By using this site, you agree to our updated Privacy Policy and our Terms of Use. Manage your Cookies Settings.
438,574 Members | 1,968 Online
Bytes IT Community
+ Ask a Question
Need help? Post your question and get tips & solutions from a community of 438,574 IT Pros & Developers. It's quick & easy.

Cluster /AWE / 3GB Switch /Memory Usage

P: n/a
Dear all,

i am planing to implement a Windows 2003 Cluster with MS SQl 2000
Enterprise Edition. I have 2 Nodes (4 * XEON MP Processors) with 8 GB
RAM per Node.

I have the need for 6 SQL instances and would like to implement a
active-active cluster. What do you think it the best way to configure
the memory for the Servers ?
I would like to run 3 instanced per Node but can i allocate 7 GB per
Node for SQL (and 1 for the Operating System) ?
What is happening then i one server fails ?

Should i plan to allocate only 3 GB per Server for SQL to make sure
that one server can handle the load for all instances if one server
fails ?

Should i use the /AWE switch only in the boot.ini to allow more than 4
GB Memory ?
or should i use the /3 GB switch as well ?

Maybe somebody can give me a hint.

Best regards,
Walter
Jul 23 '05 #1
Share this Question
Share on Google+
1 Reply


P: n/a
>
Dear all,

i am planing to implement a Windows 2003 Cluster with MS SQl 2000
Enterprise Edition. I have 2 Nodes (4 * XEON MP Processors) with 8 GB
RAM per Node.

I have the need for 6 SQL instances and would like to implement a
active-active cluster. What do you think it the best way to configure
the memory for the Servers ?
I would like to run 3 instanced per Node but can i allocate 7 GB per
Node for SQL (and 1 for the Operating System) ?
What is happening then i one server fails ?

Should i plan to allocate only 3 GB per Server for SQL to make sure
that one server can handle the load for all instances if one server
fails ?

Should i use the /AWE switch only in the boot.ini to allow more than 4
GB Memory ?
or should i use the /3 GB switch as well ?

Maybe somebody can give me a hint.

Best regards,
Walter

I am afraid 4GB of memory must go to waste. This is a typical probem of 32bit
systems, which is why 64bit is so important for RDBMS.

To use more than 4GB of memory, both /PAE (set on Windows) and AWE (set in SQL)
must be turned on. AWE memory, however, is static. Meaning SQL will allocate
all the memory specified in AWE (according to max server memory configuration
in sp_configure) to itself.

In a cluster where there is more than one instance, setting up AWE is highly
problematic because EACH instance will take up whatever memory that is spcified
in AWE. In other words, if you have two instances set to eat up 6GB each using
AWE, then they will BOTH try to take up 6GB even if they both resides in the
same node with only 8GB. The result is often disatrous and cause the OS to
crash.

In summary, you never setup AWE in a active/active cluster unless you have
plenty of memory to support the static memory in all the instances in ALL the
nodes. For example, if you want two instances to use 6GB of memory each, you
will be adviced to have each node to have 16GB of memory, this way, even when
one node holds both instances, SQL will only use 12GB of its total memory.

In your case, since you want to hold 6 instances in a two node cluster with
only 8GB of memory, it is best not use AWE at all, and let SQL allocate 4GB of
memory dynamically. You should, however, turn on /3GB.

Joe
MCDBA, MCSD, OCP
Jul 23 '05 #2

This discussion thread is closed

Replies have been disabled for this discussion.