Larry Ellison Chief Executive Officer, Oracle
Oracle Confidential – Internal/Restricted/Highly Restricted Copyright © 2014 Oracle and/or its affiliates. All rights reserved. |
Oracle Database In-Memory Powering the Real-Time Enterprise
2 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Database In-Memory
Option
Powering the Real-Time Enterprise
Available in July
3 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Database In-Memory Option: Goals
100X Faster Queries: Real-Time Analytics • Instantaneous Queries on OLTP Database or Data Warehouse
2x Faster OLTP • Insert rows 3x to 4x faster
Transparent: No application changes
• Minutes to Implement
4 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle 12c Stores Data in Both Formats Simultaneously
Optimizing Transaction and Query Performance Row Format Databases versus Column Format Databases
Row
Transactions run faster on row format – Example: insert or query a sales order – Fast processing of few rows, many columns
Column
Analytics run faster on column format – Example: report on sales totals by region – Fast accessing of few columns, many rows
ORDER
SALES
SALES
REGIO
N
5 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
BOTH row and column formats for same data/table
Simultaneously active and transactionally consistent
100X Faster Analytics in-memory column format
2X faster OLTP: row format
Innovation: Dual Format In-Memory Database
Column Format
Memory
Row Format
Memory
Analytics OLTP Sales Sales
Sales
6 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory Column Format
Pure In-Memory
Pure Columnar
2X to 20X compression: Faster Scans
No data change logging: Faster OLTP
Enabled at table or partition level
Available on all hardware platforms
Sales
Sales
7 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Scans Billions of Rows per Second per CPU Core
SIMD Compare all values in 1 cycle
Vector Compare all values in 1 cycle
Load multiple Regionvalues V
ecto
r Re
gist
er
In-Memory Column Store Sales
Example: Find all sales in region of CA
“CA”
>100X Faster
• Each CPU core scans local in-memory columns
Scans use super fast SIMD vector instructions
Billions of rows/sec scan rate per CPU core
CPU
REGION
8 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
10x Faster Joining and Combining Data
Sales Stores
Type=outlet
Example: Find total sales in outlet stores
Storeid in
15,38,64 STOREID
AMOUNT
Converts join processing into fast column scans
Joins tables 10x faster
Sum
STOREID
TYPE
9 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Generate Reports Instantly
In-Memory Report Outline
Example: Report sales of footwear in outlet stores
Products
Stores
Sales
Footwear
Sales
Dynamically creates in-memory report outline
Then report outline filled-in during fast fact scan
Reports 20x faster without predefined cubes
Out
lets
Footwear
10 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
OLTP is Slowed by Analytic Indexes
Table 1 to 3 OLTP
Indexes
10 to 20 Analytics Indexes
Most OLTP Indexes (e.g. ERP) are only used for analytic queries
Inserting one row into a table requires updating 10-20 analytic indexes: Slow!
Indexes only speed up anticipated queries & reports
11 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Column Store Replaces Analytic Indexes
Table
1 to 3 OLTP
Indexes 100x Faster analytics
Works on any columns Better for ad-hoc analytics Less tuning required
• 2x Faster OLTP and Batch • Column store not logged • Row Insert cost is lower
In-Memory Column Store
12 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Scale-Out In-Memory Database to Any Size
In Memory Column Store
Scale-Out across servers to grow memory and CPUs
In-Memory queries parallelized across servers to access local column data
Direct-to-wire InfiniBand protocol speeds messaging
In Memory Column Store
In Memory Column Store
In Memory Column Store
13 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
In-Memory Speed + Capacity of Low Cost Disk
Size not limited by memory
Data transparently moves between tiers
Each tier has specialized algorithms & compression
Speed of DRAM I/Os of Flash Cost of Disk
DISK
PCI FLASH
DRAM
Cold Data
Hottest Data
Active Data
14 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Scale-Up for Maximum In-Memory Performance
Scale-Up on large SMPs
SMP scaling removes overhead of distributing queries across servers
Memory interconnect far faster than any network
M6-32 Big Memory Machine
32 TB DRAM 32 Socket, 384 Cores
3 Terabyte/sec Bandwidth
15 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory: Extreme Availability
Pure In-Memory format does not change Oracle’s storage format, logging, backup, recovery, etc.
All Oracle’s mature availability technologies work transparently
Protection from all failures Node, site, corruption,
human error, etc.
RAC
ASM
RMAN
Data Guard & GoldenGate
16 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory: Unique Fault Tolerance
Similar to storage mirroring
Duplicate in-memory columns on another node • Enabled per table/partition • Application transparent
Downtime eliminated by using duplicate after failure
17 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory: Implement in Minutes
1. Configure Memory Capacity inmemory_size = XXX GB
2. Configure tables or partitions to be in memory alter table | partition … inmemory;
3. Drop analytic indexes to speed up OLTP
18 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory Requires Zero Application Changes
Full Functionality - No restrictions on SQL Easy to Implement - No migration of data Fully Compatible - All existing applications run unchanged Fully Multitenant - Oracle In-Memory is Cloud Ready
Uniquely Achieves All In-Memory Benefits With No Application Changes
19 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Applications
From Batch to Real-Time
With In-Memory and Engineered Systems
Oracle Database In-Memory
Oracle Applications
20 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Real-Time Enterprise
Data Driven – Rapidly make decisions based on real-time data
Agile – Respond quickly to change
Efficient – Continually improve processes and profitability
Real-Time Enterprise
AGILE
EFFICIENT
DATA-DRIVEN AGILE
21 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory Cost Management
Adjust product mix, pricing, and marketing investments to maximize profits in real-time
Recalculate cost of every component in inventory, work-in-process, in-transit shipments, and finished good
Fast analysis of cost differences across manufacturing locations for make or buy decisions
• 1.9 Billion Cost Rows • 13.8 Million Items • 14 Level BOM
From 58 Hours to 13.5 Minutes
257X Faster
22 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
PeopleSoft In-Memory Financial Analyzer
Iterative financial position analysis in real-time
Make earlier decisions for the financial period
Speed account reconciliation Shorten financial period close
• 290M Ledger Lines • 250 Business Units • 7 Step Analysis, Pivot, Drill
From 4.3 Hours to 11.5 Sec
1300X Faster
23 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle In-Memory Transportation Management
Dispatchers perform real-time monitoring and rerouting when en-route exceptions occur
Benefits of instant rerouting: Reduction in empty miles and in Driver
turnaround time Improved on-time delivery Improved Dispatcher efficiency and
Driver retention • 145M Status Records • 60M Shipment Data Records • 16K Drivers
From 16 Min to Sub-second
1030X Faster
24 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
JD Edwards Sales Order Analysis
Operational Analysis of sales orders in real-time
Find immediate answers to unanticipated customer sales questions
Eliminate batch jobs, data exports, third-party systems
1000’s of use cases across all functional areas
• 104 million sales order lines
From 22.5 Min to Sub-second
1700X Faster
25 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
JD Edwards Customer Receivables Management
Real-Time receivables summarization Balances by customer, line of business,
and currency Eliminate multiple queries, batch jobs,
data exports Similar use cases for projects,
suppliers, assets, inventory etc. • 10 million invoice lines
From 244 Min to 4 Secs
3500X Faster
26 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Value Chain Planning Demantra Promotion Planning
Marketing Managers want detailed data when analyzing promotion plans
Benefits of real-time promotion analytics: Analyze promotion profit and revenue
on real-time basis Optimize promotion spend Better assess timing and cost impacts
• 1.3 Billion Rows • Aggregate 36M rows • 2 week major sell-through report
From 1120 to 11 Seconds
102X Faster
27 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Value Chain Planning Demantra Consumption Driven Planning
Daily store-level forecasting and replenishment requires processing high volume POS data multiple times a day
Benefits of real-time planning: Consumption and shipment based
forecasting in a single scalable system Improved forecast accuracy and
customer service levels at lower cost • 400 Million Rows • 1.4 Million SKU Locations
From 12.7 Hours to 56 Minutes
13.5X Faster
28 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Value Chain Planning Supply Chain Planning and Analytics
Supply Chain analysts spend hours processing granular supply and demand data
Benefits of real-time planning: Planners share business metrics with analysts and VP’s of Supply Chain real-time
Quickly solve supply disruptions and unexpected demand fluctuations
• 360K Items • 1.2M Demands, 1M exceptions • 5.7M KPI records
From 230 to 3 minutes
76X Faster
29 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Siebel Marketing – List Import
Real-time marketing for large numbers of prospects
Rapid processing of marketing data for campaign launch Accelerate import of large
numbers of prospects Reduce data-cleansing time
From 1.9 hours to 49 sec • Import and Cleanse 1 Million Records
140X Faster
30 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Fusion Cloud Financials Subledger Period Close Exceptions
Lists all accounting events and journal entries that fail period close validation in real-time
Report is run many times at end of quarter and is a bottleneck to close
From 10 Minutes to 3 Sec • 19 Million Ledger Lines
210X Faster
31 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Oracle Fusion Cloud Financials Sales Pipeline Analytic Report
• Report potential income in real-time • Aggregate revenue from
opportunities grouped within each sales stage for a specific time period
From 52 Minutes to 24 Sec
• 1.6 million opportunities • 5.1 million revenue lines
129X Faster
32 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Become a Real-Time Enterprise Using Oracle Database In-Memory
Real-Time Enterprise
Data-Driven • Get immediate answers to any
question with real-time analytics Agile
• Eliminate latency with analytics directly on OLTP data
Efficient • Non-disruptively accelerate all
applications
AGILE
EFFICIENT
DATA-DRIVEN AGILE
33 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.
Summary: Oracle Database In-Memory
Extreme Performance: Analytics & OLTP Extreme Scale-Out & Scale-Up Extreme Availability Extreme Simplicity
Powering the Real-Time Enterprise
All In-Memory Benefits With No Application Changes
34 Copyright © 2014, Oracle and/or its affiliates. All rights reserved.