elemental links

@ Interop’s Enterprise Cloud Summit: What Elasticity really means

What Elasticity Really Means

Moderator:
Ian Rae, CEO, Syntenic

Speakers:

• Jon Beck, SVP Sales and Client Services, OpSource, Inc.
• Scott Clark, Director of Engineering Infrastructure, Broadcom
• Josh Litwin, President and CEO, Memento Press
• Geir Magnusson, Consulting Architect, Platform, Gilt
• Chad Swartz, Senior Manager, IT Operations, Preferred Hotel Group

Finally, some cloud computing practitioners, small business, enterprise, web retail and technology – Memento Press, Preferred Hotel Group, Gilt and Broadcom. 

Alistair Croll mentioned that most enterprises are reluctant to go public with their cloud stories right now, it’s early, and many enterprises don’t want to give away secrets. [This happens in every wave.  I speak with shy enterprises as well.]

Drivers:

- Small Business, 1 person IT, find SaaS to help run business

- Enterprise, viable option to expanding data center capacity.  Looking for solution to leverage other’s experience, economy of scale, in way that accounts for investment.  Pay as you use, for peak periods, is compelling idea for them.

- Web Retail, classic 0 to 60 startup story. 

- Technology, best investment option for planned growth.  Didn’t go with a complete pay-as-you-go plan.  Contracted for a fixed pool of resource that they can use in whatever way see fit.  The fixed pool has base plan, overage is pay-per-use, have opportunity to adjust plan.

Interesting points regarding Elasticity:

- OpSource, most customers are more concerned with ability to scale-down, rather than scale-up.  Don’t want to pay for unused capacity.  But, want to know capacity is available.

- Preferred Hotels emphasizes the need to evaluate/change application distribution and infrastructure setup for performance.  You might end up using more server instances in the cloud to get the same performance.  If so, need to adjust your software licenses accordingly.

- Gilt’s biggest concern is horizontal scale.  Were concerned with ‘Oprah effect’ on Gilt infrastructure, needed to ensure scale and performance.  Mentions challenges of database scaling.  Really need to look at applications and prepare them for scale.

- Broadcam, there is an appeal to participate in something bigger than them.  Leverage ecosystem, economy of scale, really smart people.  Foresees purpose built clouds, for a business activity.

Recommendations:

- Gilt, despite all the hype, this (cloud) is a fundamental change.  Application architecture is changing.  Operations management is changing.  Need to understand where cloud is going.  The hype will end, the changes will stay.

- Preferred Hotels, this is new era.  Doesn’t make sense to build-out and manage these computing environments.  Starting points: virtualization or web applications.  States “make commitment to virtualization”.  Calls out agility, time to setup new environment, as key return. 

Audience Q&A

Q: How do you know when you need to spin up a new server?

A: Transparency and monitoring – Opsource

A: Looks like ‘regular’ server, can see highs and lows, manage as ‘regular sever’. – Gilt

Q: Plan servers?

A: Have predictable demand, spikes at noon, build up before, drop off after.  Base is steady demand, elasticity for peak

A: Over provision individual servers, common configuration, allows for agility and ease of management.

Q: Application changes for performance on cloud?

A: Not the cloud, it’s virtualized environment that pushes need for change.  Rebuild servers, not necessarily applications.  But, applications (and application architecture) need to be reviewed.

My bottom line interpretation: Scale does not equal performance!

@ Interop’s Enterprise Cloud Summit: On demand economics panel

This morning’s focus in on the business case for cloud computing, ROI & migration.  "What's the promise? Despite the downsides, there's little doubt that on-demand computing services are the future of both consumer and enterprise IT.”

First up in a panel discussion: The Case for Cloud Infrastructure: On-Demand Economics

Moderator:
John Willis, Owner, Zabovo

Panelists:

• Paul Mockapetris, Chairman and Chief Scientist, Nominium, Inc.
• James Urquhart, Cisco, Technology Strategist, Data Center Solutions team
• Neil Cohen, Director of Product Marketing, Akamai
• Joe Weinman, Strategic Solution Sales, AT&T Signature Client Group

Positions/fascinations with Cloud

James Urquhart – it’s a big deal.  But, a big deal that is going take some time to evolve.  Educate yourself on opportunities and risks related to cloud computing.

Paul Mockapetris – cloud computing is problem of showing no seams at all.  security is challenge, faster you go, faster DOS attack can happen.

Neil Cohen – run applications at edge of internet, apply economics of cloud to applications you have running today that might not be able to move to the cloud.

Joe Weinman – synaptic storage management announcement yesterday.  Finds area fascinating, so many things sound plausible, but not everything adds up.  Can’t look at cost of internal vs. cloud compute alone, need to delve into drivers, might make sense to pay more per transaction for a scale transaction versus buying & managing equipment for partial (occasional) use.

ROI

Joe – Real benefits lie in spiky demand, “own the base, rent the spike”

James – Eli Lily story, recent webcast, spoke of using Amazon for research activities.  Researchers came back and said “now, we need to rethink the way we do research projects”.  Not having to acquire, setup, teardown thousands of servers gives the researchers flexibility in how, when, frequency to do work.  [This is real agility].

Look at jobs with varying elasticity, batch jobs, data analysis.

John Willis – What’s missing in cloud story, and reports like McKinsey, is the “thinking outside the box” opportunities, such as Eli Lily.  Speaks of story of organization moving data analysis to cloud on Hadoop because in-house Oracle dba not available.  Now, IT (dba) exploring other opportunities on cloud, with Hadoop.

Some questions from the audience.  Most are self-proclaimed cloud-newbies.  (honest people)

Q: Is there an ROI calculator available?

Joe: Created one, it is at complexmodels.com

Neil: Reminds folks that it’s not a direct port to the clouds.  Not all applications are cloud ready, and some (legacy) just can’t be moved.  So, need to augment ROI calculations with implications for your particular application.  [Dave Bressler did a good job addressing integration costs and the cloud at the Cloud Computing Expo]

Security?

James – the #1 enterprise concern for cloud computing is trust.  ROI won’t matter if Trust issue isn’t solved.  Recommends Hoff’s Rational Survivability blog to learn about Security & Cloud Computing.  [me too]

ROI related to applications & PaaS?

Neil: The key metric is ‘time to value’.  Can you deliver quicker with PaaS?  Will the delivered application meet the business requirements?

Joe: better to 50% over budget than 50% late.  Anything that allows you to deliver applications more quickly is of benefit.

@Cloud Camp @Interop: Un-conference, Un-panel

This evening, we are at Cloud Camp.  For background, check out this interview with Cloud Camp co-founder Dave Nielsen.  Dave is also our host today. 

The lightening rounds are done.  For insights, search on twitter with either of these tags: #cloudcamp #cloudcampinterop

Now, Dave is building an un-panel session.  He started with 5 empty chairs and no questions.  Filled the chairs with folks who raised hands to “Folks that know a lot about cloud computing”.  No one admitted expertise.  Now, he’s asking the audience to build a list of 10 questions.

Interesting, not all of the panelists have identified themselves.

1. Will there be ‘A Big Switch’? 

2. Will efforts like CSA (Cloud Security Alliance) help or hurt cloud computing?

What will happen is same as SOA, if there is a really easy way to do something, that will prevail over an standards industry effort.  Think REST vs. WS-deathstar.  -James Urquhart

3. Skills to become cloud computing expert?

Define things in terms you already know.  Cloud is an operational model: self-service, on-demand, multi-tenant etc.  How to apply this operational model to what you are doing today.  -James Urquhart

4. Standards needed for maturity?

John Willis, interesting things with standards, DMTF, hypervisor and virtual images.  Standards about provisioning.  How to provide standards on configuration?  DMTF is good place to go.  Need to start with APIs.  Standards could be driven by whoever succeeds first. 

5. Legal risks for a corporation with Cloud Computing?

Secure communications act.  Patriot act – Canada won’t store data in US.  FBI seized all servers in server farm, based on suspicious actions of one tenant.  These are really Internet issues, but cloud computing accelerates this.  Legal system needs reform for cloud computing.  -James Urquhart

6. Risk mitigation?

Alternate provider redundancy.  [Rightscale representative]

7. Move large data to cloud?

8. Will we wake up one day, 5 years from now, and say “oh crap, why did we put our data in the cloud”?

9. Does Cloud Computing put IT Service Providers out of business?

10. Where should enterprise start?

John Willis, try stuff out to get familiar with capabilities and challenges.  Be smart about what you pick.  Also, consider private clouds.

11. Bonus: is Cloud Computing a useful term?

Useful now, might go away in future, as cloud computing becomes business as usual, aka “computing”.  John Engates

Panelists pick and answer questions.  [See commentary under questions].  If audience deems “answered”, good to go.  If not, question will be discussed in an “un-conference” breakout session.

Questions 1, 8 & 9 were answered, but by folks with financial stake in outcome.  These could be interesting questions to consider at another time, with a group of business & technology futurists.

Now, there is the prep & transition to the un-conference session.  Probably tweet from there, back to blogging tomorrow at Enterprise Cloud Summit.

@ Interop’s Enterprise Cloud Summit: Drew Bartkiewicz, The Hartford, Cloud Insurance

Session: Security, Risk, Legislation, and Compliance: Pandora's New Box. 

Speakers:
Drew Bartkiewicz, Vice President of Cyber Risk and New Media Markets, The Hartford
Robert Parisi, SVP & National Technology, Network Risk & Telecommunications Practice Leader, FINPRO, Marsh USA

I happened to sit next to Drew this morning, and we chatted a little on what he’s doing at The Hartford.  Drew isn’t in IT.  He runs the business line that covers (insures) the cloud operators, and it sounds like, in the near future smart cloud consumers.  He mentioned a phenomenal month-to-month business growth, which makes me believe that not only is the cloud real, but operators are serious about protecting against business risk, for themselves and their customers.

Drew speaks to “Information malpractice”.  It is something that all businesses need to protect themselves against.  Drew says he’s not here to be “wet blanket”.  He’s a believer in the economic benefits of cloud computing, even with the additional costs of insurance. 

Drew shares that the advances of cloud computing are outpacing CIOs, acquisition directors, lawyers and legislators ability to interpret, protect against and regulate the associated risks.  

“With great power there must also come – Great Responsibility”.  - Stan Lee, writer of SpiderMan, 1960’s

Cloud Uncertainties, there is an Aggregation of Risk – for data, for interdependence, for IP

Contractual considers – data privacy, liability, interdependence.  On interdependence, even if you have great terms of use, if someone “misbehaves” in your environment, and causes issues for other tenants, you the cloud operator can be held responsible.  Reasons: your enforcement of terms of use, in terms of stringency and consistency; or, because you have the deepest pockets.

Calls out two books: Unseen Wealth, by the Brookings Institute on intangible information assets as the future for wealth creation and his upcoming book, Unseen Liability , which speaks to the risks of collecting, using and managing this information.  These risks need to be factored in as organizations assess the viability of bringing a new service to market.  [This applies to traditional “providers” and organizations/enterprises who choose to offer new services via the cloud.]

Drew says “select cloud provider judiciously”.  Stay away from providers who aren’t willing to discuss contractual, legal, privacy, compliance concerns.

Summary:

• Think Open, Think Smart. 
• Manage your points of risk with the cloud.
• Insure for uncertainty, or make sure your cloud providers are insured.
• Cloud Insurance exists

Q: Doesn’t existing errors & omissions policy cover this?

A: Not necessarily.  Doesn’t always cover E&O by third party providers (cloud operators) or that impacts non-clients.  For example, if your mistake damages multi-tenant cloud, you’ve created an error for people you don’t have any relationship with.

“Traditional Insurance was created for a world that no longer exists”. 

For me, this is an important topic for enterprise adopters to understand.  Not in the sense of “collecting reasons to ignore cloud computing”, but in the spirit of making informed decisions, and in protecting yourself, your company and your customers.

@ Interop’s Enterprise Cloud Summit: Panel on Where can things go wrong?

Where Can Things Go Wrong?

Moderator: Greg Ness, Marketing Executive, Infoblox

Panelists:

• Peter Coffee, Director, Platform Research, salesforce.com
• Randy Rowland, General Manager, Managed Hosting & Cloud Computing Services, Terremark Worldwide, Inc.
• Geva Perry, Founder, Thinking Out Cloud
• Bill McGee, Vice President, Products and Technology, Third Brigade

Panelist Points of View on Cloud Computing:

• Greg Ness – Networks, resources and monitoring
• Randy – Enterprise Cloud Platform offering
• Peter Coffee – Salesforce
• Bill McGee – Security perspective, private and public cloud
• Geva Perry – Bigger picture, Thinking out Cloud blog

Some highlights from the discussion follow…

General comment:

Peter – the problem of the cloud is anything being offered with price tag of zero, while some things are being offered with a price commensurate with the value.  Look for the latter.

Critical Cloud dependencies:

Geva: lock-in, security (although often same issues exist in enterprise). Want to consider who will eventually become hub of ecosystem, look for that player.  Geva has related blog post.  Ease of integration.

Bill: Cloud vendors (IaaS) will need to offer basic, and higher levels of security.  Higher level – physical, server, management console – is for customers (enterprises) who value security more [business requires stricter security]

Peter: Amplifies first two answers, ease of integration and strong security.  More integration, more trust comes into play.  Peter says in multi-tenant environment, everyone benefits from the security strictness of the most demanding customers.

Randy: Reliability.  Mistake people make is assuming just because something is in the cloud means it is backed up somewhere, on a cluster, provides business resilience.  Unless building brand new app that doesn’t need access to anything on premise, need to understand, plan for connectivity.  If you are looking a cloud computing as answer to enterprise IT needs, need to really work contract, compliance, know you can survive an audit.  All can be overcome.

Peter: Conversation shouldn’t be “cloud is not secure, won’t use it” or “cloud can’t integrate, won’t use it”.  Need to say, if going to the cloud, what is cost for our level of security, or integration.  Typically, people find the additional cost is not significant.  Cloud ends up being viable solution.

Geva: Lock-in, the more things are limited and specific to vendor, tightly knit, more “perfect” the solution is for the thing it was built for.  Specialized, easy to use.  Trade-off with lock-in. 

Security – different solutions emerging to solve issues.  Not necessarily from the big cloud vendors, but from the ecosystem players.  Cites enStratus as an example.  Important to consider the niche players.

Stark differences between standard IT and cloud services:

Randy – agility, how fast can IT respond to the business.  Business will go directly to SaaS providers for solutions.  [Shadow IT threat story]

Peter – Two ways to mess up: (1) non-enterprise class SaaS solutions adopted by business in silos (2) re-create traditional IT environment and practices in the cloud; need to do something in the middle.

Bill – understand if existing software has access points to be successful

Geva – automation increases self-service, adds to agility story for enterprise