Keynote by Diane Bryant, SVP and GM of the Data Center Group at Intel, at OpenStack Silicon Valley 2015.
Cloud computing provides tremendous agility and efficiency to organizations are the driver of the digital service economy. In her keynote, Diane Bryant will discuss how Intel was an early leader in adoption of cloud computing under her tenure as CIO and how this experience has shaped broader strategy to deliver tens of thousands of new clouds across the enterprise with Intel’s new Cloud for All Initiative. Attendees can expect to learn about OpenStack’s critical role in shaping the future of the enterprise data center and learn more about key industry efforts to drive enterprise readiness to the OpenStack platform.
1. OpenStack Silicon Valley | 2015
Accelerating the Next
10,000 Clouds
Diane Bryant
Senior Vice President & General Manager
Data Center Group
Intel
2. OpenStack Silicon Valley | 2015
Cloud Hype Cycle
TECHNOLOGY
TRIGGER
PEAK OF INFLATED
EXPECTATIONS
TROUGH OF
DISILLUSIONMENT
SLOPE OF
ENLIGHTENMENT
PLATEAU OF
PRODUCTIVITY
VISIBILITY
MATURITY
“The cloud is for everyone. The cloud
is
a democracy.”
Marc Benioff (2010)
CEO, Salesforce.com
“Cloud computing:
how computing services will be
delivered in the future.”
Tim O’Reilly (2009)
CEO, O’Reilly Media
“The interesting
thing is that we’ve
redefined cloud computing to include
everything that we already do.”
Larry Ellison (2008)
CEO, Oracle
"It's stupidity.
It's worse than
stupidity: it's a marketing hype
campaign.”
Richard Stallman (2008)
Creator,GNU
“Cloud computing's primary
innovation is its name..”
Glen Morris (2009)
Technology Editor
“...utilities give us
access to clean water... cloud
computing works in a similar fashion.”
Vivek Kundra (2010)
Former federalCIO of the United States
YOU
ARE
HERE
2009
3. OpenStack Silicon Valley | 2015
HPC for Design Engineering
SELF SERVE
CONFIGURATIO
N80%
UTILIZATION
INSTANT
DEPLOYMENT
4. OpenStack Silicon Valley | 2015
Bringing Cloud to the Enterprise
MANUAL
<10% UTILIZATION
90 DAYS
Traditional
Pre-2009
5. OpenStack Silicon Valley | 2015
Bringing Cloud to the Enterprise
MANUAL
20% UTILIZATION
14 DAYS
Virtualized
MANUAL
<10% UTILIZATION
90 DAYS
Traditional
2009Pre-2009
6. OpenStack Silicon Valley | 2015
Bringing Cloud to the Enterprise
SELF-SERVE
40% UTILIZATION
<3 HOURS
Cloud
MANUAL
20% UTILIZATION
14 DAYS
Virtualized
MANUAL
<10% UTILIZATION
90 DAYS
Traditional
20112009Pre-2009
7. OpenStack Silicon Valley | 2015
Built on Open Source
2011
124K
lines of code
3
projects
<100
contributors
TODAY
2.65M
lines of code
41
projects
4,000
contributors
OpenStack
8. OpenStack Silicon Valley | 2015
OpenStack Today
2009 2010 2011 2012 2013 2014
<10% of enterprises have deployed cloud
Cloud Market Growth
IntelCPUSales
Private Cloud
8%
Public Cloud
92%
9. OpenStack Silicon Valley | 2015
OpenStack Today
2009 2010 2011 2012 2013 2014
<10% of enterprises have deployed cloud
Cloud Market Growth
IntelCPUSales
Private Cloud
8%
Public Cloud
92%
FRAGMENTED solution stacks
COMPLEXITY in deploying solutions
LACK of key features
Gaps Remain
10. OpenStack Silicon Valley | 2015
Jevons Paradox
Leads to
NEW USAGES
Increase in
TECHNOLOGY
EFFICIENCY
Increases
rate of
CONSUMPTION
11. OpenStack Silicon Valley | 2015
Intel® Cloud For All
Unleash Tens of Thousands of New Clouds
in creating enterprise ready, easy to
deploy cloud solutions
for high efficiency clouds across
workloads
the industry to accelerate cloud
deployments
Invest Optimize Align
12. OpenStack Silicon Valley | 2015
Deliver enterprise ready OpenStack
to scale to 1000’s of nodes
Build largest joint developer team
to accelerate upstream contributions
Create largest developer cloud
to build and validate solutions at scale
Other names and brands may be claimed as the property of others.
13. OpenStack Silicon Valley | 2015
Enhance enterprise readiness
to ease adoption and operations
Deliver performance at scale
to ensure efficient and effective OpenStack deployments
Implement network & storage stack improvements
to enable scalable & resilient solutions
Other names and brands may be claimed as the property of others.
14. OpenStack Silicon Valley | 2015
Boris Renski
Co-Founder &
Chief Marketing Officer
Paul Voccio
Vice President Software
Development
Other names and brands may be claimed as the property of others.
15. OpenStack Silicon Valley | 2015
Boris Renski
Co-Founder &
Chief Marketing Officer
Paul Voccio
Vice President Software
Development
Other names and brands may be claimed as the property of others.
16. OpenStack Silicon Valley | 2015
Building upon existing collaborations
Rackscale
Architecture
Open Container Initiative
Clear
Linux
Project
Cloud Native Computing Foundation
Standards & Initiatives
Recent Collaboration Announcements
Other names and brands may be claimed as the property of others.
17. OpenStack Silicon Valley | 2015
Future Clouds
Traditional
System 1
App
System 2
App
Virtualized
App
Virtualization
System 1 System 2
Virtualization
App App App
Cloud
Orchestration
System 1 System N
Virtualization Virtualization
App App App App App App
Hyperscale Cloud
Orchestration
App App App App App App
Containers /
Processes
Containers /
Processes
Resource
Pool
Resource
Pool
= Resource Utilization
Bringing hyperscale to the masses
18. OpenStack Silicon Valley | 2015
Future Clouds
Bringing hyperscale to the masses
EFFICIENT application development
EFFICIENT hardware
EFFICIENT schedulers: bare metal to VM to containers
EFFICIENT execution across all clouds
Hyperscale Cloud
Orchestration
App App App App App App
Containers /
Processes
Containers /
Processes
Resource
Pool
Resource
Pool
19. OpenStack Silicon Valley | 2015
Summary
Enable the INNOVATION cycle
Deliver greater EFFICIENCY through cloud computing
http://www.forbes.com/sites/joemckendrick/2013/03/24/10-quotes-on-cloud-computing-that-really-say-it-all/
“The interesting thing about cloud computing is that we’ve redefined cloud computing to include everything that we already do. I can’t think of anything that isn’t cloud computing with all of these announcements. The computer industry is the only industry that is more fashion-driven than women’s fashion. Maybe I’m an idiot, but I have no idea what anyone is talking about. What is it? It’s complete gibberish. It’s insane. When is this idiocy going to stop?” – Larry Ellison, chairman, Oracle, referring to the term "cloud computing" in his Oracle OpenWorld 2008 speech
“I don’t need a hard disk in my computer if I can get to the server faster… carrying around these non-connected computers is byzantine by comparison.” – Steve Jobs, late chairman of Apple (1997)
“If you think you’ve seen this movie before, you are right. Cloud computing is based on the time-sharing model we leveraged years ago before we could afford our own computers. The idea is to share computing power among many companies and people, thereby reducing the cost of that computing power to those who leverage it. The value of time share and the core value of cloud computing are pretty much the same, only the resources these days are much better and more cost effective.” – David Linthicum, author, Cloud Computing and SOA Convergence in Your Enterprise: A Step-by-Step Guide
“Ultimately, the cloud is the latest example of Schumpeterian creative destruction: creating wealth for those who exploit it; and leading to the demise of those that don’t.” - Joe Weinman, Senior VP at Telx and author of Cloudonomics: The Business Value of Cloud Computing
it’s worth keeping in mind that in the cloud we’re all guinea pigs, and that means we’re all dispensable. Caveat cloudster.” – Nick Carr, author of Does IT Matter?, The Big Switch and The Shallows
“Cloud computing is often far more secure than traditional computing, because companies like Google and Amazon can attract and retain cyber-security personnel of a higher quality than many governmental agencies.” - Vivek Kundra, former federal CIO of the United States
She asked the team to start in 2009, but I think it might be useful to know that we had a running Grid in 1997, and we had it globally connected around 2005.
Resource pools (ways to make pools larger & larger -> easier to share when the lake is larger)
The better we got at sharing capacity, the better we were at driving up capacity
Small team that had to scale quickly, so HAD to automate everything
Enterprise had just a couple pools
Lets move Automation after Virtualization… for IT Virtualization put 14 days for Time to Deploy, with Automation drop it to 5 days (and the key here is that the automation was actually pretty quick, the issue was the time waiting in the ticket queue for IT)… Then Gen 1 IT Private Cloud is <3hrs, Gen 2 which you can’t fit is basically instantaneous.
Utilization first went up from <10% to 20% when we virtualized... Then up to 40% when we made it a private cloud.
Lets move Automation after Virtualization… for IT Virtualization put 14 days for Time to Deploy, with Automation drop it to 5 days (and the key here is that the automation was actually pretty quick, the issue was the time waiting in the ticket queue for IT)… Then Gen 1 IT Private Cloud is <3hrs, Gen 2 which you can’t fit is basically instantaneous.
Utilization first went up from <10% to 20% when we virtualized... Then up to 40% when we made it a private cloud.
Do you know what the $$ value of private cloud deployment has been for Intel?
I am pretty sure it is $25M NPV, I will close on the validity on Monday.
7
Fragmentation of solution stacks: End user required to navigate too many and complex API; Limited standard (open) interfaces between legacy/traditional & software defined infrastructure
Complexity of deploying solutions: Requires too high a technical acumen to design & implement (Network overlays, authorization & authentication); Integration of legacy/traditional infrastructure & applications with next generation infrastructure into brownfield datacenters; Existing enterprise infrastructure & applications and operations is not well documented
Lack of mature enterprise features: Rolling upgrades (version upgrades without downtime); Metric driven resource elasticity (auto scaling up and down); High availability of tenants & services (hardware or service failures trigger automatic failover); Overarching management tool for compute, storage, network & services (Ecosystem maturity like VMware and AWS)
High efficiency and scale difficult to achieve: The complexity of manually “bin packing” to achieve optimal efficiency requires the characterization off every workload factored with the memory, CPU and cache capacity size of a heterogeneous set of server; Operational productivity is hampered by the lack of the necessary telemetry data and automated management; Today’s scale in measured thousands of physical hosts (Everything has maximum limits memory, storage, vCPUs, network connections) where cloud native applications require scale to 10s of thousands of hosts in multiple data centers. http://www.vmware.com/pdf/vsphere6/r60/vsphere-60-configuration-maximums.pdf
Fragmentation of solution stacks: End user required to navigate too many and complex API; Limited standard (open) interfaces between legacy/traditional & software defined infrastructure
Complexity of deploying solutions: Requires too high a technical acumen to design & implement (Network overlays, authorization & authentication); Integration of legacy/traditional infrastructure & applications with next generation infrastructure into brownfield datacenters; Existing enterprise infrastructure & applications and operations is not well documented
Lack of mature enterprise features: Rolling upgrades (version upgrades without downtime); Metric driven resource elasticity (auto scaling up and down); High availability of tenants & services (hardware or service failures trigger automatic failover); Overarching management tool for compute, storage, network & services (Ecosystem maturity like VMware and AWS)
High efficiency and scale difficult to achieve: The complexity of manually “bin packing” to achieve optimal efficiency requires the characterization off every workload factored with the memory, CPU and cache capacity size of a heterogeneous set of server; Operational productivity is hampered by the lack of the necessary telemetry data and automated management; Today’s scale in measured thousands of physical hosts (Everything has maximum limits memory, storage, vCPUs, network connections) where cloud native applications require scale to 10s of thousands of hosts in multiple data centers. http://www.vmware.com/pdf/vsphere6/r60/vsphere-60-configuration-maximums.pdf
Put context to the Mirantis deal as it pertains to Rackspace (Rackspace focuses on upstream maturity and scale, Mirantis focuses on specific areas of need for enterprise adoption)
Talk about the opportunities for enhancing enterprise readiness to ease adoption and operations
As part of the Intel Cloud for All announcement on July 24, Intel is increasing our investment in Mirantis via a Series B infusion. The total round was ~100M and included investors such as Goldman Sachs, August Capital, Insight Venture Partners, Ericsson, Sapphire Ventures (formerly SAP Ventures) and WestSummit Capital. This investment was based on the intent to mature downstream distributions of OpenStack, focusing on the following technical aspects:
o Performance at scale
o Bare metal provisioning
o Ceph storage improvements
o Enterprise readiness acceleration
o Network stack improvements
o Big data optimized OpenStack solution delivery
Cloud Integrity Technology – TXT
Bringing hyperscale technologies to mainstream environments (concepts born on hyperscale)
Cloud Native Computing Foundation – enabling software developers to easily deploy containers
give me a server {VM, OS, etc} give me a container
Why would an app developer want to use a container?? – AR: Das to highlight pros/cons
Cloud Integrity Technology – TXT
Bringing hyperscale technologies to mainstream environments (concepts born on hyperscale)
Cloud Native Computing Foundation – enabling software developers to easily deploy containers
give me a server {VM, OS, etc} give me a container
Why would an app developer want to use a container?? – AR: Das to highlight pros/cons
Build hyperscale efficiency to the masses – the ability to rapidly deploy apps at scale, enable more apps to be containerized, address network and storage limitations, etc.
Enable schedulers that support from bare metal to VM to container - rather than independent fiefdoms, allow all apps and resources to be scheduled with common view
Drive cloud efficiency via rapid hardware innovation – clouds that can support heterogenous computing across microservers to Xeon phi, that can support FPGAs, Next generation memory technology and manage the hardware innovation seamlessly
Integrate cross- cloud – deploy on prem / in cloud / different service providers with a common view based on where it makes the most sense to deploy