Home | Contact Us | FAQ | Search & Site Map | Link to Us
Sign In | Join | Other 45 Sites in Network
Home
Discussion Groups
DB Engine
SQL ServerMSDESQL Server CE
Services
Analysis (Data Mining)Analysis (OLAP)DTSIntegration ServicesNotification ServicesReporting Services
Programming
CLRConnectivitySQLXML
Other Technologies
ClusteringEnglish QueryFull-Text SearchReplicationService Broker
General
Data WarehousingPerformanceSecuritySetupSQL Server ToolsOther SQL Server Topics
DirectoryUser Groups
Related Topics
MS AccessOther DB ProductsMS Server Products.NET DevelopmentVB DevelopmentJava DevelopmentMore Topics ...

SQL Server Forum / Other Technologies / Clustering / March 2005

Tip: Looking for answers? Try searching our database.

SQL 7.0 on Windows 2000

Thread view: 
Enable EMail Alerts  Start New Thread
Thread rating: 
Thom - 31 Mar 2005 14:09 GMT
Okay go ahead and shred me now, and then please see if you can answer my
question. I have a client that is living in the dark ages. Currently they are
clustered on NT4.0 and SQL 7.0 using legato.  We are moving them a step at a
time to SQL 2000/Windows 2003 clustering.  W2k and SQL7.0 using microsoft
clustering is the first baby step.  I need to know if I can name my SQL7.0
virtual server the same as the cluster.
We are trying to have minimal impact on the apps, that hit the server and
they are not sure if any is mapping to the cluster at this time.
Geoff N. Hiten - 31 Mar 2005 15:04 GMT
I wouldn't even try and  give them the same names.  I would use a DNS srv
record to repoint the old name to the new system (assuming an AD
infrastructure). This has the advantage that you can turn it off and on at
will to test.  If you have only one or two client app servers, you can do
the same thing with a client network alias.

Geoff N. Hiten
Microsoft SQL Server MVP

> Okay go ahead and shred me now, and then please see if you can answer my
> question. I have a client that is living in the dark ages. Currently they
[quoted text clipped - 6 lines]
> We are trying to have minimal impact on the apps, that hit the server and
> they are not sure if any is mapping to the cluster at this time.
Thom - 31 Mar 2005 16:17 GMT
Okay, just for a little clarification for myself,
We are in AD so if this was the setup:
physical01
physical02 - both clusteres to virtual03 ip of 192.168.0.1 (example only)
sqlvirtual04 ip of 192.168.0.2,1433 (example only)

If i went into DNS and setup a record to point 192.168.0.1 to sqlvirtual04
then the shares could be mapped as \\sqlvirtual04\d_drive
and sql would still be registered as sqlvirtual04?

Let me know if this is what you were referring to?
Since the client doesn't know if the shares were being used on the original
sql server, I would like to leave name the virtual sql the same as the
original, but allow drives to be mapped also.

> I wouldn't even try and  give them the same names.  I would use a DNS srv
> record to repoint the old name to the new system (assuming an AD
[quoted text clipped - 15 lines]
> > We are trying to have minimal impact on the apps, that hit the server and
> > they are not sure if any is mapping to the cluster at this time.
Geoff N. Hiten - 31 Mar 2005 16:27 GMT
File shares and SQL Server services are two different issues.  You can alias
a SQL server many ways, including DNS records and Client aliases.  Resolving
share names is a different issue.  A SQL Server 2000 system normally doesn't
use file shares, except in replication snapshot resolution.  Also, the SRV
record includes a service name and an IP address and a port number.
Normally, these records are registered by the service so other systems can
find it.  This is how AD finds its various management roles "under the
covers".

GNH

> Okay, just for a little clarification for myself,
> We are in AD so if this was the setup:
[quoted text clipped - 38 lines]
>> > and
>> > they are not sure if any is mapping to the cluster at this time.
Thom - 31 Mar 2005 16:51 GMT
Sorry for the confusion... sql will not be using the file shares.  but
individuals may be using them... ie.  copy some data over to the windows
share and then use dts to load the data.

So with legato, and I don't know much about it, the user would map a drive
to \\servername\d_drive and then copy it over.  since it was a default
instance the sql had the same name as the server.

Now with MSCS we have a two physical servers, nodea and nodeb, the get
clustered and given the name of the old server <servername> from above. The
user would map drive to \\servername\d_drive and copy the file as before.

But now I have virtual sql that doesn't have the same name as before
<servername> from above.  Since the customer isn't sure about the use of the
file shares I would like to be able to have people still connect to sql
server using <servername> as before with legato.  but I have already used
that name for the windows cluster name.

That is my true dilemna, will this work using DNS and pointing the old
server name and port 1433 to the new sql virtual name?

Thanks for the assitance Geoff.

> File shares and SQL Server services are two different issues.  You can alias
> a SQL server many ways, including DNS records and Client aliases.  Resolving
[quoted text clipped - 49 lines]
> >> > and
> >> > they are not sure if any is mapping to the cluster at this time.
Geoff N. Hiten - 31 Mar 2005 19:05 GMT
You can create a clustered file share and group it with the SQL server
instance so it follows along with the cluster from node to node.  Another
way would be for you to set up a remote file share on a non-clustered
instance and use a UNC name to connect DTS with the data file.

If you already have used the old name for your cluster, you will have huge
problems with a DNS alias  You may even prevent the cluster from coming
online.  You probably need to nuke the cluster and rebuild it from scratch
using entirely new names.  Then you can use alias techniques via DNS or
Client tools.  Port 1433 is not an issue.  Both the DNS record and the
Client Network utility will allow you to specify a port number as part of
the alias.  If the file share is in the same virtual server group, you will
use the same network name for both, simplifying the mapping process.

Geoff N. Hiten
Microsoft SQL Server MVP

> Sorry for the confusion... sql will not be using the file shares.  but
> individuals may be using them... ie.  copy some data over to the windows
[quoted text clipped - 87 lines]
>> >> > and
>> >> > they are not sure if any is mapping to the cluster at this time.
 
Sign In
Join
My Latest Posts
My Monitored Threads
My Blog
My Photo Gallery
My Profile
My Homepage

Start New Thread
Enable EMail Alerts
Rate this Thread



©2009 Advenet LLC   Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.