Cluster Services

  • Thread starter Thread starter Scott
  • Start date Start date
S

Scott

Running a Win2K domain. MSCS is used as a file/apps
server and SQL 2000. Running now in an Active/Passive
mode, would like to make it Active/Active with the
file/apps on NodeA and SQL on NodeB. Couple of
questions, 1. Running it this way (Active/Active), would
SQL failover to NodeA if NodeB fails? And would the
file/apps failover to NodeB if NodeA fails? 2. If the
answer above is yes (I thought I read somewhere that it
would, but can't remember), how do I switch it from
Active/Passive to Active/Active?

Thanks in Advance,


Scott
 
Hi Scott,

From an OS/Cluster perspective, Active and Passive modes are "states" that the cluster is in. If you have resources running on both nodes simultaneously, then that is considered Active/Active
state. If you have resources only running on one of the nodes in the cluster, then it is considered Active/Passive state. This is the same for clusters with more than 2 nodes as well (I.E.
Active/Active/Passive). There is no "switching" from Active/Active to Active/Passive, it's simply relative to where the resources are running.

From an application perspective such as SQL, an Active/Active configuration would mean that you have two separate instances of SQL - one running on each node simultaneously (this would
then of course mean that you are in an Active/Active state as far as the cluster is concerned).

To answer your first question, yes, the SQL should fail over to the other node even though it's already running other resources. What you need to do is complete some benchmark testing to
ensure that each node in your cluster will be able to service all resources if you run into a problem with one of the nodes.

This message is provided as is and does not imply any warranty or gaurantee of support of any kind.


--------------------
| Content-Class: urn:content-classes:message
| From: "Scott" <[email protected]>
| Sender: "Scott" <[email protected]>
| Subject: Cluster Services
| Date: Thu, 8 Jan 2004 12:49:16 -0800
| Lines: 16
| Message-ID: <[email protected]>
| MIME-Version: 1.0
| Content-Type: text/plain;
| charset="iso-8859-1"
| Content-Transfer-Encoding: 7bit
| X-Newsreader: Microsoft CDO for Windows 2000
| Thread-Index: AcPWKOGBDriHIMehT+q3rf/HbJHNZA==
| X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4910.0300
| Newsgroups: microsoft.public.win2000.advanced_server
| Path: cpmsftngxa07.phx.gbl
| Xref: cpmsftngxa07.phx.gbl microsoft.public.win2000.advanced_server:15724
| NNTP-Posting-Host: tk2msftngxa13.phx.gbl 10.40.1.165
| X-Tomcat-NG: microsoft.public.win2000.advanced_server
|
| Running a Win2K domain. MSCS is used as a file/apps
| server and SQL 2000. Running now in an Active/Passive
| mode, would like to make it Active/Active with the
| file/apps on NodeA and SQL on NodeB. Couple of
| questions, 1. Running it this way (Active/Active), would
| SQL failover to NodeA if NodeB fails? And would the
| file/apps failover to NodeB if NodeA fails? 2. If the
| answer above is yes (I thought I read somewhere that it
| would, but can't remember), how do I switch it from
| Active/Passive to Active/Active?
|
| Thanks in Advance,
|
|
| Scott
|
|
 
Thanks Kevin for your response.

Follow-up question, if SQL is installed on both nodes,
but the dependencies indicate that it is dependent on a
detached raid5 drive which is only on one drive, how
would I move SQL to the second node?

Scott

-----Original Message-----
Hi Scott,

From an OS/Cluster perspective, Active and Passive modes
are "states" that the cluster is in. If you have
resources running on both nodes simultaneously, then that
is considered Active/Active
state. If you have resources only running on one of the
nodes in the cluster, then it is considered
Active/Passive state. This is the same for clusters with
more than 2 nodes as well (I.E.
Active/Active/Passive). There is no "switching" from
Active/Active to Active/Passive, it's simply relative to
where the resources are running.
From an application perspective such as SQL, an
Active/Active configuration would mean that you have two
separate instances of SQL - one running on each node
simultaneously (this would
then of course mean that you are in an Active/Active
state as far as the cluster is concerned).
To answer your first question, yes, the SQL should fail
over to the other node even though it's already running
other resources. What you need to do is complete some
benchmark testing to
ensure that each node in your cluster will be able to
service all resources if you run into a problem with one
of the nodes.
This message is provided as is and does not imply any
warranty or gaurantee of support of any kind.
 
Back
Top