4. McObject Company
Telecom & Networking ↑
Aerospace & Defense
Consumer Electronics
Financial ↑
Industrial & Process Control
Energy & Smart Grid
Mobile Database
Telematics
Web services
(persistent memory caching)
Reference Applications
5. McObject President & CEO
Steve T. Graves :
Co-founder
“In-memory database
systems. Interview with
Steve Graves, McObject.”
– Roberto V. Zicari ( ODBMS )
6. Gartner
Massimo Pezzini :
The Next Generation
Architecture:
In-Memory Computing
Who's Who in In-Memory
DBMSs
Published: 10. 09. 2012
+ McObject eXtremeDB
+ ALTIBASE HDB
7. DBMS2.com
Curt Monash :
Many kinds of memory-
centric data management
+ SAP HANA
+ IBM solidDB
+ McObject eXtremeDB
+ HP H-Store & VoltDB
+ Oracle TimesTen
8. (
STAC®
SECURITIES TECHNOLOGY ANALYSIS CENTER )
9. STAC Benchmark Council Members
McObject Joins
STAC Benchmark
Council – May 2012
DB Vendor :
1. KX Systems
2. McObject, LLC
11. STAC Workload - Market Data
Examples of products
Domain Caption Workload summary Example metrics
to be tested*
- Latency from exchange message hitting
Taking inbound market data messages Ticker plant software, ticker
the wire to normalized update coming
Direct feed from plant appliances, full-service
STAC-M1 integration exchanges, normalizing and caching them,
through API
- Max throughput to 5 clients with 99th
trading platforms that offer a
and making them available via an API. market data API
percentile latency less than 1 ms
- Latency from the moment a message is
Taking normalized market data streams ready for distribution to the moment it
Market data platforms,
Market data through a publisher API and delivering exits the subscriber API.
STAC-M2 distribution them - Latency to “undisturbed consumers”
messaging software, messaging
appliances
to multiple consumers via a subscriber API. when other consumers are starved of
resources
- First-result latency
Querying a time-series dataset and
(time to get back the first result)
Time-series applying
- Last-result latency
STAC-M3 data
management
various analytics. Taking streaming data,
applying basic analytics to it, and persisting
(time to get back all results)
Tick databases
- Write latency
it to a time-series store.
- Storage efficiency
15. eXtremeDB FE Architecture
▪ Core In-Memory Database System (IMDS) Design
– As an in-memory database system (IMDS), eXtremeDB gives your application speed without rewrites
or expensive new hardware.
▪ Short Execution Path, Tiny Footprint (Approximately 150K)
– Small code size and minimal overhead (database system memory footprint is as small as 15% of
managed data volume) means less RAM is required.
▪ Optional On-Disk or Hybrid Storage
– McObject's eXtremeDB Fusion edition
▪ Columnar Layout for Time Series Data
– Traditional DBMSs bring rows of data into L1/L2 cache for processing. But financial data – such as
trades and quotes – is naturally columnar, and handled more efficiently by a column-based layout.
17. Powerful Run-Time Features
ACID Transactions High Availability and Clustering
Transaction Logging 64-Bit Support
Multi-Version Concurrency Control ( MVCC ) Open Replication
Cache Prioritization Event Notifications
Deterministic Rule-Based SQL Optimizer Pattern Search
Security Features Binary Schema Evolution
Remote Procedure Calls ( RPCs ) Database Striping / Mirroring
XML Import / Export Kernel Mode Deployment
GUI-Based Performance Monitoring & API
18. Unmatched Developer Flexibility
B-Tree, R-Tree, Patricia Trie, KD-Tree and
C/C++, SQL, JAVA, C# APIs
hash Indexes
Wide Range of Supported Data Types Designed To Prevent Database Corruption
Custom Collations Broad Platform Support
Source Code Available Proven Solution
Unmatched Developer Support
19. Managing Market Data
with eXtremeDB Financial Edition
- Flexible data layout
- Vector-based statistical
function library
( boolean, add, subract, multiply,
divide, compare, not, and, or, xor,
conversion, weighted sum, weighted
average, covariance, correlation,
conditional operations, difference,
concatenation, max, min, sum,
product, count, average, variance,
standard deviation, user-defined
functions and more… )
- Handles real-time and
historical data