Presentation on the role of the (relational) database in modern enterprise application architecture and on the major themes and development in the evolution of the Oracle Database through the years, up to and including 12c. This presentation was created for and delivered to students in Computer Science at Fontys Hogeschool in Eindhoven on April 25th 2014.
3. 3
Agenda
• Wie of wat is AMIS?
• Hoe verging het Chiel na zijn studie in Eindhoven?
• Hoe zit het met die Database?
• Vragen & Opmerkingen
4. AMIS
• Leuke organisatie!
• Oracle & Java specialisten (pakweg 90 man & vrouw)
• Gevestigd in Nieuwegein
• Opgericht in 1991 door studenten van de Universiteit Twente
– A M I S
• Opdrachten in
– Heerlen, Joure, Almelo, Nijmegen, Deventer
– Utrecht, Den Haag, Amsterdam, Rotterdam
– New Caledonia, Mongolië, Canada,
Denemarken, België, Koeweit,
Faroër eilanden, Duitsland, Verenigde Staten,
…
• Stage-opdrachten
– Mail: eva.van.der.kleij@amis.nl
5. The Presenter:
Lucas Jellema
• Lives in The Netherlands
(close to Amsterdam)
• Started doing Oracle in 1994 with
Oracle Consulting (Oracle Designer, Forms, Database)
• Joined AMIS in 2002 – now working as CTO,
Consultant (Architect, Technical Lead, Programmer)
and Trainer
• Oracle ACE (2005) & ACE Director (2006)
• Author of „Oracle SOA Suite 11g Handbook‟
(Oracle Press, 2010)
• Presenter at Oracle OpenWorld, JavaOne and
many Oracle and Java User Group Conferences
• Frequent blogger at http://technology.amis.nl
• Active with SQL & PL/SQL, Java EE & ADF,
SOA, BPM & more Fusion Middleware
6. 6
AMIS
• Projecten rond (enterprise) web applicaties en portalen, B2B integratie,
“mobilisering”, 24/7 beschikbaarheid, User Experience, business process
management
• Klanten waaronder:
– Politie Nederland, Randstad, Raad voor de Rechtspraak, NS, ProRail, Eneco
– Pensioenfondsen, financiële instellingen, logistieke bedrijven, software producenten
• Technische kreten
– Database, Middleware
SOA, BI, IoT,
Big Data, UX, HTML 5,
(No)SQL, Events,
Java, grid, IdM, XML,
Cloud, BPM, VM,
Provisioning, Scalability
Engineered Systems
8. THE EVOLUTION OF THE DATABASE –
ROLE OF THE DATABASE IN
APPLICATION ARCHITECTURE
8
9. Overview
• Role of the Database in Modern Architecture
• More than a container for data
• Evolution of the Oracle Database
• Recent Trends around the Database
• Q & A
16. More than a container for data
• “We could also do that in the database”
• in the database? Huh?
RDBMS
≈>>
17. Brief history of Oracle
Database
• 1970: Paper “A Relational Model of Data for Large Shared Data Banks” by
Ted Codd (IBM)
• 1977: Software Development Laboratories was founded
• Late „70s – Project “Oracle” – executed by SDL for the CIA
• 1979: SDL became Relational Software, Inc. and introduced Oracle V2
(built in PDP-11)
– Basic SQL functionality (query, join) but no transactions
• 1982: RSI became Oracle Corporation
• 1983: Oracle 3 - rewritten in C, ported to UNIX platforms and extended
with COMMIT and ROLLBACK
• 1984: Oracle 4 – read consistency
• 1985-1990: Oracle 5, 5.1 and 6: client/server, distributed queries, row
level locking & hot back ups
• 1993: Oracle7… more than just a database
20. PL/SQL Virtual Machine
• Ever since Oracle7, the Oracle RDBMS contains a PL/SQL Virtual
Machine
• That means: write [PL/SQL] once, run anywhere (as long as the RDBMS
is running)
• This portability of PL/SQL across platforms is pretty extraordinary!
PL/SQL
21. 993 2009
10g7.0 8.0 8i 9i
internet
xml
JVM inside
ANSI/standards
interMedia
Spatial
OLAP/BI/Analytics
Data Warehouse/ETL
The Grid
11g
PL/SQL:
Triggers,
Packages
SOA
SOX
Constraints
20011997 2004
Data Mining
Semantic Query
OO
RegExp
‘13
XE
NoSQL
Evolution of the Oracle Database
12c
CBO
RAC
Flashback & Total Recall
Data Guard
R
24/7 EBR
Cloud
Multi-platform porting
Materialized Views & Query Rewrite
SQL2Hadoop
Pattern Matching
Temporal DB
Multitenant
In Memory DB
JSON
Data Masking
22. Backup and recovery
• Back up is not relevant….
• .. unless you have an established and test recovery strategy!
• Backup should be done „hot and on line‟
• Recovery should be
– Unnecessary
– Online
– Focused
– Quick
• Oracle Backup & Recovery facilities
– Data Guard – (remote)
synchronized Stand By
Database
– Flashback Recovery
– RMAN
23. 24/7 – High Availability
• Unplanned Downtime
– Real Application Clusters (RAC)
– Data Guard Standby Database for fail over
– Hot Patching
– Back-up Air Conditioning & Power Generator
• Planned Downtime
– Online Redefinition
– Edition Based Redefinition for zero down time database application upgrades
– NoSQL “read ahead & write behind” cache layer
24. Compression
• When the system‟s performance is I/O bound
– Reading/writing data from/to storage cannot keep at the same rate as the CPUs
• there is spare capacity in the CPUs to alleviate the I/O burden
– Through a reduction in the volume of data to be read and written
– by zipping before write and unzipping after read
• The same information gets across in fewer bytes and therefore with less
I/O
• It even reduces storage
requirements
– Has a green bonus!
25. Flashback and Total Recall
• Oracle Database has long had two capabilities for „remembering‟ the old
situation regarding records
– Uncommitted transactions whose changes are local
– Read consistency, even for long running queries
• In 9i, 10g and 11g this feature
has been extended to remember
„the old situation‟ for far longer
– Undo Data => Flashback Archive
• This allows for
– Historical queries
– Trend analysis over time
– Point-in-time recovery
• At database, schema, table
or even record level
• 12c adds Valid Time Modeling (aka Temporal
Validity): database keeps track of begin and end
date of the valid period of records
26. Valid time aware flashback
queries
• Select all employees who were employed at a certain moment in time
• Perform all queries as if we traveled back in time
• Example: Run an end-of-year report in the end of January
SELECT *
FROM EMP AS OF TIMESTAMP TO_TIMESTAMP('01-JUN-2012 12.00.01 PM')
EXECUTE DBMS_FLASHBACK_ARCHIVE.enable_at_time
( 'ASOF'
, TO_TIMESTAMP('29-MAR-20414 13:14:15')
);
27. Role of the Database in
Modern Architecture
• Single Point of …
• What goes where …
29. Java Class processing an
HTTP request: The servlet
• The JVM can handle HTTP requests via the Servlet Container
• HTTP requests are routed to a custom Java Class that writes the (usually
HTML response)
• The Servlet infrastructure handles the actual HTTP response to the
invoker
JVM
Servlet
WebLogic Server
30. PL/SQL Package processing an HTTP
request: The embedded PL/SQL gateway
• The Database (the PL/SQL VM) can handle HTTP requests via the
Embedded PL/SQL Gateway
• HTTP requests are routed to a custom PL/SQL Package that writes the
(usually HTML response)
• The EPG infrastructure handles the actual HTTP response to the invoker
Custom
PL/SQL
package
E
P
G
htp
31. The Talking Database
Details on the Employee.
Employee name is Smith, his job
is Analyst. He works in
department 20…
EMP
33. Reaching out from the
database
Database
dbms_epg
(mod_plsql)
http
XMLDB
http
ftp
webdav
utl_http
http
Advanced
Queuing
AQ
jms
Queue
REST
Web
Service
File, O/S
JDBC,
SQL Net,
DB Link
36. Do not do it…
More often than required
• If it has been produced before…
– Reuse before re-produce!
• If it has been shipped before…
– Reuse instead of re-ship
• … provided it is still fresh
Web Browser
RDBMS
Java EE Application Server
37. Do not do it…
More often than required
• Save on network trips, context
switches and tiers to cross
• Save on „reproducing‟
same results Web Browser
RDBMS
Java EE Application Server
-JS data
(memory)
-Cookies
- HTML 5 db
Edge
Cache
Cache
Cluster Fail-Over
(Session State)
Result Store
Write Behind Client Result
Cache
Result Cache
Materialized
View
38. Fast Data –
Real Time Event processing
• The league of real time events
– Continuous stream of a multitude of tiny events with hardly any payload, to
analyze & aggregate
– Sent from physical sensors (temperature, pressure, RFID, security gates), process
sensors, Twitter, manufacturing equipment, database triggers, web servers, ESBs,
stock trade tickers, sport statistics, RSS, network switches, …
39. No SQL
• Replacing Relational Database?
– Front End & Middle Tier Cache for read only data provisioning
– Non-Transactional processing of large volumes of data
– Not Only SQL
– Working closely with „back-end‟ relational database
– Working with Hadoop for off-line, parallel processing
40. Do not do it…
In a suboptimal place
• Do not perform a task in a resource that is not ideally suited for that task
– If it directly contributes to overall performance
41. Do not do it…
In a suboptimal place
• Leverage database for what it‟s good at
– Data Integrity – Primary Key /Unique Key /Foreign Key
– Aggregation
– Sorting
– Data Rule enforcement
– Bulk DML and Copy data
– Analytical Functions, Model clause, Rollup
• Specialized engines for
– HTML rendering and Session Management
– Imaging and Document Processing
– Match and Search
– Speech Recognition
– Cryptography
– 3D
– ….
42. Recent Trends around the
Database
• Self managing => role of DBA
• Cloud
• Appliances & Engineered Systems (the ExaData machine)
– From I/O bound back to CPU bound
• NoSQL (data gathering) and R (data analysis) ...
• 12c
– Multitenant, In Memory Database, Temporal Validity, Data Masking, JSON
44. Cloud
• Automated (self-service) provisioning of database resources
• Multi-tenant
• Metering and per-usage billing
• 24/7 uptime – hot patching, fail-over
• Fine grained recovery, upgrade, authorizations
• Scalable
• „The Oracle Public Cloud‟ IaaS
PaaS
SaaS
45. Appliances & Engineered
Systems
• Pre-configured, installed, plug‟n‟play
• One stop solution in case of issues
• Software/hardware mutually optimized (*
– Infini-band, Storage Cells, Flash-memory (between RAM & Disk)
46. Data Masking
• Gartner reports that: data masking should be mandatory for enterprises
using copies of sensitive production data for application development,
analytics or training.
• They also believe the market is
expanding into production and
unstructured data protection.
47. Data Redaction
• At runtime, you can optionally have the query results modified to
reset/scramble/randomize sensitive data
– Through „data redaction‟ policies associated with tables and view and applied at
query time
• Because the data is masked in real-time, Data Redaction is well suited to
environments in which data is constantly changing.
• You can create the Data Redaction policies in one central location and
easily manage them from there.
SQL
engine SQL
POLICY
POLICY
RESULTS
48. Oracle 12c (released summer 2013)
C is for…
Complete
Complementary
Cloud
Consolidation
Container
Crowd
Control
Core
Central
Cool Carefree
Classified
50. Managing dozens of databases
means…
– Installing
– Configuring
– Securing
– Monitoring
– Patching
– Upgrading
– Backing up
• many database instances on potentially a large number of machines
• Using dedicated resources for each individual database instance
– 20 processes
– Memory for SGA
– Disk space for generic objects such as most of the SYS schema
62. 62
Fast Cloning a PDB
PDBs can be cloned from
remote CDBs
PDBs can be cloned from
within the same CDB
63. Manage Many as One with
Multitenant
Backup databases as one; recover at pluggable database level
One Backup
Point-in-time recovery
At pluggable database level
68. 68
Oracle Technology Network
http://otn.oracle.com
• Gratis download van vrijwel alle Oracle producten
• Gebruiken voor studie, onderzoek, prototype development, …
• Geen beperking in tijd of functionaliteit
• Betalen als je een
systeem in productie
(gaat) nemen
• Handig: Pre-Built
Virtual Machines
– PHP
– BigData
– SOA & BPM
– Java EE
– BI
– Solaris
70. 70
Summary
• Database is core of enterprise IT
– Persistent when the plug is pulled
• Consolidation is important theme
– More efficient usage of hardware resources and
of human resources
– Agile scaling – quick, simple, cheap
– Cloud as ultimate target
– Also: “logical consolidation” – centralize data
• Database is bottleneck – a single point of failure
• Logical consistency demands require transactions to use locks on data –
scalability at database level is reduced as a result
– CAP theorem – consistency, availability and partition tolerance can not all be
achieved
• Evolution of databases continues rapidly – the relevance of the Database
Platform is quite strong today
• IT is a smart career choice – and AMIS is a fun company
Cache – spreekuit: kasjeKastjesBrowser: Client (browser, cookie or Java Script memory; HTML 5 offers persistent, cross session local db like storage)App Server : Edge (WebServer)JVM (and cluster)Cross cluster shared cachedb or memory gridDatabase (not requery at least)
Copy data in PL/SQL (rather than bring from DB to Middletier, copy, send back again)
Add New in title
By managing many as one, we maximize OpEx reduction. Here we see a single backup strategy for the entire container database… with the granularity of a PITR capability at the individual PDB level.