Cluster Set up Question

Hello,
We are implementing clustering (Active/Passive) in our Production environment.
So we will have two windows server 2003 for SQL SERVER. (ProdDbServer1 and
ProdDbServer2)
I will install SQL Server on both the machine and i will create my database
(FinDB)
on both the machine's SQL SERVER with all the tables,store procs, functions
etc.
So lets say if my Active Server is in action and accepting all the OLTP
transaction
and for some reason if it goes fail...How can passive server can take over,
how
passive server can be at the same point of the fail of active server ?
Bacically how both the Database will be in sink with each other ?
Pls help me
A cluster consists of two machines and a storage system that is physically
connected to both systems. The Cluster software arbitrates ownership of the
disk resource so that only one server can access it at a time. The cluster
software also monitor the sql instance on the active host node for failure.
If the first node fails, ownership of the SQL instance and the corresponding
disk resource is transferred to the other node. From the outside, it looks
like the SQL Server stopped and restarted.
Geoff N. Hiten
Senior Database Administrator
Microsoft SQL Server MVP
"mvp" <mvp@.discussions.microsoft.com> wrote in message
news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
> Hello,
> We are implementing clustering (Active/Passive) in our Production
> environment.
> So we will have two windows server 2003 for SQL SERVER. (ProdDbServer1 and
> ProdDbServer2)
> I will install SQL Server on both the machine and i will create my
> database
> (FinDB)
> on both the machine's SQL SERVER with all the tables,store procs,
> functions
> etc.
> So lets say if my Active Server is in action and accepting all the OLTP
> transaction
> and for some reason if it goes fail...How can passive server can take
> over,
> how
> passive server can be at the same point of the fail of active server ?
> Bacically how both the Database will be in sink with each other ?
> Pls help me
|||Thanks Geoff.
So From your reply, I understand that i will have to install SQL SERVER 2000
on both the windows 2003 server.
I have following questions.
1) After installing SQL SERVER on both the windows m/c, When i create
database on both the machine, Would i have to mentioned same Data files and
Transaction Log Path ( Because we will have storage will be in SAN), when i
create DB ?
2) For example if i create one table in the FinDB of on ServerA and then if
i go to FinDB of ServerB, I will see that table there too, if i will have
Datafiles and Transaction Log Path are same for both the servers ?
"Geoff N. Hiten" wrote:

> A cluster consists of two machines and a storage system that is physically
> connected to both systems. The Cluster software arbitrates ownership of the
> disk resource so that only one server can access it at a time. The cluster
> software also monitor the sql instance on the active host node for failure.
> If the first node fails, ownership of the SQL instance and the corresponding
> disk resource is transferred to the other node. From the outside, it looks
> like the SQL Server stopped and restarted.
> --
> Geoff N. Hiten
> Senior Database Administrator
> Microsoft SQL Server MVP
>
> "mvp" <mvp@.discussions.microsoft.com> wrote in message
> news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
>
>
|||Thanks Geoff.
So From your reply, I understand that i will have to install SQL SERVER 2000
on both the windows 2003 server.
I have following questions.
1) After installing SQL SERVER on both the windows m/c, When i create
database on both the machine, Would i have to mentioned same Data files and
Transaction Log Path ( Because we will have storage will be in SAN), when i
create DB ?
2) For example if i create one table in the FinDB of on ServerA and then if
i go to FinDB of ServerB, I will see that table there too, if i will have
Datafiles and Transaction Log Path are same for both the servers ?
"Geoff N. Hiten" wrote:

> A cluster consists of two machines and a storage system that is physically
> connected to both systems. The Cluster software arbitrates ownership of the
> disk resource so that only one server can access it at a time. The cluster
> software also monitor the sql instance on the active host node for failure.
> If the first node fails, ownership of the SQL instance and the corresponding
> disk resource is transferred to the other node. From the outside, it looks
> like the SQL Server stopped and restarted.
> --
> Geoff N. Hiten
> Senior Database Administrator
> Microsoft SQL Server MVP
>
> "mvp" <mvp@.discussions.microsoft.com> wrote in message
> news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
>
>
|||Does anyone want to rethink their perspective on why active/active and
active/passive should be used as terminology?
"mvp" <mvp@.discussions.microsoft.com> wrote in message
news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
> Hello,
> We are implementing clustering (Active/Passive) in our Production
> environment.
> So we will have two windows server 2003 for SQL SERVER. (ProdDbServer1 and
> ProdDbServer2)
> I will install SQL Server on both the machine and i will create my
> database
> (FinDB)
> on both the machine's SQL SERVER with all the tables,store procs,
> functions
> etc.
> So lets say if my Active Server is in action and accepting all the OLTP
> transaction
> and for some reason if it goes fail...How can passive server can take
> over,
> how
> passive server can be at the same point of the fail of active server ?
> Bacically how both the Database will be in sink with each other ?
> Pls help me
|||If you install SQL Server twice, you will have 2 physically separate
instances of SQL Server running within a single cluster. If you then create
your database in each of these instances, you will have 2 physically
separate copies of the database. In order to synchronize data between the
two databases, you would have to implement replication, log shipping, or
database mirroring (2005 only). If one of the servers failed, you would
have to change your application connections to point at the other SQL Server
instance.
"mvp" <mvp@.discussions.microsoft.com> wrote in message
news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
> Hello,
> We are implementing clustering (Active/Passive) in our Production
> environment.
> So we will have two windows server 2003 for SQL SERVER. (ProdDbServer1 and
> ProdDbServer2)
> I will install SQL Server on both the machine and i will create my
> database
> (FinDB)
> on both the machine's SQL SERVER with all the tables,store procs,
> functions
> etc.
> So lets say if my Active Server is in action and accepting all the OLTP
> transaction
> and for some reason if it goes fail...How can passive server can take
> over,
> how
> passive server can be at the same point of the fail of active server ?
> Bacically how both the Database will be in sink with each other ?
> Pls help me
|||No. You create the cluster. You then install a single copy of SQL Server
into the cluster. You then create your database inside that instance. You
do not install SQL Server directly to a machine. It is installed to the
cluster.
"mvp" <mvp@.discussions.microsoft.com> wrote in message
news:95A55E02-1BCB-47E1-9458-E01080486B3D@.microsoft.com...[vbcol=seagreen]
> Thanks Geoff.
> So From your reply, I understand that i will have to install SQL SERVER
> 2000
> on both the windows 2003 server.
> I have following questions.
> 1) After installing SQL SERVER on both the windows m/c, When i create
> database on both the machine, Would i have to mentioned same Data files
> and
> Transaction Log Path ( Because we will have storage will be in SAN), when
> i
> create DB ?
> 2) For example if i create one table in the FinDB of on ServerA and then
> if
> i go to FinDB of ServerB, I will see that table there too, if i will have
> Datafiles and Transaction Log Path are same for both the servers ?
> "Geoff N. Hiten" wrote:
|||The SQL Installer is cluster-aware. That is, it handles all the details of
cluster-wide installation. It creates the virtual server and installs the
local binaries on all nodes. It also groups all the cluster resources and
sets the dependencies correctly. Once SQL is running and you are connected,
there is almost no difference in SQL operations between clustered and
non-clustered SQL Servers. You end up connecting to a virtual SQL Server
instance that looks identical regardless of which node is actually hosting
it at the moment.
Here is an excellent overview of SQL failover clustering that you may find
helpful:
http://www.microsoft.com/technet/pro.../failclus.mspx
Geoff N. Hiten
Senior Database Administrator
Microsoft SQL Server MVP
"mvp" <mvp@.discussions.microsoft.com> wrote in message
news:95A55E02-1BCB-47E1-9458-E01080486B3D@.microsoft.com...[vbcol=seagreen]
> Thanks Geoff.
> So From your reply, I understand that i will have to install SQL SERVER
> 2000
> on both the windows 2003 server.
> I have following questions.
> 1) After installing SQL SERVER on both the windows m/c, When i create
> database on both the machine, Would i have to mentioned same Data files
> and
> Transaction Log Path ( Because we will have storage will be in SAN), when
> i
> create DB ?
> 2) For example if i create one table in the FinDB of on ServerA and then
> if
> i go to FinDB of ServerB, I will see that table there too, if i will have
> Datafiles and Transaction Log Path are same for both the servers ?
> "Geoff N. Hiten" wrote:
|||Thanks for the reply.
No i do not want two sql server instance in single cluster, I just want one
instance in my cluster. so what should i do, install SQL SERVER 2000 Software
and create my database in just one machine (ProdSqlServerA)?
If i do like this and if i go to another sever (ProdSqlServerB), would i see
SQL SERVER and my created Db on that maching too ?
"Z" wrote:

> If you install SQL Server twice, you will have 2 physically separate
> instances of SQL Server running within a single cluster. If you then create
> your database in each of these instances, you will have 2 physically
> separate copies of the database. In order to synchronize data between the
> two databases, you would have to implement replication, log shipping, or
> database mirroring (2005 only). If one of the servers failed, you would
> have to change your application connections to point at the other SQL Server
> instance.
>
> "mvp" <mvp@.discussions.microsoft.com> wrote in message
> news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
>
>
|||No thanks! I don't want Mike and Geoff mad at me anymore on this subject.
Cheers,
Rod
MVP - Windows Server - Clustering
http://www.nw-america.com - Clustering Website
http://msmvps.com/clustering - Blog
http://www.clusterhelp.com - Cluster Training
"Z" <z@.z.com> wrote in message
news:%23MIgYMvFGHA.984@.tk2msftngp13.phx.gbl...
> Does anyone want to rethink their perspective on why active/active and
> active/passive should be used as terminology?
>
> "mvp" <mvp@.discussions.microsoft.com> wrote in message
> news:3F5342B1-3AB6-4CD7-B763-33587B1BED0E@.microsoft.com...
>
sqlsql