Re: VMWare .vmx needed....

Someone will have to explain to me the difference -the *practical*
difference- from Oracle's perspective of a node and an instance
failure.

If you want to show queries pausing and then resuming from where they
left off, for example, (transparent application failover) you can
certainly do that by pulling a plug from a server. You can also do it
by issuing a shutdown abort whilst connected to one of the instances
running on a single node. The effect will be the same in both cases,
and I can't see what having the second node gives you (apart from
trouble, given the virtualisation that's going on already at that
point).

You can similarly monitor the ONS messages in 10g and see them
responding to instance failures (caused by shutdown aborts or killing
off the SMON process of one of the instances, say).

For Oracle, a node failure merely results in an instance failure.
Therefore, multi-instance single-node RAC setups are equally good for
demo purposes as their multi-instance multi-node cousins.

If you must practice traditional clustering with virtual machines,
then you'll have to add disk.locking = "false" to your vmx
configuration file to permit the sharing of the file representing the
clustered hard disk. You'll also need to make that file a 'solid' one
by pre-allocating all the space to it when you create it: it can't
grow dynamically, in other words. Good luck with it, though: it was
unsupported in VMware 3.x when I first used it, and if it's still
there at all in VMware 5 or 6, I'll be mildly surprised.


oracle_man wrote:
All,

Does anyone have a vmware .vmx file configured for Oracle 10.2.0.x
with a 2 node rac cluster running RHEL4? I have the one installed
from a demo dvd from Oracle, but it's only a single node cluster with
2 instances running on it. I want to be able to simulate node
failures for a training meeting for my company in a few days.

Email the .vmx file directly to me at rheadrick AT dbspecialists.com

Any help much appreciated,

Rich

.

Relevant Pages

  • Re: How fault tolerant can Linux be?
    ... point of failure. ... it would involve having redundant hardware; ... so you have to cluster. ... it is a disk or a PSU. ...
    (comp.os.linux.hardware)
  • Re: Failover Question
    ... >> When we induced a failure, ... >> the SQLCLUSTER1 instance manages the database on the ... >> primary server in the cluster and the SQLCLUSTER2 ... >> the SQLCLUSTER1 instance and database, ...
    (microsoft.public.sqlserver.clustering)
  • Re: And diseases of the blood!
    ... huge disaster occurring for a major medical facility. ... We can all think of worst case failure scenarios and terrorist attacks, ... >> So basically, you and I create a cluster of, lets say, SQL servers. ... > worth, say $250,000, to provide this in the event your SAN fails (pick a ...
    (microsoft.public.windows.server.clustering)
  • Re: 2000 Caravan - Primary Instrument Cluster Failure
    ... > I recently had a complete primary instrument cluster failure. ... > speedometer, fuel, temperature, digital odometer and digital ... After some research in the net I?ve found the failure and repair ...
    (rec.autos.makers.chrysler)
  • Re: Failover Question
    ... > Windows 2003 Cluster for our customer. ... > application came back with a failure after a while because ... > the SQLCLUSTER1 instance manages the database on the ... > the SQLCLUSTER1 instance and database, ...
    (microsoft.public.sqlserver.clustering)