July 5, 2008 at 12:28 pm
hello all.
im planning to change sql server port on my cluster sql server 2005.
i have done some search on the net to take every thing into considertaion, any how i have read some thing about installing sqlbrowser on the second node, but i couldnt find out how to do it?
and if i didnt installe it on the seconde node what will happen?
..>>..
MobashA
July 6, 2008 at 10:53 am
Yuo already know the BROKER is not CLUSTER Aware, so you have to isntall it on both the nodes.
you Need to Broker ofcpourse for the REMOTE clients to Communicate with the SERVER/NODE in case of a Failover etc. Broker, keep track of all the Ports and Named Pipes etc..
Maninder
www.dbanation.com
July 6, 2008 at 12:26 pm
Be very careful changing the port on a cluster. Make sure that the port is indeed available. If it is not and you assign the port number of a port that is used the sql instance will not come online and when you attempt to change the port back to a unused port or dynamic port and bring the instance online it will fail again. When you look at configuration manager you will find that the bad port is there again.
The problem is with the cluster service registry replication. If you get in this state the incorrect entry will get replicated back to the node no matter how many times you change it or even changing the registry directly.
If you get into this state do the following:
1. Download the Cluster Recovery Utility from Microsoft
2. Log into the node that currently owns the failed resource.
3. Change the following registry key to a good port or the one that worked prior to the change (replace the X with the correct entry for your instance):
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Microsoft SQL Server\MSSQL.X\MSSQLServer\SuperSocketNetLib\Tcp\IPAll
4. Checkpoint the Cluster using the Utility.
5. You should now be able to bring the SQL Server resource online.
6. Repeat this process for each node in the cluster.
For the SQL Browser it will need to be installed on the second node
Gary L. Fry
Senior Database Administrator III
MCTS, MCITP SQL Server 2008 Database Administration
Colonial Life/UNUM
July 6, 2008 at 10:32 pm
It is important to check that the port you select is not currently in use, and not registered (Google TCP reserved ports). As long you do not have many (or any!) other network applications on the cluster you should be OK.
Also, try to pick a range of say 5 numbers, in case you would like to install additional instances in the future.
July 6, 2008 at 11:58 pm
thanks, all this is very help full, but how can i install SQLBrowser??
i could find nothing on the net.
..>>..
MobashA
July 7, 2008 at 1:53 am
Are you sure it isn't already installed?
Have a look in the services applet and see if you can find a service called "SQL Server Browser".
July 7, 2008 at 4:30 am
The standard SQL Server installation will install the SQL Browser service on both nodes. (Note: this is not a cluster aware 'service'.)
To check if it is installed & running, perform this on all nodes: Start -> run -> services.msc. Check if there is a service named 'SQL Server Browser'
Alternatively, you can check the status on both nodes from SQL Server Configuration Manager.
July 7, 2008 at 5:15 am
am i missing some thing here, how the installation will install it on both nodes and the service is not a cluster aware??
The standard SQL Server installation will install the SQL Browser service on both nodes. (Note: this is not a cluster aware 'service'.)
..>>..
MobashA
July 7, 2008 at 6:08 pm
Just because the service is not cluster aware, does not mean the installation isn't 😛
SQL Browser is installed on all nodes when you install the database engine.
Have you checked if it is installed / running on both nodes, per my last post?
July 9, 2008 at 10:04 pm
When you install on a cluster and you indicated that you are installing a failover clustered instance the installation process will install the necessary components on all members (servers) in the cluster.
It does this so that when a instance fails overs to another server it has all the necessary components to run. This includes cluster aware and non cluster aware components such as sql browser, sql native client, and other dll's. The only thing it does not install is the workstation components. Those you will need to install separately on all servers in the cluster that you intend on running the SSMS on.
Hope this answers you question. If not let me know and I will try to explain it further.
Gary L. Fry
Senior Database Administrator III
MCTS, MCITP SQL Server 2008 Database Administration
Colonial Life/UNUM
July 10, 2008 at 1:01 am
ur answer was v.clear.
so i dont need to install SQLbrowser right.
but can u explain what dose it mean cluster aware and not cluster aware
..>>..
MobashA
July 10, 2008 at 3:57 am
Cluster-aware application is an application that calls the cluster APIs to determine the context under which it is running (such as the virtual server name etc.) and can failover between nodes for high availability.
If you want to know more about cluster concepts and APIs then refer to Windows Operating Platform SDK. In addition, there are several examples in the Platform SDK that can be used to demonstrate Server cluster integration.
-Satya SK Jayanty
SQL Server MVP (Follow me @sqlmaster)
Author of SQL Server 2008 R2 Administration CookBook
SQL Server Knowledge Sharing network
July 10, 2008 at 3:42 pm
Yep, Cluster aware components are ones that are written to support the Cluster API's. These API's include the following:
Cluster API - Works with the Cluster Objects and the Cluster Service
Resource API - Manages resources through the Resource Monitor and Resource DLL
Cluster Administrator Extension API - Enables Custom Resources to be administrated by the Cluster Administrator.
In other words in order for a component or application to be cluster aware it must written implementing these API's.
There are ways to set up non cluster aware components (such as SSIS) in a clustered envrionment by using a generic resource but it is very limited and only gives a basic level of support.
If you look at Services in a cluster you will notice that all services controlled by the cluster are set to manual instead of automatic. (Look at the sql server service). This is because they are under the control of the Cluster Services vs the operating system. In fact it is a bad practice to stop a cluster controlled service through the Services Snap-in. If you stop a clustered controlled service using that method the cluster service may see that as a failure and move the resource to another node.
Some good resources for explaining this are the following:
http://sql-server-performance.com/articles/per/How_SQL_Cluster_Resource_DLLs_work_p1.aspx
and
These ariticles will give you a much more detailed view of all of the Clustering API's
Hope this helps. Let me know if you need more info.
Gary L. Fry
Senior Database Administrator III
MCTS, MCITP SQL Server 2008 Database Administration
Colonial Life/UNUM
July 10, 2008 at 4:16 pm
thats good, but lets get back to the main question,
if i have a cluster server, then how can i change the sql port?
just change the port and every thing will be fine, or when there is a fail over the sql browser wont start and the user will fail to connect to sql server..
..>>..
MobashA
July 11, 2008 at 2:37 am
You should change the port in SQL Server Configuration Manager.
You only need to make this change once as this will take affect on ALL nodes (ie it is 'cluster aware').
If SQL Server does not have enough permission to do so, you will also need to update the SPNs (Service Principal Names) manually. The is a lot of information on the web on how this is done, but check out: http://support.microsoft.com/kb/909801 or Google 'setspn.exe'
Viewing 15 posts - 1 through 15 (of 16 total)
You must be logged in to reply to this topic. Login to reply