New Technology Implementation
Overview
Barreling into a technology decision can lead to situations where the intended benefits are greatly out weighed by negative ramifications that were never planned for.
Though you cannot plan for every possible contingency, the questions that follow will head off the most naive oversites when considering a new technology or various technical solutions. It is standard for large organizations to ask these questions:
- Is there a business requirement from the customer for it?
- What is the cost and who will be funding it?
- Does the I.T. infrastructure support it (network bandwidth, system: cpu, io, disk space, RAM)?
- Is it easy to support?
- Does it require any specialized training?
- Specialized training being defined here as: any training over and above Oracle University Administration Workshop I and II classes.
- If it requires specialized training how soon can you send your DBAs and how many can you send?
- Can it easily be re-implemented if there is a disaster leaving you with zero computer assets?
- Are you in an environment that requires strict adherence implementing routine Oracle patches? If so how difficult is it to apply them after the proposed solution has been implemented?
- How deep is its reliance on particular hardware or software versions?
- Does implementing the proposed solution make the following maintenance tasks significantly harder:
- Resizing tablespaces.
- Adding, changing, dropping, moving: control, data, redo or archive log files.
- What needs to be changed for your proposed solution to work with the next major Oracle version?
- Is there a critical problem that cannot be solved without it?
- What happens if it is not implemented, i.e. if we do nothing?
New Technology Sidebar
In most Fortune 200 environments new technologies must prove themselves. Just because it has been made to work in a small test environment does not warrant the risk of rolling it out. In environments where risk of failure is not an option, other DBA's need to validate your test findings. If success cannot be duplicated, implementing the preposed new technology has no business in production.
The validation phase is also instrumental in determining if a technology is supportable or a tech support nightmare. A technology costing the business mission overwhelming DBA hours and profitability in most I.T. environments is not considered a "solution".