Jerry Held

1
(No Transcript)
2
Oracle TimesTen In-Memory Database
3
Agenda

• TimesTen Architecture
• TimesTen Options
• Q A

4
Oracle TimesTen In-Memory Database
A memory-optimized application-tier database for
blazing-fast response times and real-time data
caching
Presentation Tier
Application-tier data management for
performance critical systems
Application Tier
Enterprise Database Tier
5
Lightning Fast Response
28 millionths of a second
25 millionths of a second
microseconds
9 millionths of a second
70 select 20 update 10 insert
TimesTen In-Memory Database 6.0, 4-CPU, 3 GHz x86
Xeon, 32-bit RHLinux
6
Scalable on Commodity Platforms
Transactions per second
70 select 20 update 10 insert
TimesTen In-Memory Database 6.0, 4-CPU, 3 GHz x86
Xeon, 32-bit RHLinux
7
Combined Performance
8
Why Is TimesTen So Fast?

• TimesTen needs fewer CPU instructions to
  accomplish the same work as a disk RDBMS
• Physical memory addresses are used inside the
  engine
• No lookups of logical addresses to physical
  addresses
• No buffer cache management overhead

9
Optimized In-Memory Database

• Entire database is in memory
• Optimized memory layout and algorithms
• Data store shared by multiple threads and
  processes
• Embedded or client-server
• Standard APIs (ODBC, JDBC, SQL-92)
• Easy installation and configuration

Network
...
In-MemoryDatabase(s)
Client-Server
Transaction Logs
Checkpoint files
Application-Tier Platform
10
Oracle TimesTen IMDB Supported Platforms
Operating System Processor Type
AIX 5L 5.2, 5.3 (32/64 bit) IBM Power
HP-UX 11i (32/64 bit) HP PA-RISC
HP-UX 11i v2 (64 bit) Intel Itanium-2
MontaVista Linux CGE 3.0, 3.1 (32/64 bit) Intel (IA32, EM64T)
Red Hat Enterprise Linux 3.0, 4.0 (32/64 bit) Intel IA-32, EM64T, Itanium-2, AMD Opteron
Solaris 8, 9, 10 (32/64 bit) UltraSparc Opteron
SuSE Enterprise Server 9 (32/64 bit) Intel IA-32 EM64T, AMD Opteron
Tru64 Unix 5.1B (64 bit) AlphaChip EV68
Windows Server 2000, 2003, XP Intel IA-32
11
Logging and Data Persistence

• Buffered and synchronous transaction logging
  options
• Application programs can control logging option
  at the transaction level
• Transaction logs and checkpoint files persist on
  disk
• Dual checkpoint files for recovery

12
Data Publishing - Transaction Log API (XLA)

• Transaction Log API (XLA)
• Track real-time data changes
• Monitor transaction updates
• Propagate changes to external applications
• Implement real-time event notification and
  processing

13
ReplicationTimesTen to TimesTen
14
Replication TimesTen to TimesTen

• N-way, peer-to-peer
• Active/standby or active/active, multi-level
• Table or data-store level
• Async sync transmit
• Auto recovery, master catch-up
• Conflict detection/resolution
• Online upgrade

15
Configuration FlexibilityReplication TimesTen
to TimesTen

• Flexible Configurations
• Replicate tables or data-store

16
Cache Connect to Oracle
17
Integration with Oracle DataCache Connect to
Oracle
Instantly Responsive
The Cache Connect to Oracle option combines the
blazing-fast response times of Oracle TimesTen
with the scalability and capacity of Oracle
Database 10g RAC
Tightly Integrated
18
Cache Connect to Oracle

• Cache full or subset of tables from Oracle DB
• Subset of rows
• Subset of columns
• Full tables
• Cache can be read-only or updatable
• Bi-directional updates
• From TimesTen to Oracle DB
• From Oracle DB to TimesTen
• Support multiple Cache installations per Oracle
  database
• Replication optional

19
Cache Group Definition
20
Read-only Cache Groups

• Read-only Cache Groups
• Updates disallowed
• Updates can be sent to Oracle (via the
  Pass-through feature)
• Updates in Oracle automatically refreshed to the
  TimesTen Cache Group
• Resilience to failures

21
Example Valuable Customers
Caching premier customer info CREATE CACHE GROUP
PremierCustomers FROM CUSTOMER ( NAME CHAR(100)
NOT NULL, ADDR CHAR(100) ) WHERE CUSTOMERS.ORDER
gt 500
PremierCustomers
CUSTOMER
22
Updatable Cache-Groups

• Updates in Cache Group
• Synchronous or asynchronous write-through to
  Oracle
• Synchronous write-through transactions
  synchronously committed in Oracle
• Asynchronous write-through transactions
  batch-committed in Oracle
• Resilience to failures

23
Complementary Database Strengths
Database Characteristic Oracle Database 10g Oracle TimesTen In-Memory Database
Target Applications Mission-critical Mission-critical
Data Model Relational - SQL Relational - SQL
Optimization Disk-centric Memory-centric
Typical deployment Database tier Application tier
Architecture Client/server Direct data access
Response Times Millisecs - to seconds Micro- to milliseconds
Data Capacity Tens of terabytes Tens of gigabytes
Scalability Cluster (RAC) SMP SMP
24
Oracle Times Ten Customers
Inside Enterprise Products
Inside Networking VAS Solutions
TimesTens real-time caching and data management
software provides Essential performance for
Fixed Mobile Convergence (FMC)
25
Use Cases
26
Fast, On-Target Customer ServiceLeading Call
Center Solutions Vendor

• Oracle TimesTen Usage
• Reference data lookups (agent skillsets)
• Real-time data matching (calls to agents)
• Dynamic state management (case agent status)
• Real-time reporting (call center monitoring)
• Performance Metrics
• 40,000 (avg) - 100,000 (peak) calls per hour
• 5,000 10,000 agents
• Configuration
• 2-CPU Server (plus hot-standby)
• Windows O/S with .NET applications
• 100 Megabyte Oracle TimesTen
• Value of Orcle TimesTen
• Instant, accurate matches
• High throughput per server

Enterprise Contact Server
Agents
Skills Based Routing
Real-Time Historical Reporting
Agent Task Manager
Scheduled Callback
Active
Standby
IP-PBX
ACD
IM
27
Dynamic PersonalizationHosted CRM Provider

• Oracle TimesTen Usage
• Caching personalization preferences
• Performance Metrics
• 10 million requests per day
• 250 peak requests per second
• Configuration
• 4-CPU servers
• Linux with Java applications
• 2 gigabyte Oracle TimesTen
• Oracle RDBMS
• Value of Oracle TimesTen
• Offloading backend systems
• Low latency response
• Customized interactions
• High-availability replication

Worldwide Subscribers
Region 2 Application Servers
Region 1 Application Servers
Load Balancer
Load Balancer
Active
Standby
Standby
Active
Master Database
Master Database
Oracle
28
Fixed-Income Trading SystemGlobal Broker-Dealer
Different Geographies or Security Classes

• Oracle TimesTen Usage
• Order capture / processing
• Event publishing
• Performance Metrics
• 1,000 fixed-income trades/hr
• 20 trader alerts/sec
• Configuration
• 4-CPU server plus standby
• Solaris with C applications
• 2 gigabyte Oracle TimesTen
• Value of Oracle TimesTen
• Fast order execution
• Trader alerting

Orders Inquiries Notifications
Trader Alerts
Trader Alerts
Active
Active
Standby
Standby
XLA Application
XLA Application
Closed Orders
Msg Bus
AggregatedReporting
Global Order Repository
29
Example Real-Time Billing
Tier 1 European Mobile Operator

• TimesTen Usage
• Event capture
• Reference data lookups
• Balance management
• Performance Metrics
• Sub 10ms response
• 70/30 read/update workload
• Configuration
• 8-CPU servers
• 10 gigabyte TimesTen
• Oracle RDBMS
• Value of TimesTen
• Real-time offers for high-volume customers
• Commercial, standards-based

Real-Time Offers for Premium Subscribers
BSC
MSC
SS7
INPrepaid
BatchBilling
30
Real-Time Authentication Spending Controls
Wireless Service Provider in Turkey
VAS Authentication

• Oracle TimesTen Usage
• Event capture (billing events, VAS usage
  statistics)
• Reference data lookups (black/white list, budget)
• Performance Metrics
• Real-time read only access for 20 million
  subscribers
• Microseconds response time
• Configuration
• 2 Servers (hot standby)
• Sun/Solaris platforms
• 4GB Gigabytes in TimesTen (caching from Oracle)
• Oracle RDBMS
• Value of Oracle TimesTen
• Scale to support high margin value-added services
• Meet government mandate for spending controls
• Eliminate network latency and timeouts

RAN
SMSC
SDP
Authentication
AuthenticationLayer
Oracle
VAS Budget
Subscribers
Oracle
Oracle
Oracle
Permissions
SubscriberType Data
ReplicationDatabase
31
Bulletproof VoIP Services that Work
Leading IP Centrex Vendor
IP Centrex

• Oracle TimesTen IMDB Usage
• Reference data lookups (subscriber profile mgmt.)
• Real-time data matching (routing info location)
• Dynamic state management (active call status)
• Performance Metrics
• Application Server (AS) 50,000 users
• Network Server (NS) 1M users 95/5 read/update
  workload
• Configuration
• 2, 4, 8-CPU Servers (plus hot-standby)
• Sun/Solaris platforms
• 2 Gigabytes TimesTen (DB of record)
• Value of Oracle TimesTen IMDB
• Carrier-grade reliability
• Highly scalable throughput across wide geography
• Enables advanced IP Centrex productivity
  applications

Remote Location
Headquarters
ApplicationServer (AS)
NetworkServer (NS)
Softswitch
ProxyServers
MediaGateway
MediaGateway
MediationGateway
MediationGateway
32
Taking the Latency Out of Push-To-Talk
Push-to-Talk Vendor

• Oracle TimesTen IMDB Usage
• List management (buddy list presence/availabilit
  y)
• Dynamic state management (network active
  sessions)
• Event capture (FCAPS)
• Performance Metrics
• 60/40 read/update workload
• Latency is the key differentiator
• Configuration
• 2-CPU Servers (plus hot-standby)
• RedHat Linux platforms
• 300-500 Megabytes/server TimesTen (DB of record)
• Value of Oracle TimesTen IMDB
• Real-time user session authentication
• Advanced rich user profiles presence
• Carrier-grade reliability

Push-to-Talk
2G/2.5G
UMTS
BSC
xGSN
UTRAN
MSC
BTS
CSCF
HLR/HSS
PTT System

• Call Setup
• Authentication
• Presence
• Buddy List
• Billing

33
Enabling a Real-Time World
Over 1,500 companies worldwide use Oracle TimesTen
34

Q/A
35
(No Transcript)