Fingerprinting the Datacenter: Automated Classification of
Download
Report
Transcript Fingerprinting the Datacenter: Automated Classification of
Fingerprinting the
Datacenter
Offense
Mykell Miller, Gautam Bhawsar
Crisis Detection
Why don’t you detect unidentified crises?
How severe is each crisis?
May affect recovery method
May affect detection probability and method
Useful for operators to know
Fingerprints are averaged over epochs in
the crisis
Everywhere else you use the median
Time to identification
What are you doing for 10 minutes?
Why have the time between epochs be
15 minutes?
Will shorter lead to more accurate or quicker
identification?
Data Center
What is the configuration?
Fat tree, etc.
What kind of servers?
Type of servers affects type and frequency
of crises
Using only Windows
Study cannot be applied to other OS
Data Set
Only 19 examples
8 types of crises had only 1 instance each
Training set of only 5 examples
No statistics on the user application
being run
Don’t know if it’s representative