Are cloud based virtual labs cost effective? (CSEDU 2012)
-
Upload
nane-kratzke -
Category
Education
-
view
763 -
download
0
description
Transcript of Are cloud based virtual labs cost effective? (CSEDU 2012)
ARE CLOUD ENABLED VIRTUAL LABS ECONOMICAL?
A Case Study Analyzing Cloud Based Virtual Labs for Educational Purposes
Nane Kratzke
1 Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems
The next 25 minutes are about ...
• Definition of virtual labs
• Analyzed use case • Some use case specific findings about costs • Some use case specific findings about resource usage
• Decision making is not always obvious => How to decide?
• Findings and advices • Establishing virtual labs • Is the concept transferable? • Advices for practical courses • Pragmatic rules for cost control
• Conclusion and outlook
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 2
What is a virtual lab (definition)
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 3
A Virtual Lab is a collection of compute, storage and networking resources provided by an educational organization for educational or research purposes.
It can be provided to a single or a small group of students.
Provided resources are available for
• short-term use,
• accessible via internet and
• are billed by actual resource consumption.
Analyzed use case
• Web technology lecture/practical course for 49 computer science students (bachelor) in summer 2011.
• Project 1: Website for a scientific conference on robotic sailing • Project 2: Google map based automatic sailbot tracking service • All groups were assigned cloud service accounts provided by Amazon
Web Services (AWS). • Analysis of billing as well as usage data provided by AWS.
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 4
Cost analysis
5 Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems
Total costs: 846.99 $
Total students: 49
Cost per student: 17.28 $ CW 13 CW 14 – CW 17 CW 18 – CW 21 CW 22 – CW 25
(A)Costs per Month (aligned to Weeks)
Calendar Weeks (CW)
Cos
ts in
US
D
0100
200
300
400
500
instancehour (62%)
datastorage (34%)
adressing (3%)datatransfer (0%)
(B)Main Cost Drivers
B 1 (5%)
B 2 (7%)
B 3 (7%)B 4 (4%)B 5 (6%)
A 1 (10%)
A 2 (31%)
A 3 (10%)
A 4 (19%)
(C)Costresponsibilty of Groups
(D) Histogram of Costs per Group
Cost Ranges in USD
# G
roup
s
0 50 100 150 200 250 300
01
23
4
Main identified cost drivers:
(1) Server uptime (2/3)
(2) Data storage (1/3)
Server Usage Analysis
6 Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems
13 14 15 16 17 18 19 20 21 22 23 24 25
Average Box UsageMaximum Box Usage in an hour
(A)Maximum and Average Box Usage
Calendar Week
Use
d S
erve
r Box
es
010
2030
4050
13 14 15 16 17 18 19 20 21 22 23 24 25
(B)Accumulated Processing Hours per Week
Calendar Week
Pro
cess
ing
Hou
rs
0500
1000
1500
2000
14 16 18 20 22 24
0.0
0.2
0.4
0.6
0.8
1.0
(C)Average Box to Maximum Box Ratio
according to Weinman
Calendar Week
Avg
to M
ax B
ox U
sage
Rat
io
Training
Project 24x7 Migration
Average to Peak Ratio per Week
7 Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems
13 14 15 16 17 18 19 20 21 22 23 24 25
Average Box UsageMaximum Box Usage in an hour
(A)Maximum and Average Box Usage
Calendar Week
Use
d S
erve
r Box
es
010
2030
4050
13 14 15 16 17 18 19 20 21 22 23 24 25
(B)Accumulated Processing Hours per Week
Calendar WeekP
roce
ssin
g H
ours
0500
1000
1500
2000
14 16 18 20 22 24
0.0
0.2
0.4
0.6
0.8
1.0
(C)Average Box to Maximum Box Ratio
according to Weinman
Calendar Week
Avg
to M
ax B
ox U
sage
Rat
io
Cloud computing is economical reasonable
Cloud computing might be reasonable
Cloud computing is economical not reasonable
Economical Decision Analysis A four step process to decide for or against cloud based virtual labs
Determine your atp ratio
Determine your dedicated costs
Determine your maximal cloud costs
Determine appropriate cloud ressources
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 8
Max instances: 49
Processing hours: 7612
Average: 7612 / (26 * 7 * 24) = 1.74
Overall atp ratio: 1.74 / 49 = 0.035
13 14 15 16 17 18 19 20 21 22 23 24 25
Average Box UsageMaximum Box Usage in an hour
(A)Maximum and Average Box Usage
Calendar Week
Use
d S
erve
r Box
es
010
2030
4050
13 14 15 16 17 18 19 20 21 22 23 24 25
(B)Accumulated Processing Hours per Week
Calendar Week
Pro
cess
ing
Hou
rs
0500
1000
1500
2000
14 16 18 20 22 24
0.0
0.2
0.4
0.6
0.8
1.0
(C)Average Box to Maximum Box Ratio
according to Weinman
Calendar Week
Avg
to M
ax B
ox U
sage
Rat
io
Economical Decision Analysis A four step process to decide for or against cloud based virtual labs
Determine your atp ratio
Determine your dedicated costs
Determine your maximal cloud costs
Determine appropriate cloud ressources
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 9
Example Server: 500 US Dollar Amortization: 3 years
!
d3years(500$) =500$
3• 365 • 24h= 0.019$ h
„In other words, even if cloud services cost, say, twice as much, a pure cloud solution makes sense for those demand curves where the peak-to-average ratio is two-to-one or higher.“ Weinman, Mathematical Proof of the Inevitability of Cloud Computing, 2011
Economical Decision Analysis A four step process to decide for or against cloud based virtual labs
Determine your atp ratio
Determine your dedicated costs
Determine your maximal cloud costs
Determine appropriate cloud ressources
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 10
!
cMax =0.019$ h0.035
= 0.54 $h
According to Weinman the peak-to-average ratio should be greater than the ratio between the variable costs c and your (assumed) dedicated costs d:
Economical Decision Analysis A four step process to decide for or against cloud based virtual labs
Determine your atp ratio
Determine your dedicated costs
Determine your maximal cloud costs
Determine appropriate cloud ressources
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 11
!
cMax =0.019$ h0.035
" 0.54 $h
Exam
ple:
Am
azon
Web
Ser
vice
s EC
2-Pr
icin
gs fo
r EU
regi
on, 1
9th
Mar
ch, 2
012
Economical Decision Analysis A four step process to decide for or against cloud based virtual labs
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 12
A virtual lab provides a more than 25 times cost advantage.
The measured ATP ratio of 0.035 means in fact a 1/0.035 == 28.57 times cost advantage.
This means for the presented use case:
Compared to necessary investment efforts for a classical dedicated lab.
General Advices for Implementing Virtual Labs
• Use your first run to determine the ATP ratio (and make a profound decision pro or contra virtual labs) #1
• Initial training phase (may be even an virtual lab training course) #2
• Cloud computing gives flexibility. Use it! No presence is required (after initial training) #3
• Watch your costs and establish a cost awareness #4
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 13
Mindkeeping cost control rules for students
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 14
motivation rule
Conclusions
• Cloud computing loves peak load scenarios (be happy) • 25 times cost advantage (analyzed use case)
• Cloud generated costs are use case specific (be carefull) • Decision making must not be obvious • Four step decision making model (to determine your ATP ratio)
• Main cost drivers are (try to minimize) • Server uptime • Data storage (server volumes)
• Uneconomical use cases (try to avoid) • 24x7 and • constant loads
• It is likely that cloud based virtual labs might be an economical option for several practical courses, but ...
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 15
Outlook
• Provided cloud management tools are not designed for educational purposes so far • Missing multi account handling • Missing cross region handling • Missing comfortable batch processing of
recurring administrational tasks
• Our focus: Management software to set up and manage virtual labs • Reduce administrational efforts for teachers • Automatic average to peak ratio calculation • Decision support by analysing generated
usage as well as cost data • Mobile access via tablets/smartphones ...
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 16
Thank you for listening
Prof. Dr. rer. nat. Nane Kratzke Computer Science and Business Information Systems 17
Prof. Dr. Nane Kratzke Computer Science and Business Information Systems Lübeck University of Applied Sciences Mönkhofer Weg 239 23562 Lübeck Germany
Mail: [email protected]
Slideshare: i21aneka
XING: Nane_Kratzke
LinkedIn: nanekratzke
WEB: http://praktische-informatik.fh-luebeck.de
You can find this presentation here: http://www.slideshare.net/i21aneka/csedu-2012