Re: Portents of VMS death

From: jlsue (jefflsxxxz_at_sbcglobal.net)
Date: 05/28/03 

Date: Wed, 28 May 2003 16:30:52 GMT

On Mon, 26 May 2003 13:58:56 -0400, JF Mezei <jfmezei.spamnot@istop.com>
wrote:

>
>But in the real world, if you have 2-3 nodes in your cluster, one production,
>one for testing/backup and one for quorum vote, you don't have that luxury.
>When the production node fails, you must start the production application on
>the backup node, and if the application has its own drivers, you need to first
>unlaod the test drivers and then reload the production ones. You need to reboot.
>

That's not a real-world cluster. That's an accident waiting to happen.
Clustering development/test systems with production is ludicrous design.

The *reason* for a cluster is a scalable, available server for
applications. That implies multiple servers providing access to the same
applications in a load-balanced way.

Fail-over clustering can be done by the cheapest crap out there. Those
configurations are not highly available and should not be implemented where
availability is a means to achieve certain business goals.

Relevant Pages

  • Re: Disaster Recovery Planning
    ... > shipping to keep the DR databases in sync with the production DBs; ... > automate all the log shipping role changes (150+ databases is a bit too ... The cluster nodes were ... >> I've never seen a disaster recovery plan for SQL Server. ...
    (microsoft.public.sqlserver.server)
  • Re: Replicating Drives to use for Development Instances
    ... with the -S you can write a new Disk Signature:) ... MVP - Windows Server - Clustering ... > site B failover) and a cluster with one node at each site. ... > production and a development instance on the same system. ...
    (microsoft.public.windows.server.clustering)
  • Re: System organization preference question
    ... parts of the startup chain to the files located on that shared disk. ... This so that a cluster with multiple system disks has a central disk ... with one or more system disks being off-line during rolling upgrades. ... production -- I do typically leave the files in the default location. ...
    (comp.os.vms)
  • Re: Portents of VMS death
    ... >> in a cluster with production would get you hit ... Especially in the financial sector. ... We had two production and one ... developent cluster - attempts to combine one of the production clusters with ...
    (comp.os.vms)
  • Re: Running REXX program in a batch job
    ... Applications development to write code for production batch, but batch REXX, including batch REXX that runs in batch TSO or batch ISPF environments, has been frequently used by Technical Services to quickly implement a number of simple utilities needed by Applications Development. ... One thing which may make this practical in our environment which I gather may not be true in yours - our production support people are the same people who do application development. ...
    (bit.listserv.ibm-main)