Private Cloud reduces risk calculations by 50%

TIBCO (DataSynapse)
Grid Server
David Broeren
Business Architect
NAB

© 2008 TIBCO Software Inc. All Rights Reserved. Confidential and Proprietary.

Company Overview

 National Australia Bank is among Australia’s largest
financial institutions by market capitalisation and one of
the world's top 30 financial services companies.
 NAB operates across 10 countries serving 10 million
banking and wealth management customers.
 Wholesale Banking, a division of NAB, has more than
2,500 people across Australia, New Zealand, Asia, the
United Kingdom and the United States.
 Wholesale Banking provides customers with access to
international financial markets and a range of
specialised funding, investment, asset services and risk
management capabilities.
http://www.wholesale.nabgroup.com/AboutUs/Pages/Whoweare.aspx
2

Introduction: Setting the Context?

What is Cloud Computing? What is a Grid?
App
5
Software as a Service
Cluster Server or
Compute Container
Virtual Server Grid

App App
1 1
Deployment Grid Server or Server or
Container Container
Grid (Desktop)
Client1

Desktop or Client1
Client2
Server or Server or
Citrix Svr Container Container
Hardware Virtualisation
Grid (Server)
Container Container
App App Data
2 1 Grid
Virtualisation
App App App App
2 1 1 4
Server or
Server or Server or Container
Container Container

3

Challenges

 One example of why is it important to a Bank
Input Output Nomenclature
Curve
Rates Data Bootstrapping

Curve
Interpolation

Price
+ Perturbations
Rates Data

+ Bootstrap Curve Perturbed Curve
Rates Data

1 Month Bill

2 Month Bill

3 Month Bill
Perturbed
Curve

15 Year

25 Year
1 Day Bill
+

10 Year
Interpolation Interpolated

18 mth

24 mth

4 Year
5 Year
7 Year
Future

Future

Future
12mth

3 Year
Future
6 mth
Curve

Cash Future Swap
 For example: Value at Risk (a regulatory risk control)
 Minimum 2 years of daily history
 Over 5 Risk Factors
= ~ 500 (trading days) * 5 (risk factors) * entire trade population
= 2,500 valuations per trade each time to calculate
4

Benefits and Return On Investment

Calypso
 Commercial Off the Shelf Implementation
 The time to install the TIBCO DataSynapse GridServer components on the NAB
Test Environment took approximately two hours.
 The time taken to install and deploy the Calypso integration layer was
approximately two hours
 To the limits of the available hardware the scalability of the resulting grid-enabled
Calypso deployment is almost completely linear for on-grid processing.
 Results
 In pre-grid testing, the Calypso Scheduling engine reached a capability limit. With
grid processing this instability was removed.
 Calypso task preparation and collection contains a bottleneck which skews end to
end performance as measured from Calypso. For example, the results are written
to a file by a single process.
 Near linear scalability across the grid
PLUS:
• Expanded full portfolio recalculation, previously limited due to capacity
constraints
 Now: Live and expanding
5

Benefits and Return On Investment

BGS Provision Model:
 Excel based Monte Carlo simulation:
 The BGS Provision Model is delivered as a set of C++ dlls
 The entry point method accepts as input a set of data, including Trades,
Schedules, Scenarios, Curves, Volatilities, etc.
 The calculation is a 3-factor Monte Carlo simulation.
 After each calculation the dlls return a vector of values
 Spreadsheet
• performs the shifting operations,
• marshals the input data and calls the entry point method of the BGS Provision
Model dlls 7 times; once for each of the interest rate shifts.
• The dll invocation is effected directly from Visual Basic code

 Desire to increase to a minimum of 512 shifts, each containing 50,000 paths.
 Linearly, this would increase the elapsed time for the calculations in excess of 300
times,
 Pre-grid, this would take a minimum run time on a single PC of over 50 hours!

6

Architecture


The Master Worker Pattern

http://www4.java.no/presentations/javazone/2006/slides/4623.pdf

8

Implementation

 TIBCO GridServer
additional benefits:
 Deployment of code
managed centrally
 “Profile” matching:
 Hardware, runtimes,
libraries, etc
 Ease of deployment
 Audit trail of usage (i.e.
chargeback capability)

9

A shift in paradigm

 Non-grid

Monitor CPU
Capacity must be sized for maximum demand
• If reach maximum demand risk catastrophic failover

 Grid

Not possible to overwhelm workers
• Only take work when available
• Can reliably work at 100% utilisation
• If worker fails, grid manager re-assigns task
• Focus;
 sustained put that exceeds take

10

Best Practices

11

Candidates

 Grid Server
“Model” implementations
• Look for the “for” loops
• Be aware, it will expose bottlenecks and design flaws – this is a good thing!

Master Worker pattern implementations
• Anything that can be broken into tasks and carried out in parallel

Breaking the back of Batch

 Fabric Server
Server deployments
• Extreme availability websites
• “Peaky” applications

12

Lessons Learned

 Difficult to touch and feel

 Baby steps

13

Wrap Up and Summary

 Consider a multi-dimensioned business case:
 Business Sponsors
• More comprehensive decision analytics; Revenue/Risk drivers

 Technology Leadership
• Cost reduction; Less servers -> the no maintenance utopia
• Resilience
• Quality of Service
• Service charge models, pay for capability not server

 Corporate Social Responsibility

14

Private Cloud reduces risk calculations by 50%

Recomendados

Recomendados

Mais conteúdo relacionado

Mais procurados

Mais procurados (9)

Destaque

Destaque (9)

Semelhante a Private Cloud reduces risk calculations by 50%

Semelhante a Private Cloud reduces risk calculations by 50% (20)

Último

Último (20)

Private Cloud reduces risk calculations by 50%