELECT *
INTO TABLE ITAB1
FROM A_TABLE
WHERE A_ID IN S_A_ID.
SELECT *
INTO TABLE ITAB2
FROM B_TABLE
WHERE A_ID IN S_A_ID.
SELECT *
INTO TABLE ITAB3
FROM C_TABLE
WHERE A_ID IN S_A_ID.
To download to PC or application server, please check this sample codes.
http://www.sapdevelopment.co.uk/file/file_downloadpc.htm
http://www.sapdevelopment.co.uk/file/file_downloadsap.htm
Regards,
Ferry Lianto
Please reward points if helpful.

Getting single "generic" WLS 4.5.1 app to serve multiple DNS names, including SSL

Hi everybody,
I am trying to figure out a strategy to enable a single web application
based on WLS 4.5.1 (SP13) running on Linux to serve multiple DNS
names, with some specific pages to be transmitted using SSL.
While the JSP pages differ in colouring and screen layout (but not in
logic),
the entire EJB business logic is identical for all DNS names. So are the
servlets.
The application data to be presented to the user can easily be mapped,
and presently (with the application responding to a single DNS name)
it is cached using a JNDI tree. To minimize coding effort, and
to continue utilizing the present caching concept, only expanding it to
support
multiple similar "virtual" applications, I would prefer to
continue using a single WLS instance when making the application "generic".
But as a consequence of this, I could only present a single SSL certificate
to the user if I had WLS answer internet HTTP requests directly.
What would you think should be done to get the browsers
"display the closed padlock while in the order form" for multiple DNS names
served by this same application instance?
One approach might be to put Apache in front of WLS, and configure it
to serve multiple virtual hosts (one for each DNS name to be served,
encoding the differences between the request URLs somehow in the ServerPath
directive, and providing a separate SSL certificate for each of those
virtual
hosts).
Does this make sense to you? Any experience in this area? Which Apache
and WL modules should I be using - which combination
of mod_proxy, mod_ssl (by Apache), mod_wl, mod_wl_ssl[_raven] ?
Any input would be greatly appreciated! Thanks in advance!
Ruben
[A minor add-on question is: Is the situation truly different with
WLS 6.0? I understand it supports virtual hosting on its own. Can multiple
(web) applications running under a single common WLS "umbrella" server
instance share the same application data? I would expect so - is this true?]

Ruben, I can't help you with the specifics of the configuration but your
general strategy seems correct to me. WLS does not handle multiple certs
and one must front it with a Web server if multiple certs are required.
Neil Smithline
WLS Security Team
BEA Systems
"Ruben-B Reincke" <[email protected]> wrote in message
news:[email protected]...
Hi everybody,
I am trying to figure out a strategy to enable a single web application
based on WLS 4.5.1 (SP13) running on Linux to serve multiple DNS
names, with some specific pages to be transmitted using SSL.
While the JSP pages differ in colouring and screen layout (but not in
logic),
the entire EJB business logic is identical for all DNS names. So are the
servlets.
The application data to be presented to the user can easily be mapped,
and presently (with the application responding to a single DNS name)
it is cached using a JNDI tree. To minimize coding effort, and
to continue utilizing the present caching concept, only expanding it to
support
multiple similar "virtual" applications, I would prefer to
continue using a single WLS instance when making the application"generic".
>
But as a consequence of this, I could only present a single SSLcertificate
to the user if I had WLS answer internet HTTP requests directly.
What would you think should be done to get the browsers
"display the closed padlock while in the order form" for multiple DNSnames
served by this same application instance?
One approach might be to put Apache in front of WLS, and configure it
to serve multiple virtual hosts (one for each DNS name to be served,
encoding the differences between the request URLs somehow in theServerPath
directive, and providing a separate SSL certificate for each of those
virtual
hosts).
Does this make sense to you? Any experience in this area? Which Apache
and WL modules should I be using - which combination
of mod_proxy, mod_ssl (by Apache), mod_wl, mod_wl_ssl[_raven] ?
Any input would be greatly appreciated! Thanks in advance!
Ruben
[A minor add-on question is: Is the situation truly different with
WLS 6.0? I understand it supports virtual hosting on its own. Can multiple
(web) applications running under a single common WLS "umbrella" server
instance share the same application data? I would expect so - is this
true?

RESTful Authentication to Service Bus for Windows Server - Multiple Scopes?

We currently have a three-node farm set up and have tested authentication and sending/receiving messages to each node individually, with which we've had no issues. However, we use NetScaler as a network load balancer, and once we configured NetScaler
in front of the three nodes we've encountered issues with RESTful authentication using OAuth.
The issue appears to be that when specifying a single scope (ex. for "server1/namespace"), if NetScaler routes the request to server1 it will provide a token back but if routed to server2 or server3 they fail with the following warning event logged:
Security token service finished processing a request with failure. TrackingId: (guid + server receiving request), Absolute Uri: https://(load balancer uri):9355/namespace/$STS/OAuth/, Message: System.UnauthorizedAccessException: Invalid user.
at Microsoft.Cloud.ServiceBus.Security.OAuthAuthenticationRequest.Parse(Message message, Int32 maxArrayLength)
at Microsoft.Cloud.ServiceBus.Security.SecurityTokenServiceProtocolHandler.GetTokenAsyncResult.ParseAuthenticationRequest(Message requestMessage, Int32 maxArrayLength)
at Microsoft.Cloud.ServiceBus.Security.SecurityTokenServiceProtocolHandler.GetTokenAsyncResult.<GetAsyncSteps>d__2c.MoveNext()
at Microsoft.ServiceBus.Messaging.IteratorAsyncResult`1.EnumerateSteps(CurrentThreadType state)
at Microsoft.ServiceBus.Messaging.IteratorAsyncResult`1.Start()
After reading the OAuth specifications, I tried changing the scope to be a space-delimited list of all three nodes (ex: "server1/namespace server2/namespace server3/namespace"), but this failed with a similar event being logged, though this one
specifying that the namespace could not be resolved for the scope provided.
When the server receiving the request matches the server in the scope and a token is provided, sending/receiving appears to function as normal across all three nodes. Is there a different way to specify multiple servers in the scope or otherwise solve/mitigate
this issue? I've read a couple threads regarding the support and use of NLB for load-balancing -- I think our infrastructure/network teams would prefer using NetScaler if possible, though any suggestions are welcome. Thanks!

Hi,
Since this issue is more related to Microsoft Azure Pack. I move it to the Azure Pack forum. It is appropriate and more experts will assist you.
Regards,
Jambor
We are trying to better understand customer views on social support experience, so your participation in this interview project would be greatly appreciated if you have time. Thanks for helping make community forums a great place.
Click
HERE to participate the survey.

Upload client data to task server multiple times in same day

I am just getting started setting up the task server piece of ARD. My task server is an Xserve(PowerPC 10.4) and my client computers are mostly eMacs running 10.4. My client computers are all running Deep Freeze by Faronics which causes the hard drive to be restored to it's original state after a restart.
My question is - where does the report information about the client computer get stored before it is uploaded to the task server. Is it cached on the local client machine? My concern is that if the client computers don't upload their data until the end of the day and a student restarts a computer during the day we won't have a whole day of data recorded for that machine.
If this is the case, is there any way to upload the data to the task server multiple times in the same day?
Thanks!

I am collecting User History and Application Usage. Where is the cache stored - on the local client machine or on the task server? If it's stored on the local machine and one of the students restarts the computer the data will be gone because of the Deep Freeze program. That's why I'm trying to figure out a way to upload the data multiple times throughout the day.
I'm hoping that the data is cached on the task server so that I don't have to try to upload multiple times.

Subversion serving multiple repositories with systemd startup

Hello.
I'm trying to set up a simple svn server, serving multiple repositories.
I was able to correctly set up single repository using guide from archwiki, now I'm trying to serve multiple repositories.
I've modified the /etc/conf.d/svnserve file to point to parent directory of those repositories, so instead of
-r /path/to/repos/repo1
it reads
-r /path/to/repos
However when I try to start subversion I get the following output
? svnserve.service
Loaded: error (Reason: Is a directory)
Active: inactive (dead) since Thu 2014-11-20 21:17:01 CET; 18min ago
Main PID: 24680 (code=killed, signal=TERM)
Now, when I try to start svnserve from shell, using this command
svnserve -d -r /path/to/repos
it starts normally and I am able to access all repositories. Adding and removing -d in /etc/conf.d/svnserve has no effect on error message.
Any thoughts?

The problem has been solved.
The error message:
Loaded: error (Reason: Is a directory)
came from systemctl, specifically from /etc/systemd/system/svnserve.service being a directory.
The fix was to remove that directory
sudo rm -rf /etc/systemd/system/svnserve.service
and enabling/starting svnserve
systemctl enable svnserve
systemctl start svnserve
Hope this will help someone having similar problem.

How to place multiple servers in Maintenance Mode in SCOM 2012

Hi All,
I am trying to place multiple servers in Maintenance Mode in SCOM 2012 but i could not find any PoSh script or any other solution to do it.
Could someone help me please?
Regards, Srini

Hi,
Please find the script.
need to have access and you can provide list of servers in text file.
param([int32]$TimeMin, [string]$Reason, [string]$Comment)
$api = new-object -comObject 'MOM.ScriptAPI'
Import-Module operationsmanager
New-SCOMManagementGroupConnection
$Servers = Get-Content "ServerList.txt"
$Time = (Get-Date).Addminutes($TimeMin)
Foreach ($Server in $Servers)
#Get Computer instance
$ComputerClass = Get-SCOMClass -Name Microsoft.Windows.Computer
$ComputerClassInstance = Get-SCOMClassInstance -Class $ComputerClass | Where {$_.DisplayName -eq $Server}
If ($ComputerClassInstance.InMaintenanceMode -eq $true)
#Write-Host $Server " is in maintenance mode"
Write-Host $Server" already under Maintenance Mode or not accessible by SCOM, skipped from script execution" -foregroundcolor "red"
$api.LogScriptEvent('MaintenanceModeScript.ps1', 201, 1, "$Server already found under MM, skipped from script execution")
Else
#Write-Host $Server " is NOT in maintenance mode, attempting to put in MM......."
If ($ComputerClassInstance -ne $Null)
$HealthServiceWatcherClass = Get-SCOMClass -name:Microsoft.SystemCenter.HealthServiceWatcher
#Get Health Service Watcher Class instance of the server
$HSWClass = Get-SCOMClass -Name Microsoft.SystemCenter.HealthServiceWatcher
$HSWClassIns = Get-SCOMClassInstance -Class $HSWClass | Where {$_.DisplayName -eq $Server}
#Starting the maintenance mode
Start-SCOMMaintenanceMode -Instance $HSWClassIns -EndTime $Time -Reason $Reason -Comment
$Comment
Start-SCOMMaintenanceMode -Instance $ComputerClassInstance -EndTime $Time -Reason $Reason
-Comment $Comment
Write-Host "Health Service Watcher and Agent server "$Server " kept in maintenance
mode for $TimeMin minutes" -foregroundcolor "green"
$api.LogScriptEvent('MaintenanceModeScript.ps1', 200, 0, "$Server kept in maintenance
mode for $TimeMin minutes")
Else
Write-Host $Server" not found in Domain" -foregroundcolor "red"
$api.LogScriptEvent('MaintenanceModeScript.ps1', 202, 1, "$Server not found in domain")
Regards
sridhar v

Separating a database onto multiple servers

We are working on a product that uses a set of SQL Server databases that are currently all in one instance and on one server. Now that we are growing, we need more processing power/space. So we want to move the databases onto multiple servers.
So we would not need to "shard", we created separate databases (with the same schema) for each company that purchased the product. We could readily put different company databases onto the different servers.
The question is this, how to we handle the routing to the different servers?
If we would have "sharded" the data in one database with different keys, SQL Server tools would have handled the routing for us. Because we actually separated the databases, do we have to do the routing manually? (That is, do we need to figure
out that when company A logs in that they need to go to database server aa.aa.aa while compay B needs to go to server bb.bb.bb?) Or is there a product that can help us with the routing?
Thanks!
www.insteptech.com ;
msmvps.com/blogs/deborahk
We are volunteers and ask only that if we are able to help you, that you mark our reply as your answer. THANKS!

Hello,
Based on your description, it seems that you want to copy all schema from a "Shared" database into other servers and separate data into the tables based on the company name.
If I understand correctly, you can try to use the
SQL Server Import and Export Data Wizard to copy data from table or view to destination server and specify query to restrict data for copy. Or you can use linked server and
SELECT INTO statement with condition to move the data.
As per my understanding about the "Handling routing", balancing capability. In order to meet your requirement, you have to load balance through your application's design.
Regards,
Fanny Liu
If you have any feedback on our support, please click here.
Fanny Liu
TechNet Community Support

1 Server multiple instances or 1 instance or even Multiple servers?

Similar Messages

Maybe you are looking for