2nd instance of an active\active cluster Fails to allocate min memory
-
Has anyone experienced this strange type of problem before.
I'm currently running 2 SQl Instance as an active/active cluster.
Both servers are identical running windows 2000 Advance server.
SQl Server Enterprise 2000 with settings exactly the same on both servers.
with 8GB of memory, both servers are new builds, Windows sees all the memory, boot.ini as 3gb pae switches and SQl’s are awe configured.
Instance 1 running as default virtual server on node 1
Min server memory 5092
Max Server memory 7406
Uses all available memory.
Instance 2 running as named instance of virtual server on node 2
Min server memory 5092
Max Server memory 7406
Fails to allocate min memory error in log
Runs using only 3.73 GB as max
But I'm getting excessive page because more memory is needed.
-
The problem I think is that even though you have 2 different nodes the following is happening (I think):
Assumptions:
You are referencing the cluster name for failover purposes
Issue:
The Cluster level places both nodes into 1 virtual server name. Each of the nodes has 8GB RAM available unfortunately, the virtual server ONLY has 8GB available.
The cluster starts instance 1 and says here is your 5GB RAM. It then tells the 2nd instance here is what is left AFTER the OS takes what it needs to stay running.
I understand what you attempted to do and I would have thought that it would have worked until I started thinking about it.
I am only now starting to learn more about clusters and may be completely off base on the issue.....
Good Luck,
AJ
Good Hunting!AJ Ahrens
webmaster@kritter.net -
I should have added current active usage is node 1 uses 7.43 GB and node 2 uses 3.73GB which is over 11GB
Thanks anyway
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply