Guidelines

Guidelines on use of Waterloo computing and network resources

Preamble

Computing and network resources are important components of the University of Waterloo's infrastructure. These guidelines govern the appropriate and ethical use of these resources, inform users of expectations and responsibilities assumed in the use of the University of Waterloo computing and network resources, and clarify the context.

Please refer to the Information Systems and Technology's webpage titled "guidelines on use of Waterloo computing and network resources".

Departmental hardware specification policy

Taken from "State of Departmental Computing 2008-Section Four: Future Growth- Policy Recommendations For The Future".

Hardware specifications exist to ensure driver support, maximum return on investment, and cost-effectiveness. For example, it's not reasonable for an Information Technology (IT) staff member to spend four hours fixing a six-year-old Dell computer that is a model they're unfamiliar with and that is worth $20.

  • Newly purchased hardware to be supported by the IT staff must be purchased from the departmental computer specification.
  • Existing hardware must be a maximum of four years old to be supported by IT staff. Older lab equipment driven systems are the only exception to this rule.

A common practice of organizational IT is to implement "rolling policies". These are policies that minimize downtime by replacing the hardware on schedule:

The rolling departmental server policy

  • Mission-critical systems must be replaced after three years of uptime.
  • Non-mission critical systems that can afford more than three days of downtime must be replaced after five years.

Rolling administrative/technical staff workstation policy

  • These are mission-critical systems used to run the department. The replacement schedule is three years. Unlike faculty and researchers, staff members typically don't have laptops that would keep downtime more bearable.

Rolling undergraduate computing facilities policy

  • These are high usage machines used solely by our undergraduate students. They cannot afford downtime especially during peak periods when resources are at 100% usage. This is also an attractive feature to add to marketing our program to prospective students. The replacement schedule is three years.  

Approved unanimously at the department meeting held on Thursday, October 16, 2008.

Civil and Environmental Engineering Information Technology (CEE IT) cost recovery policy

Overview

To sustain our rapidly growing and vibrant research computing environment while not limiting the services provided to researchers, staff, or students, a mechanism needs to be established. The CEE IT team's skills continue to enhance and enrich our research environment.  This document outlines a departmental policy designed to both ensure fairness of the use of IT staff time, and to recover costs associated with the use of IT staff time for services provided for research computing resources.

Research services

Research services are here defined as any services provided by the CEE IT staff in explicit support of research activities which fall outside of their primary responsibilities (i.e., supporting undergraduate and administrative computing, and desktop and laptop support for researchers).

Research IT services subject to cost recovery include:

  • Hardware server deployment and support
  • Software server deployment and support
  • Multi-functional copiers
  • Programmatic solutions (i.e., scripting, programming, application development)

The following research-related services are not subject to cost recovery:

  • Desktop/laptop computing support
  • Integration into the University of Waterloo IT infrastructure (i.e., Quest, FORE, Nexus, Jobmine, ACE, etc.)
  • Regular and ongoing service calls

These lists are non-comprehensive and other services not explicitly stated should be discussed with the CEE IT manager.

Cost recovery model

There are many different models in use across campus with the simplest of these being an hourly cost recovery rate.

  • Cost recovery will not be used for every single service provided (service calls, daily support, troubleshooting, etc) for existing IT hardware that meets the minimum Civil and Environmental Engineering (CEE) specifications (minimum supported hardware specification is described in the “State of Computing 2008” Report - Hardware specifications exist to ensure driver support, maximum return on investment and cost-effectiveness).
  • Cost-recovery will be applied on a per-project basis for major hardware or software services (e.g., server deployment). The cost will include support for the recommended support period of the server or service.
  • A cost recovery rate of $60 per hour will be used.
  • Infrastructure costs, including housing the equipment, and electrical and cooling costs (with the exception of massive multi-system deployments, such as high performance computing clusters) will not be subject to cost recovery.

Examples

Example 1. File server deployment

A typical file server deployment would be associated with a cost recovery of approximately $1200, based on:

  • Consultation, system specification, requests for quotation for system = three hours = $180
  • Receiving and installing hardware into rack = one hour = $60
  • Installing windows OS (Operating System) and software = three hours = $180
  • Testing hardware = one hour = $60
  • Installing, configuring RAID (redundant array of inexpensive disks) array and testing array for failure = five hours = $300
  • Customisation/training for user = one hour = $60
  • Live migration of an existing server = six hours = $360
  • Rack mount space for equipment = free
  • Networking equipment to support server (gigabit connection, patch cables, etc) = free
  • Cooling for equipment = free
  • Use of existing electrical facilities = free
  • Future on-site support of equipment (hardware failures, etc) = free

Example 2.  Existing file server

Existing server hardware that still meets the CEE minimum hardware specifications may:

  1. Opt-in at a half-price pro-rated service plan for the remainder of their warranty period. 
  2. Commence hourly rate, half-priced for the remainder of the warranty period.

Services performed on existing research hardware that does not meet our minimum hardware specifications will commence cost-recovery on an hourly basis.

For a typical new fileserver (with three year warranty) with a deployment that would now cost $1200, but was deployed one year ago, the following cost recovery would apply:

  • Pro-rated cost is two-thirds of $1200 = $800
  • Half of pro-rated cost is $400
  • Cost to researcher is $400

For a server that is older than four years and is in need servicing:

  • OS upgrade and user migration = four hours = $240
  • Hard drive replacement = one hour = $60
  • Data migration = two hours = $120
  • Cost to researcher is $420

Approved at the December 16th, 2010 Department Meeting.