Multi-Site Replication on a Single-Site Budget

Multi-Site Replication on a Single-Site BudgetKyle Berger

Kyle Berger Biography

Executive Director of Technology

Technology Administrator of the Year for Texas

Oversee all aspects of educational technology

Technology administrator in education for over 10 years

Featured in industry publications for proven ROI, technology advancements within the K-12 market

Photo Courtesy: The Alvarado Star

Alvarado Independent School District – Texas

K-12 – 3,200 students 7 site locations spread

over 96 square miles 2,000 desktops 900 laptops –every

student in grades 4-6 13 TB Compellent

SAN since 2007

Audience Response Requested

1-1000 1001-3000 3001-5000 5001-8000 8001-1000 1001+

17% 17% 17%17%17%17%

1. 1-1000

2. 1001-3000

3. 3001-5000

4. 5001-8000

5. 8001-10,000

6. 10,001 +

Q: Number of end users in your organization?

Critical Server Applications and Storage

Key Network Software– Novell– GroupWise– Skyward (student information system)– RSCCC ( HR and finance software)

Storage– 4,000 network accounts

IT Challenges

Funding– K- 12 funding in regards to

technology– Never guaranteed– Always changing

Reliability as well as accessibility

– 24/7 access to data– Protection and recovery is key– Data integrity

What We Were Looking For

Reliable Cost effective Disaster recovery,

co-location– True ROI for the

school district – Little overhead

Do have the same challenge?

Yes Some No Don't Want

25% 25%25%25%

Q: Do you have the same challenges?

1. Yes

2. To some degree

3. No

4. Don’t want to think about it

The Solution

A DR consortium between K-12 schools

Connect Compellent systems across TX

Multi-site replication with out added cost or overhead

Compellent’s replication software connects locations like a business enterprise

How it Works

Using existing Internet bandwidth, dial up connections in off-peak hours

Schools replicate changed data from their primary SAN to another school’s primary location

Chargeback feature allows schools to recoup cost for storage use

Min of 500 gigs set aside for DRC usage Static mappings with ACL to allow for secure

communication between SANs

The solution

Districts across Texas connected

ACROSS THE COUNTRY

Parameters of the DRC and Testing

DRC is user driven– User group board of directors

Quarterly health checks of Compellent device– Results given to members

Two annual enrollment periods into the DRC– Allow for data migration and

placement within the DRC

Data integrity tested twice a year between sites

Replication is encrypted between sites

Benefits of the Overall Solution

Leveraged current investment in SAN Added remote survivability at zero cost Lower data center footprint No added energy cost (true green co-locate) No need to learn a new system Support from your peers when you need it most Compliance of data security is a non issue

– Each district held to the same data standards for security

ROI

Easy to establish, not extended cost outside of the SAN and its software

Compellent features and flexibility, leverage the investment for a ROI that goes beyond the initial investment

No added…– Energy cost – Bandwidth– Hosting fee – Cooling cost

Lessons Learned

Time and coordination of initial connection

Limited site connectivity Initial sync, available

bandwidth and data push User adoption – beyond

technology departments Different SAN versions? Remote site down times? Bandwidth throttling and

expectations Replication software not

frequent purchase first SAN

The Future

5 school districts across Texas– End of 2009

Additional states– Jan 2010

Cross county – Multi-site DR for K-12 mission

critical applications

Disaster Recovery Consortium (DRC)

Audience Response Requested

25%

25%25%

25%

1 2 3 4

Q: What would you like to hear more about?

1. Technical information

2. Agreement between schools

3. ROI

4. None

50%

50%

Audience Response Requested

1. Yes

2. No

Q: Would you like to join this or a similar solution?

Wrap Up!

Q & AKyle Berger

Kberger@alvaradoisd.net

Pablo Coronado – EST Group

pcoronado@est-grp.com

Session Evaluation

Please rate this session by responding to the following questions.

Your feedback helps us to improve the quality of future sessions.

Unlike the previous polling questions, your responses will not be visible on the screen but will be captured and recorded for the presenter to view offline.

Should you have additional feedback, please utilize one of the hard copy evaluation forms at the back of the room to provide us with your written comments.

Q: This subject is ____ to performing my job?

Not

App

licab

le

Slig

htly U

sefu

l

Impo

rtant

Ver

y Im

porta

nt

Crit

ical

20%

10%

50%

7%

13%

1. Not Applicable

2. Slightly Useful

3. Important

4. Very Important

5. Critical

Session Evaluation: Audience Response Requested

Session Evaluation: Audience Response Requested

A Lot

Les

s Det

ail

Les

s Det

ail

The

Sam

e Am

ount

...

Mor

e Det

ail

A Lot

Mor

e Det

ail

7%

30%27%

17%20%

1. A Lot Less Detail

2. Less Detail

3. The Same Amount of Detail

4. More Detail

5. A Lot More Detail

Q: I would like ____ in the material presented.

Session Evaluation: Audience Response Requested

Poor

Bel

ow A

vera

ge

Ave

rage

Good

Exc

eptio

nal

13%

23%

27%

13%

23%

1. Poor

2. Below Average

3. Average

4. Good

5. Exceptional

Q: I would rate the presenter as ____.