Upload
sergey-a-razin
View
325
Download
0
Embed Size (px)
DESCRIPTION
Traditional datacenter is broken up into a number of silos: network, storage, virtualization, and application. The emerging Software Defined Datacenter movement breaks those silos and creates a playground for innovation, convergence, and new opportunities to reveal the hidden and unknown. During this session I will describe what the Software Defined Data Center hype is all about, how it breaks the traditionally established silos while creating the opportunities for data driven orchestration powered by Machine Learning principals that will make cloud providers and enterprises FINALLY realize the value of virtualization and ultimately deliver the self-driving data center where open initiative is at the front and center.
Copyright © 2011 SIOS Inc. | SIOS Confidential Information
Self-Driving Data Center The Nugget of Software Defined Movement
About SIOS
• Founded in 1999
• SIOS Technology Corp. is a wholly owned subsidiary of SIOS Technology, Inc., a Tokyo based, publicly traded company
• Worldwide presence with offices throughout the US, UK and Japan
• Over 50,000 licenses installed globally
• Leading Technology Partnerships
Out of the Box Linux Knowledge
Services
• SAP
• WebSphere MQ
• Any Custom App
Applications • Apache
• Samba
• NFS
• LVM
• SW Raid (md)
Databases Storage
• Oracle
• MySQL
• PostgreSQL
• Sybase
• DB2
• Informix
• DMMP
• NAS
• EMC PowerPath
• Hitachi HDLM
• IBM SDD
• Data Replication
Protect
Any
Application
Technology Trends
Hardware Commoditization
Convergence
CLOUD BigData
Technology
Business Model
Technology Trends
Hardware Commoditization
Convergence
CLOUD BigData
Technology
Business Model
Customer Problem
Reduce Capex and Opex while
achieving scale, performance, and
availability.
Software
Black Box
Agility and Flexibility
Customer Problem
Cloud and Virtualization
Data Center Today Storage Network Cloud/Virtualization
Software
Black Box
• Has scale and capacity
• Proprietary Solution (Software and Hardware)
• Not heterogeneous
• Expensive
• Restrictive (vendor lock-in)
Data Center Emerging (Open Source Foundation)
Storage Network
Software White
Box
• Has scale and capacity
• Open Solution (Software and Hardware)
• Heterogeneous
• Cost-effective
• No vendor lock-in
Cloud/Virtualization
Data Center Emerging (Open Source Foundation)
Cloud/Virtualization
Software Defined from the
beginning and causes to the
disruption in silos underneath.
Data Center Emerging (Open Source Foundation)
Storage
Blog post: http://blog.techdozor.org/index.php/2013/09/30/san-vs-opensource/
• Distributed block storage takes each local storage
system and, in much the same way as RAID
combines multiple drives into one single array. No
SPOF, point of management locus is centralized
• Enterprise level features: scalability, thin-
provisioning, snapshotting, cloning, tiering, caching,
HA/DR
• Performance: ability to spread the load from a heavy
use virtual drive across multiple storage nodes in a
cluster.
• Finally: All of the above is all about CapEx, OpEx,
and TCO
Storage
Network
Data Center Emerging (Open Source Foundation)
• What is it? It is a centralized management of the
network infrastructure that enables automation, analytics, and orchestration of network services via the standardized interfaces.
• Why SDN? Efficiency – optimized existing
application, services, and infrastructure
Scale – ability to keep up with the dynamic scaling of application and corresponding services
Innovation – from “black box” to
Why Software Defined *
• To achieve ITaaS, i.e. all aspects of infrastructure
are virtualizaed and delivered as a service
• Make application infrastructure aware for
automation, provisioning, etc.
Is Software Defined * DEAD? What’s Next?
Hardware Defined *?
Back to the problem at hand
Cloud and Virtualization
Back to the problem at hand
Back to the problem at hand
Physical Servers
Public
Cloud
Private
Cloud
Hybrid
Cloud
Hosted Private
Cloud
Virtualized Servers
On-Premise Off-Premise
Back to the problem at hand
Storage Network Cloud/Virtualization
API API API
Software Stack Software Stack Software Stack
Gaps in Current Approaches
Gaps in Current Approaches
Gaps in Current Approaches
Converged Predictive analytics is GAP
Data Complexity
We see it coming – Google Alerts
2014
IT Operations Analytics (born 2014)
• Source: http://www.itoa-landscape.org/
Analytics powered by Machine Learning
Analytics and Machine Learning
Next Generation Datacenter
Five Key Requirements for ML App
Presentation
Scalability is the KEY!
Adoptability, and Portability
Data Strategy (security, privacy, and governance)
Algorithm and Data Normalization
Do It Yourself!
http://www.r-project.org/
http://www.cs.waikato.ac.nz/ml/weka//
http://madlib.net/
https://mahout.apache.org/
Questions?
Contact Info
Sergey A. Razin, Ph.D.
Chief Technology Officer
SIOS Technology Corp.
Email: [email protected]
LinkedIn: http://www.linkedin.com/in/sergeyrazin
Follow me: http://www.twitter.com/TechDozor
Read me: http://blog.techdozor.org
Thank you!