Reporting and Statistics of Enterprise Lab Management

1
Reporting and Statistics of Enterprise Lab
Management

• Matthew Maynard Senior Team Leader Common-Use
  Computing Infrastructure

Learning Environment Services, Information
Services
2
Who we are and what we do

• Matthew Maynard, Griffith University for 7.5
  years
• 5 of those have been with CommonUse computing
• Learning Centres, non-school-owned computer
  laboratories, lecture theaters and seminar rooms,
  laptop fleet, self-help kiosks.
• Over 2200 pcs on 7 different campuses
• Great team of only 4 staff members
• Build image, image deployments, application
  packages, antivirus, student printing backend
• Server and Novell Tree Administrators

3
Some things we are going to cover

• Point and click lab management,
• What is it?
• Why might you need it?
• How do you achieve it?
• Overview of tools and systems used to achieve
  point and click method
• Planning using reporting and statistics

4
Background

• Set up Fleets Standards aiming toward point
  and click lab management
• Leveraging of the use of fleet standards to
  create reports and statistics
• Head Toward
• Software metering and its uses
• Streaming OS / Multiple OS cost analysis

5
From Easy to Easier

• Rolling out images - easy
• Managing large fleets of different hardware types
  across many different location - easy
• Asset management Dynamic asset DB - easy
• Antivirus and software deployment - easy
• Reporting and auditing of facilities easy

6
Whats Hard

• Increased user demand on resources
• User expectations
• Increased user knowledge of systems
• Extended communication
• Flexibility to change

7
What is it?

• My Holy Grail concept of working from home
• How many here can achieve this?
• Its about not having to roam about in the middle
  of the night re-imaging labs hoping they work
• Tried and tested, robust, full confidence

8
Why do you need it

• Reduces staff, less on the ground, 5001
• In fact we had staff reduction with PC increase
• Reduces complexity, simple management
• Reduces cost, lower overtime
• Systems do the bulk of the work
• Process reports when finished
• Best case full fleet re-image in 3 hrs

9
How do you get it?

• You need vision at least 18-24 months
• Time and continuous planning to meet the
  objectives for where you want to go
• Staff who know the goal and can work toward it
• Staff with varying skills set (big help)
• Easy if someone is showing you the way

10
Sticky questions

• Is rolling out software images a major burden?
• How do you cope if your image has errors or needs
  additional software changes after deployment?
• How do you measure your resource usage and use
  that to plan for the future?
• If your management doubled your computer fleet
  tomorrow how would you cope?

Learning Environment Services, Information
Services
11
Easy Answers (The 5-ations)

• Centralisation
• Standardisation
• Automation
• Innovation
• Motivation

12
Centralisation

• Nathan campus is the central hub for management
• Single point of build, test, manage and plan
• Master server for replication out to all other
  campuses
• Replicates student data
• Replicates network deployed software
• Replicates images
• Replicates TFTP boot files
• Central reporting of consoles and databases

13
Network Delivered Applications Data Sets PXE /
TFTP boot files Images
14
Standardisation

• 1 image for all hardware types
• 1 image for all campuses
• 1 image for Learning Centres, Labs, Lecture
  Theatres and Seminar rooms.
• Standard look and feel everywhere
• Standard drive mappings
• Standard data availability
• Standard print system

15
Automation

• Automatically detects machine type and role and
  dynamically configures
• Automatic configuration of campus- or lab-based
  variables
• Fully configures machine and reports all data to
  database before finally deep-freezing
• Scripting is the key to single image

16
Innovation

• Many unique solutions for problems provided from
  the wide staff skill set
• Out-of-the-box thinking no idea too crazy
• Focus on what we can do not what we cant
• Service- focused not system- focused
• Got to be smarter and more efficient

17
Motivation

• Small yet highly effective team
• Highly self-motivated
• Lots of research, testing and developing
• Test facilities
• Continuous improvement and evolution of systems
• Never stand still

18
Major Systems Overview

• PXE (Argon CMS)
• WUOL
• Scripting
• DeepFreeze Enterprise
• PCID (in-house)
• MySQL
• SCEMS (in-house)

19
SCEMS

• Student Computing Environment Management System
• Utilisation of facilities
• Student tracking
• Integrated with timetabling for room bookings
• Computer vacancy

20
Statistics and Reporting

• Standards allow for consistent data input into
  management systems.
• .VB net
• Named the same (for DB lookup and workflow)
• Same DB connectors (oracle MySQL)
• Same management suite (PXE)

21
Statistics and Reporting

• Track user login / logout for security and
  duration
• Polls every 5 mins back to database
• By tracking PC usage you can tally to get
  utilisation
• Utilisation of a PC then leads into utilisation
  of labs
• Labs leads to campus and so on.

22
Statistics and Reporting

• Usually less then 0.5 of fleet unserviceable
• You could tell most-used PC for a year
• Student logged in the longest
• Most number of logins per year

23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
Where to now with reporting

• Software metering to help determine what is
  base/core software and what's networked to lessen
  image size
• utilisation vs. location for planning next
  installment of PCs
• Most use student to PC ratio
• Best use of licenses

27
Where to now (overall)

• Finalise a scheduling system to book lab
  management work.
• Less monitoring more reporting
• Coupled with email / sms for reports
• Streaming OSs and zero client (No imaging ever)
• Remove my hand from the mouse and go to bed. Even
  more automation and less manual driving.

28
Thanks

• Thanks for listening and I hope you enjoyed
• m.maynard_at_griffith.edu.au