Failover Nodes - Varying Hardware

Discussion in 'microsoft.public.sqlserver.clustering' started by Randy, Apr 8, 2009.

  1. Randy

    Randy Guest

    We are replacing one SQL 2000 cluster node. The existing two nodes have 4GB
    with the /3GB switch - no PAE. AWE is not enabled. The new node will have
    15.2 GB with PAE. Should I still use the /3GB switch even though there is
    more than 8 GB?

    Randy, Apr 8, 2009
    1. Advertisements

  2. To answer your question, /3GB should stop at no higher than 12 GB.

    As for replacing cluster nodes with asymmetrical hardware, that is not
    supported on Windows 2000/2003. There are very good technical reasons why
    matched HCL-listed hardware is required.

    Then again, with SQL 2000, you are pretty much out of support anyway.
    Geoff N. Hiten, Apr 8, 2009
    1. Advertisements

  3. Randy

    Randy Guest

    Thanks. I will not use the switch and suggest we replace both nodes, not
    just the bad one. Unfortunately, the active node had been patched and
    requires a reboot. I can't add a new nodes until it's rebooted. Looks like
    we are going to be down.
    Randy, Apr 8, 2009
  4. One more reason to go to SQL 2008. Everything is rolling upgrades. One or
    two failovers is all it takes.

    Geoff N. Hiten, Apr 8, 2009
  5. Randy

    Randy Guest

    One of these days.... Thanks.

    Randy, Apr 8, 2009
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.