Virt Cloud.4

ITEC - TOWARD CLOUD COMPUTING

Dan Brint: [email protected]

OAS

1Sunday, November 21, 2010

mailto:[email protected]


AGENDA

Retrospective

Operations and SOPs

Performance Engineering

‘Cloud’ Computing

ITEC’s Cloud Strategy


RETROSPECTIVE

•A little history (close your eyes for 3 minutes if you must)


•Services @ITEC (a +20 year evolution)

• We now directly provide technical support for Confluence, @task and Elluminate as u-wide programs.

RETROSPECTIVE

!   "#$%&'(!   ")'#$*+,(-./0'1/(

!   -234$#'(56/0#678*2+(

-8))2#0(

!   "#$%&'(!   ")'#$*+,(-./0'1/(

9'120'(-'#:6%'( !  -08;'+0(<+=2#1$*2+(

-./0'1/(!  >8/6+'//(<+0'&&6,'+%'(!  ?67#$#.(9'%2#;/(@,10(!  56/0$+%'(?'$#+6+,(!  A+0'#)#6/'(B<C<(!  A+0'#)#6/'(D#2E(@,10(

F2/*+,(


•Virtualization @ITEC

•Storage (2000): EVA, Shark, DS4500, HP24000, Equallogic

•Server (2002): pSeries,VMWare, OVM

•Network (200X): vlans, multiple ISPs, BGP

RETROSPECTIVE


•Why did we virtualize?

• Hardware, power & cooling savings via consolidation

• Efficiencies in technology management

• Take advantage of the educational cycle (LMS, Library, SIS)

• Consolidation of not only technology but skill sets and expertise as well

RETROSPECTIVE


•Our Banner on RAC POC

• Working with SUNY’s strategic partners

• Proved we could run Banner on virtualized / clustered Database and Application Server technologies

• but ...Database Cluster (RAC)

RAC Node1

db1

RAC Node2

db4

RAC Node3

db4

Banner on Linux n-tier

db2 db2

db5

db3 db3 db3

WWW 01

Binghamton

Buffalo StateDutchess

JCCNassau

SuffolkDB10DB 9 SICASDB 8 OldWestbur

DB 1 Binghamton

DB 7 NiagaraDB 6 NewPaltz

DutchessBuffalo State

JCCNassau

DB 2DB 3DB 4DB 5

Web tierINB and SSB

Banner tiergurjobs, command line ‘stuff’

Database tierdatabases

RAC NodeY

db10

db9

RAC NodeZ

db2

db1 db1

db10

Old WestburyNewPaltz

NewPaltzNiagara

Old Westbury

WWW 02

Binghamton


JCCNassau

NewPaltzNiagara

Old Westbury

WWW 03

Binghamton


JCCNassau

NewPaltzNiagara

Old Westbury

WWW x

Binghamton


JCCNassau

NewPaltzNiagara

Old Westbury

WWW y

Binghamton


JCCNassau

NewPaltzNiagara

Old Westbury

WWW z

Binghamton


JCCNassau

NewPaltzNiagara

Old Westbury

Niagara

Binghamton Buffalo State Dutchess JCC

SICASSuffolk

SICASSuffolk

SICASSuffolk

SICASSuffolk

SICASSuffolk

SICASSuffolk

SICAS Suffolk

Nassau


•Objective look at our apps and business needs

•No single app needed more compute power beyond a single x86 server and no apps had true 24 x 7 x 365 uptime requirements

•Chose database ‘virtualization’ via OVM instead of RAC

• DBA team hesitant to jump into an emerging technology• Systems team confident and proficient with VM technology

RETROSPECTIVE


RETROSPECTIVE

•Let’s put that in our mental parking lot for a moment


OPERATIONS / SOPS

•Towards a more consistent datacenter management experience


•Enterprise Manager as a tool for managing technology

•Engaged TUSC to help with template development

•Delivered Dashboards to SLN helpdesk (finally)

OPERATIONS / SOPS


•@task as a tool for managing tasks

•Confluence as a tool for managing knowledge

OPERATIONS / SOPS


•Adoptions of Project Managers and project management methodologies

•Organizing our projects and business processes

•Reviewing and standardizing our operating procedures

OPERATIONS / SOPS


•Long and short of it is that we are improving our services AND making our own lives more tolerable by documenting, standardizing and proceduralizing

•Only after we have achieved this can we really look to automation to take it to the next level

OPERATIONS / SOPS


OPERATIONS / SOPS

•Put that in the parking lot too :)


PERFORMANCE ENGINEERING

•Some enlightening work with SLN & ANGEL / Blackboard


•As a result of the ‘overwhelming’ success of ANGEL adoption during fall 2009 we sought to understand

•SLN engaged Blackboard’s performance engineering team to help us better understand the past and predict the future



• Analyzed our web logs

• Determined peak loads (peak of peak)

• Constructed simulations to test capacity, throughput, ...

• Performed tests on a lab representing our new environment to find potential bottlenecks & efficiencies

•FYI - No silver bullets (unexpected growth)



•After the analysis was complete we performed what blackboard calls a TKT - a technical knowledge transfer.

• I waited with bated breath* for this !!!


The phrase means anxiously or with great anticipation.I don't remember and can't find my source on this but I think that "Bated" is a shortened version of "abated", which means "to slow down". In the case of "bated breath" this would mean to slow down your breathing or hold your breath. Curiously, people hold their breath when in anticipation. Perhaps so as not to be distracted by breathing."Bated" is no longer commonly used, causing people to believe the expression to be "with baited breath". This common misspelling leads to confusion and strange imagery.

http://www.phrases.org.uk/bulletin_board/5/messages/14.html


“ It’s as though someone put the bloody SCIENCE back in Computer Science !!!!

You know theories, hypothesis, tests, analysis, conclusions ... all that good stuff we did back in school ! ”



•A couple of big takeaways from the TKT:

•The science of performance engineering

•Emphasis on process, language, communication

•The resource commitment to implement

•Potential as an internal resource we consume, a value add to our current customers, a standalone service we offer for anyone in SUNY.




•Another one goes into our mental parking lot


CLOUD COMPUTING

•Good lord he went 15 minutes without saying cloud - now we’re in for it ???!!!

•Time for some definitions


•Melissa Woo (University of Wisconsin-Milwaukee )

• “I find the term cloud computing to be extremely overloaded. In fact, a recent report by McKinsey & Company identified twenty-two different definitions for cloud computing.”

• “A not-yet-perfect storm of technologies that allows one to consume services in blissful ignorance of what's under the hood and how to maintain it.”

CLOUD COMPUTING


•Larry Ellison (Oracle)

• "I can't think of anything that isn't cloud computing.... It's complete gibberish....When is the idiocy going to stop?"

CLOUD COMPUTING


•NIST (NationalInstitute of Standards andTechnology)

Definition of Cloud Computing: Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.

http://csrc.nist.gov/groups/SNS/cloud-computing/

CLOUD COMPUTING




CLOUD COMPUTING

1. On-demand self-service

A consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with each service’s provider.

2. Broad network access

Capabilities are available over the network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, laptops, and PDAs).Cha

ract

erist

ics


CLOUD COMPUTING

3. Resource pooling

The provider’s computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand. There is a sense of location independence in that the customer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or datacenter). Examples of resources include storage, processing, memory, network bandwidth, and virtual machines.

Cha

ract

erist

ics


CLOUD COMPUTINGC

hara

cter

istic

s 4. Rapid elasticity

Capabilities can be rapidly and elastically provisioned, in some cases automatically, to quickly scale out and rapidly released to quickly scale in. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be purchased in any quantity at any time.

5. Measured ServiceCloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported providing transparency for both the provider and consumer of the utilized service.

Cha

ract

erist

ics


Service Models:

•Software as a Service (SaaS)

•Platform as a Service (PaaS)

• Infrastructure as a Service (IaaS)

CLOUD COMPUTING

Think of them as decreasing levels of “blissful ignorance”


CLOUD COMPUTING

1. Software as a Service (SaaS)

The capability provided to the consumer is to use the provider’s applications running on a cloud infrastructure. The applications are accessible from various client devices through a thin client interface such as a web browser (e.g., web-based email).

The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings.

Serv

ice

Mod

els


CLOUD COMPUTING

2. Platform as a Service (PaaS)

The capability provided to the consumer is to deploy onto the cloud infrastructure consumer-created or acquired applications created using programming languages and tools supported by the provider.

The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, or storage, but has control over the deployed applications and possibly application hosting environment configurations.

Serv

ice

Mod

els


CLOUD COMPUTING

3. Infrastructure as a Service (IaaS)

The capability provided to the consumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications.

The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, deployed applications, and possibly limited control of select networking components (e.g., host firewalls).

Serv

ice

Mod

els


Deployment Models:

•Private Cloud

•Public Cloud

•Hybrid Cloud

•Community Cloud

CLOUD COMPUTING


CLOUD COMPUTINGD

eplo

ymen

t M

odel

s

1. Private Cloud

The cloud infrastructure is operated solely for an organization. It may be managed by the organization or a third party and may exist on premise or off premise.

2. Public Cloud

The cloud infrastructure is made available to the general public or a large industry group and is owned by an organization selling cloud services.


CLOUD COMPUTINGD

eplo

ymen

t M

odel

s

3. Hybrid Cloud

The cloud infrastructure is a composition of two or more clouds (private, community, or public) that remain unique entities but are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting for load-balancing between clouds).


CLOUD COMPUTINGD

eplo

ymen

t M

odel

s

4. Community Cloud

The cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be managed by the organizations or a third party and may exist on premise or off premise.

SUNY

Goodness?


•And with that our parking lot is full

CLOUD COMPUTING


Retrospective

Operations and SOPs


‘Cloud’ Computing

REVIEW


Retrospective

Cloud Computing

SOPs


•Service Evolution

•Virtualization

•Banner POC

•Grid Control

•Standardization

•PM Methodologies

•Definitions

•Community Cloud

•Predictable Future

•Healthy dose of science


•Good lord look at all the circles up there ...

CLOUD STRATEGY


•Weaving these 4 loosely coupled ideas together we can now assemble ITEC’s cloud strategy

•First - a tiny bit about our technology/business model

•How we see ourselves evolving into a cloud computing like model

•Some current initiatives under way

CLOUD STRATEGY

Keep in mind, I’m mostly concerned about database and application server technology bits of this strategy


• Campuses use the apps

• SICAS, OLIS, SLN manage the big apps

• ITEC manages the infrastructure (and some of the smaller apps)

CLOUD STRATEGY


•TODAY - we make use of virtualization at the storage, network and server level

•But we still end up (in essence) deploying dedicated environments for every customer for (mostly) every app

• i.e. we have not virtualized our databases or applications

•Time for a little ‘cloud’ comparative analysis

CLOUD STRATEGY


CLOUD STRATEGY

On Demand

Self Service

Broad Network Access

Resource Pooling

Rapid Elasticity

Measured Service

1/4 Check Check 1/2 Check 1/4 Check 1/2 Check

On Demand: Sort of, (we are very accommodating)

Self Service:No

Dual ISPs & BGP

SUNY VPN

Burstable consumption.

OS, Network, Storage YES,

Database, Apps NO,

(People Yes)

We can respond fairly quickly for the resources that we pool

but it’s mostly manual.

We have grid control collecting a lot of data points but we’re not looking at them WRT charge back.

Cha

ract

erist

ics

(tod

ay)


CLOUD STRATEGY

Software aaS Platform aaS Infrastructure aaS

Check Check No Check

We serve up @task, Confluence, Lyris, Elluminate to all ITEC members. Apps are 100% maintained by ITEC and our customers simply use them (with some very limited configuration capabilities)

This is really the core of what we do today.

Banner, ANGEL, ExLibris all operate in this model with other u-wide entities providing application management.

We provide VMs that OAS uses as part of their DR solution. We provide the hardware and basic networking and storage while OAS manages the rest.

Serv

ice

Mod

els

(tod

ay)


CLOUD STRATEGY

Private Cloud Public Cloud Hybrid Community

Tiny Bit NO NO 95% yes

We have a few things for our own internal use but these are very minimal

Too many concerns all of which are well discussed in the public.

By definition of being partially Public = No.

Amazingly YES.

Instinct told us that was true but looking at it with clinical perspective this is most certainly what we do today.

Dep

loym

ent

Mod

els

(tod

ay)


•Summary of our shortcomings:

•Agility and self-service

•Need more virtualization, standardization and automation

•Observation that you need more structure in order to be more flexible

CLOUD STRATEGY


•Strategy?

•Continue to develop our SOPs and then automate

•Virtualize our databases and app servers

•Choose reasonable pilot projects

•Draw on expertise where ever it lies

CLOUD STRATEGY

ITEC as a Cloud – Current and Future Projects & The ITEC Cloud sessions by Mike Notarius


So in a concrete sense - what are we doing about it ?

• We have a project to revisit our operations and develop well documented SOPs

• We’re developing a project to start exploiting EMs provisioning pack

• We’re re-embracing RAC and application server clustering (perhaps WebLogic Suite)

CLOUD STRATEGY


CLOUD STRATEGY

• Re-assessed our current and near future business and application needs

• at least 2x databases to manage

• More databases are approaching near 24x7 (LMS IS 24x7)

• CPU requirements are increasing

• Business need to be more agile and elastic

• and lastly - it’s now a matured (mainstream) technology

So, why the change of heart with respect to RAC?


•Specifically with respect to RAC:

•Three projects • LMS POC with RAC (and DBFS)

• SBIP - Cloud based Business Intelligence • Project Z

CLOUD STRATEGY


CLOUD STRATEGY

• Our LMS TEAM is already investigating Blackboard and Moodle as options to ANGEL

• At the end of these projects we are adding in RAC technology on the back end plus DBFS - a database hosted file system

• Performance, scalability, absolute point-in-time recovery

• Executing load/stress testing utilizing OATS - Oracle Application Testing Suite (possibly engaging Blackboard to help us)

#1 - LMS POC / RAC


CLOUD STRATEGY

• Hosting SICAS BI program in an internal BI Cloud

• Dramatic cost savings and efficiencies for all entities involved

• At the expense of flexibility

• Single (Enterprise Model) OBIEE install

• Multiple campus ODS in a single DB utilizingOracle Virtual Private Database

#2 - SBIP BI Cloud


CLOUD STRATEGY

GRID / Cluster

GRID / Cluster

Rac DB Infrastructure

BI Server Cluster(OBIEE 11g)

Web Server Cluster(WebLogic 11g)

Local BI Admin Server

IdentityInfrastructure

Campus ERP

Campus ERP

Campus ERP

Shared ODS (VPD)

Database


CLOUD STRATEGY

• A less conceptual version for those of you who like details


CLOUD STRATEGY

• And yes thats ‘Zed’ not ‘Zee’

• Developed by Germans

• Purloined by Canadians

• (Made by Americans)

• An evolutionary REVOLUTIONARY approach to hosted database management

#3 - Project Z


CLOUD STRATEGY

Z as in Zoolander the world greatest male supermodel?The man who brought us ....

MagnumFerarriBlue Steel Le Tigra


We’re talking Hansel.

In your face, off the wall, smashin’ the mold, Hansel !!!


CLOUD STRATEGY

• Self Service cloning of production databases to a cloud environment at ITEC

• Campuses will be able to initiate (via automation) clones of PROD databases into this environment

• Using RMAN and Grid Control we will clone the DB into the ITEC cloud

• Utilizes a cluster of x86 servers running Oracle RAC on Linux to deliver the database to you (and only you)

#3 - Self Service Cloning


CLOUD STRATEGY

• Benefits

• Removes the need for dedicated development database servers (i.e. saves money)

• Gives us time to become proficient in the technology in a lower risk environment (i.e. increased quality of life)

• Costs

• Investment in automation, infrastructure, training (internal to ITEC)



CLOUD STRATEGY

• Probably will not be as on-demand as one might imagine and will be an overnight process (network considerations)

• Cloning will be to a named database so that we can pre-configure integration with other application and web servers

• We will probably limit a campus to X non-prod databases



CLOUD STRATEGY

• Each of these 3 projects enable us to move forward in a manner that is low(er) risk than if we just said “we’re doing cloud tomorrow” and let er rip.

Moving forward


CLOUD STRATEGY

• Education - to better understand what we were being compared to and diffuse the hype

• Analysis - understand what our business needs are and where we stand today and where we should be tomorrow

• Insight - in several respects (but not all) we’re already doing computing in a ‘cloud’ style

• Found a way to move forward while managing risk

Summary

One more thing ....


CLOUD STRATEGY

• Cloud computing is an evolving body of work

• For as much as many may downplay vendor/media hype the fact is that numerous high profile organizations are working diligently to define cloud computing standards:

• NIST, IEEE, OGF (open grid forum), iTL (Information Technology Laboratory), ...

• As an example visit http://www.nist.gov/itl/cloud/cloudworkshopii.cfm


http://www.nist.gov/itl/cloud/cloudworkshopii.cfm




CLOUD STRATEGY

Standards Acceleration to Jumpstart Adoption of Cloud Computing (SAJACC)

The goal of the SAJACC initiative is to drive the formation of high-quality cloud computing standards by providing worked examples showing how key use cases can be supported on cloud systems that implement a set of documented and public cloud system specifications.

The SAJACC initiative will develop and maintain a set of cloud system use cases through an open and ongoing process engaging industry, other Government agencies, and academia. Simultaneously, the SAJACC initiative will collect and generate cloud system specifications through a similarly open and ongoing process.

The SAJACC initiative will develop tests that show the extent to which specific use cases can be supported by cloud systems that implement documented and public cloud system specifications, and will publish test results on the SAJACC web portal (this web site).

The SAJACC web portal will provide pointers to known cloud system implementations, use case documents, upcoming events, and will also provide a convenient means to provide feedback to the SAJACC team. These resources will serve to both accelerate the development of high- quality cloud computing standards and reduce technical uncertainty during the interim adoption period before many cloud computing standards are formalized.

http://www.nist.gov/itl/cloud/index.cfm




• Thank you

• I’m always looking for feedback

• [email protected]

Questions or Thoughts?




Documents

Virt Cloud.4