Proven Low-Cost Database for Your Business

Tech Notes

InterBase® SMP:
Proven, Low-Cost Database for Your
Business
Embarcadero Technologies

February 2009

Corporate Headquarters EMEA Headquarters Asia-Pacific Headquarters
100 California Street, 12th Floor York House L7. 313 La Trobe Street
San Francisco, California 94111 18 York Road Melbourne VIC 3000
Maidenhead, Berkshire Australia
SL6 1SF, United Kingdom

InterBase SMP: Proven, Low-Cost Database for Your Business

INTRODUCTION
While large enterprises often require complex database systems that support thousands of
concurrent users, terabytes of data, and a highly-trained support staff, the needs of most
businesses and applications are more modest. InterBase SMP is a proven, highly-reliable, low-
cost database that can easily support hundreds of concurrent users and gigabytes of data with
no support during normal operation.

The most important characteristics of a database for the smaller enterprise are high reliability
and low maintenance for a low total cost of ownership (TCO). This paper examines the features
that determine reliability and total cost of ownership and explains why InterBase has been the
best choice for thousands of businesses and applications for over twenty years.

When evaluating the reliability of a database, it is important to ask the following questions:

• Can I lose data if a workstation crashes?
• Can I lose data if the server crashes?
• Can I lose data if the hard drive in the database server fails?
• Can I lose data if the database server is stolen or destroyed?

CRASH RECOVERY
Because InterBase SMP is a database server, client workstation crashes have no effect on the
database. If you have a server crash, caused by loss of power or some other event, InterBase
recovers reliably and immediately. When you restart InterBase all active transactions are
automatically rolled back and the database is returned to a logically consistent state without any
action on your part.

InterBase not only recovers automatically on restart after a server crash, it also recovers instantly
as there is no change log that must be processed to rollback active transactions. Instead,
InterBase simply changes the status bits for each active transaction to rolled back and leaves
the row versions created by the rolled back transactions in the database. The versioning engine
automatically ignores rolled back row versions and the garbage collector removes them
automatically as tables are visited during normal database use.

INTERBASE SMP: HIGH-PERFORMANCE
DATABASE ON MULTI-CORE CPUS
WHEN PERFORMANCE MATTERS
InterBase SMP is one of the fastest native databases available, and utilizing its built-in SMP
(symmetric multiprocessing) it scales to multiple CPUs, including multi-core CPUs. Performance
enhancements and optimizations have been added in a number of areas, such as batch
updates, query optimizations, and OLTP improvement via journaling. InterBase implements a
multi-generational architecture providing unique versioning capabilities that result in high data

Embarcadero Technologies -1-


availability for transaction processing users and decision support users—simultaneously.
Support for multiple simultaneous instances of InterBase on the same machine provides
application isolation and load balancing.

DATA PROTECTION AND RECOVERY
InterBase provides a full suite of features to protect your data from disaster. The database can
be backed up while in use. Both full and incremental backups are supported to minimize
backup time. Backups automatically maintain a duplicate copy of the database on another
machine. You can even use the backup database, in read only mode, while backups from the
production server are updating it. A single command will switch the backup database to
read/write mode so you can put it into production in an emergency.

The InterBase incremental backup system makes managing large databases easier because you
do not have to do periodic full backups. Other database systems that support incremental
backups require you to perform a full backup, then periodic incremental backups. However, to
recover the database you must restore the full backup then restore each incremental backup.

This means you must do periodic full backups to limit the number of incremental backups that
must be restored and thus limit the recovery time. Since InterBase incremental backups update
a working copy of the database you never have to do a full backup after the first one and
recovery is instantaneous; just switch the backup database from read only to read/write mode
and you are back on line. There is no restore operation.

Enabling transaction journaling records all database changes in the InterBase ‘write ahead’ log.
The journal file can be placed on a separate physical disk drive for protection from disk failure.
Journal files can be automatically archived to a remote server and the database can be
recovered from the remote server even if the production server has been physically destroyed.
Journaling also allows point-in-time recovery to protect from accidental deletion of data or
other logical damage to the database.

MINIMIZING COST
There are many factors that contribute to the total cost of ownership (TCO) of a database.
When evaluating the TCO, important questions to ask include:

• How much maintenance does the database require?
• What are the hardware requirements?
• Does the database give me a choice of operating system platforms?
• How much will it cost to train my development staff to use the database?
• How easy is it to diagnose a performance problem?
• What are the licensing costs?
• What level of support is available from the database vendor?
• Does the database provide features to accelerate development time?

DATABASE MAINTENANCE
A database that requires routine maintenance adds to TCO in one of two ways. You may be
able to automate the maintenance tasks but automating routine maintenance adds to
development time and cost. If the database requires maintenance that cannot be automated,



you will need staff to perform the maintenance tasks and total deployment costs will increase
over time.

InterBase databases require no maintenance. InterBase has features that are designed
specifically for use at sites where no DBA or other support is available. Because InterBase is self-
tuning and self-maintaining there is nothing for a DBA to do during normal operation.

SYSTEM RESOURCE REQUIREMENTS
The more hardware your database requires the higher the cost. A full installation of InterBase
requires less than 35 MB on disk. If you do not install the documentation and examples
InterBase requires less than 19 MB of disk space. Furthermore, InterBase requires just 32
megabytes of memory on all platforms.

MULTI-PLATFORM SUPPORT
To minimize hardware costs your database needs to run on a platform that is already available
at your site or it must give you the option to deploy a low-cost platform such as Linux. InterBase
runs on Windows®, Mac OS®/X, Linux and Solaris™ to give you platform flexibility now and in
the future.

TRAINING TIME
The time required to train your development team is a significant component of both the total
cost of ownership and the time it takes to deliver your project. It is also a good measure of
product complexity.

InterBase is so easy to learn and administer that the most comprehensive training course for it,
designed to train complete database novices to the point that they can confidently administer
InterBase databases, lasts just five days. The jump-start course for experienced SQL database
developers lasts just two days and many experienced developers find that they require no
training at all. Instead, the InterBase manuals are all they need.

MONITORING PERFORMANCE
Whenever you deploy a multi-user database system, there is always a chance that users will
experience poor performance. Users may get access to the database with a third-party query
and reporting tool and concoct queries that place a huge load on the database.

The important question is, when the unexpected happens, how many staff-hours will it take to
diagnose and correct the problem? InterBase includes a sophisticated performance monitor
that offers both graphical and SQL interfaces. Not only can you see exactly what is happening in
the InterBase server at any time, you can also log selected parameters over time for later review.
This makes diagnosing performance problems a snap, so you can maximize database
performance and availability.

VENDOR SUPPORT
Many of today’s database applications are mission critical. In these environments support from
your database vendor should be available whenever you need it. InterBase is backed by



Embarcadero with a full suite of support options from ‘per incident’ support to ‘around the
world around the clock’ support contracts. In many cases there are support personnel near you
or your deployed systems to help resolve issues.

LICENSING COSTS
InterBase licensing costs are among the lowest of all commercial databases. This is especially
true for systems that use multi-core CPUs. InterBase SMP 2009 is licensed per system rather
than per core. This allows you to scale up performance with multi-core systems without scaling
up costs. When combined with very low to zero maintenance costs, high-reliability, and other
features described in this paper, InterBase provides the lowest TCO for most business
applications.

In addition, customers who license Embarcadero® All-Access™ receive a license of InterBase
SMP 2009 Server edition with their All-Access license. This InterBase license provides support
for 5, 10, 25, or unlimited user connections, depending on the selected All-Access level.

MINIMIZE DEVELOPMENT TIME
A major component of database TCO is the cost of developing, maintaining, and enhancing
your applications that leverage the database. If your database does not provide the features
that make development easy, your development team can waste a lot of time coding around
the limitations of the database. InterBase provides a rich feature set that that makes developing
a wide range of applications faster and easier than with a less full-featured database.

HIGH CONCURRENCY
InterBase provides high concurrency in a mixed read/write environment because readers get a
stable snapshot of the database without blocking writers. Inserts, updates and deletes never
have to wait on read transactions regardless of their isolation level. In fact, snapshot transaction
isolation is the default in InterBase. No matter how many SELECT statements you execute
during the life of a snapshot transaction you will always get a stable consistent snapshot of the
entire database as it was at the instant the transaction started.

To prevent a row that has been updated by an active transaction from being changed by
another transaction, InterBase uses row level locking. The “locks” are actually implemented
through the versioning engine, not by traditional two-phase locking, so you never have to worry
about lock escalation, conversion deadlocks, or a large number of locks causing a severe
resource load on the server. These problems simply do not exist in InterBase.

TRANSACTION SUPPORT
InterBase provides full ACID transaction support with both snapshot and read committed
isolation levels. InterBase also supports two phase commit allowing a single transaction to
include changes to two or more databases.

SQL SUPPORT
InterBase provides a rich SQL implementation including ANSI 92 support with many extensions.
A powerful cost-based optimizer provides optimum SQL execution speed. Using bitmap



technology, indexes are combined to optimize performance with complex selection criteria.
InterBase lets you extend SQL on any platform with user-defined functions. User-defined
functions are implemented as DLLs on Windows and as shared object libraries on Linux and
Solaris so you can create them using the language you prefer.

STORED PROCEDURES, TRIGGERS & VIEWS
InterBase provides both traditional executable stored procedures and selectable stored
procedures. Selectable stored procedures are used in SQL SELECT statements in place of
tables and, like a table, return a set of rows. You define the columns that comprise the returned
rows when you write the stored procedure, then use any combination of SQL statements and
computations to supply a value for each column. Using selectable stored procedures,
developers save hours by easily creating virtual tables that are impossible to create with views.

For example, assume you need to produce a report that shows total sales for each of the last
four quarters and for the last 12 months grouped by territory. With most databases you would
create a temporary table with columns for the sales territory, the four quarterly totals and the
twelve month total.

Next you would insert a row for each territory; run a series of UPDATE/SELECT statements to
compute the totals for each territory and insert them into the temporary table; and finally select
the data from the temporary table to produce the report.

Using an InterBase selectable stored procedure, you declare output parameters to hold the
sales territory, quarterly totals and twelve month total. The body of the stored procedure loops
through the sales territories and selects the totals directly into the output parameters. That is all
there is to it. To produce the report you just SELECT * FROM PROCEDURE_NAME.

InterBase’s trigger implementation also reduces the amount of code developers have to write.
InterBase provides both before and after triggers for insert, update and delete operations. The
old and new value of each column is available in both before and after triggers. Before insert
and before update triggers let you validate or change the value in any column before the row is
inserted or updated. You can also prevent the insert, update or delete from taking place by
raising an exception in a before trigger. InterBase triggers are easy to write because the trigger
fires for each row changed, not once for each statement, so you do not have to write looping
code to iterate through the affected rows. InterBase also allows multiple triggers for each event,
with complete control of trigger firing order, so you can modularize code for easy development,
testing and maintenance.

Like most databases, InterBase lets you define a view using any valid SQL SELECT statement.
Views are frequently used to flatten or denormalize the data model by joining multiple tables.
However, many views that use joins or aggregate functions cannot be updated. This can force
you to write additional code in your client applications to handle changes to the data. InterBase
solves this problem by letting you create before triggers on a view. When a user makes a
change the trigger fires and you can easily control which tables, columns and rows are changed.

EVENTS
When client applications must be notified of events in the database, many database
management systems require you to poll the database at regular intervals or write complex



add-on components. Events are built into InterBase. If your application needs to be notified
when a part’s quantity on hand drops below the reorder point all you need is an after update
trigger to check the quantity and raise an event. Your client application simply registers interest
in the event and is notified immediately when the event fires. Of course, you never have to
worry about spurious events fired by transactions that subsequently rollback. No matter where
you raise an InterBase event, the event is not actually sent to the client until the transaction
commits.

METADATA
InterBase provides a full range of data types for floating point, fixed point and integer numeric
data; date, time and timestamp data, Boolean data, fixed and variable length character data
and BLOBs for regular binary and character data. InterBase supports the null state for all data
types. User defined data types (called domains) save both development and maintenance time
by letting you define standard types for names, addresses, part numbers and other values that
may appear in multiple tables. Altering the domain changes the type everywhere it is used.
InterBase supports multiple character sets including Unicode. A character set can be specified
at the database, table or column level.

InterBase tables also provide computed columns so you never have to store values that can be
computed on the fly. Default value, primary key, foreign key, and check constraints make both
referential integrity and domain integrity enforcement automatic.

SECURITY
InterBase provides both server-wide and embedded database security models. With server-
wide security, any user with a valid user name and password can log onto the InterBase server
then connect to any database. If you use embedded user authentication, user names and
passwords are stored in the database making it impossible for someone to kidnap your
database and open it on another server. InterBase provides role based access control to
database objects using the standard SQL GRANT and REVOKE commands.

DEPENDENCY CHECKING
InterBase will not allow you to drop or alter any database object when another object depends
on it. For example, you cannot drop a table if that table is used in a view or stored procedure.
Likewise, you cannot drop a stored procedure that is called from a trigger or another stored
procedure, nor can you drop a view that is used by a stored procedure or trigger. With
InterBase you never have to worry about making a metadata change that will cause another
object to fail.

A LOOK AT ALTERNATIVES
There are other databases you can consider for your applications but none of them were
designed specifically for the smaller enterprise. Alternatives to InterBase can be divided into
commercial products, such as Oracle® 10g, Microsoft® SQL Server 2008 and MySQL 5.1, and
open source databases, for example, PostgreSQL 8.3 and Firebird 2.1. MySQL can only be
distributed free of charge as part of your application if the entire application is distributed



under the GPL or an OSI approved open source license. 1 Since very few business applications
are distributed under an open source license, MySQL is included with the other commercial
databases.

Commercial databases are developed and supported by a single organization that can give you
quick in-depth support in an emergency. This is because the support staff has direct and
immediate access to the R&D engineers who develop the product. If you encounter a bug or
other problem with a commercial database, you can get a patch or work-around without waiting
for the next scheduled product release.

Royalty-free, open source databases appear to have a significant cost advantage because there
are no licensing costs. However, these kinds of databases often have a significant disadvantage
in terms of support and hidden maintenance costs. These types of databases are often
developed by groups of volunteers and lack a single, viable, commercial vendor to provide
support. Instead, you must rely on support from the volunteer development community and/or
third-party support vendors that do not have the level of access to the development team which
you find with commercial database products.

Because none of these commercial or open source products were designed specifically to
support the smaller enterprise, they all have technical weaknesses when compared to InterBase.
The following sections examine some of their limitations.

ORACLE 10G
Oracle is one of the most powerful, complex, and expensive databases available today. If you
need to support thousands of users and terabytes of data, Oracle is an excellent choice.
However, applications in small to medium size enterprises do not need to scale to that level.
Oracle has hundreds of tuning and configuration parameters and is so complex that Oracle
certifies DBAs at four different skill levels. 2 Oracle XE can be deployed at no cost but has the
following limitations.

• One database per machine
• One instance per machine
• Uses only one CPU
• Uses a maximum of one gigabyte of memory
• Database size is limited to four gigabytes of user data

In reality, Oracle XE cannot be used for most applications in any but the smallest businesses
due to the first two restrictions. Consider what happens when you try to add a second
application to an existing application server and discover that another application using Oracle
XE is already installed on that machine? Since Oracle XE only allows one database per machine
and one Oracle XE instance per machine, you cannot install a second application that uses
Oracle XE.

1
MySQL is not free when deployed with your application unless your entire application is deployed under the GPL or
an OSI approved open source license. In all other cases you must purchase OEM licenses to deploy MySQL.
Reference: http://www.mysql.com/about/legal/licensing/oem/

2
http://education.oracle.com/pls/web_prod-plq-dad/db_pages.getpage?page_id=141



Your only choices are to install another application server or upgrade to the full Oracle product.
Suddenly Oracle XE has become a very expensive choice. In addition, Oracle XE has a disk
footprint of over 1.1 gigabytes. That is 30 times the size of a full InterBase installation and 62
times the size of InterBase without documentation and examples.

MICROSOFT SQL SERVER 2008
Microsoft SQL Server is another enterprise database with the features and complexity necessary
to handle very large numbers of users and very large databases. SQL Server uses a complex
locking model that can cause poor concurrency if developers do not understand how to avoid
its traps. “As the Microsoft SQL Server Database Engine acquires low-level locks, it also places
intent locks on the objects containing the lower-level objects.” 3 “The Database Engine does not
escalate row or key-range locks to page locks, but escalates them directly to table locks.

Similarly, page locks are always escalated to table locks.” 4 Having entire tables locked without
warning causes serious concurrency limitations in a multi-user environment.
To avoid the pitfalls of lock escalation and conversion deadlocks you will need one or more
certified DBAs on your development team. Certification as a Microsoft SQL Server developer
and DBA requires 18 days of classroom instruction plus three examinations.
Developing applications with Microsoft SQL Server typically requires substantially more code
compared to InterBase. One reason is Microsoft SQL Server’s trigger implementation. Microsoft
SQL Server does not have before triggers. If you need to calculate or modify values being
inserted or updated, you must create an instead of trigger and write the code to perform the
insert or update, in addition to your data validation or calculation code. Even that option may
not be available. “Instead of, update, and delete triggers are not allowed on tables that are
targets of cascaded referential integrity constraints.” 5 In addition, only one instead of trigger is
allowed for each action so you cannot modularize your code. Microsoft SQL Server triggers fire
once per statement, not once for each affected row. So, in addition to the code you would have
to write with InterBase, you must declare a cursor and write code to iterate over the affected
rows. Microsoft SQL Server 2008 Express Edition can be deployed at no cost but it has
limitations.
• Uses only one CPU
• Database size is limited to 4 gigabytes
• Buffer pool memory is limited to 1 gigabyte

Many of today’s applications must store more than four gigabytes of data. If you need to store
BLOB data such as text documents, scanned images, photographs, sound or video, you can fill
a four gigabyte database very quickly. If there is any possibility that any of your sites may
exceed these limits, you will have to monitor your installations carefully so you can upgrade to
the full version of Microsoft SQL Server before you run out of space in the database or
performance degrades. Since SQL Server 2005 Express Edition is just a limited version of
Microsoft SQL Server, developers require the same level of knowledge and skill to develop
applications using the Express Edition that they must have to work with the full product.

3
http://msdn2.microsoft.com/en-us/library/ms184286.aspx
4
http://msdn2.microsoft.com/en-us/library/ms184286.aspx
5
Books Online – DML Trigger planning guidelines



MYSQL 5.1
MySQL is not a typical database management system but rather a front-end for a family of nine
different storage engines. 6 All of the engines have different table formats, concurrency control
mechanisms and other features. This means that you cannot simply decide to use MySQL. You
must evaluate the storage engines against your application’s requirements and select the one
whose characteristics best meet your application’s needs. For example, the MyISAM engine was
designed for fast processing of SELECT statements but uses table locks for insert, update and
delete statements. This can cause poor performance in a multi-user environment where data is
being changed.

When selecting a storage engine you must also carefully consider its longevity. For example,
the ISAM engine has been deprecated and replaced by the MyISAM engine while the Berkeley
DB engine has been dropped. In addition, some of the storage engines are third-party
products. The InnoDB engine, arguably the most feature rich of all of the storage engines, is
now owned by Oracle Corporation. Will the InnoDB engine still be available five years from
now? And, if so, at what cost? Uncertainty about the future adds risk to the decision to use a
third-party storage engine.

MySQL’s implementation of triggers and stored procedures is limited. Triggers are not
activated by the action of foreign keys7. If you need to provide an audit trail and you have
foreign keys 7 with the ON DELETE CASCADE or ON UPDATE CASCADE option, you cannot
use a trigger to log changes because the child table’s trigger will not fire when your updates are
cascaded. In addition, triggers cannot modify any table that is already being used, either for
reading or writing, by the statement that invoked the trigger. 8 You must also be careful about
how you write and use stored procedures as all the restrictions that apply to triggers, but not to
stored procedures, do apply to stored procedures that are called by triggers.

Views have a number of surprising limitations. In some cases indexes on base tables are not
used. 9 You can drop or alter a table that is used in a view without warning. You will not learn
that the view is now invalid until you try to use it. 10 A database backup using the mysqldump
utility will fail if a view exists that was created by a user that does not have the SHOW VIEW
privilege. 11

MySQL has many other limitations that make it a relatively poor choice for small to medium size
enterprise applications.

• MySQL has very limited performance monitoring for maximizing database performance
and availability

6
MySQL manual page 561
7
http://dev.mysql.com/doc/refman/5.0/en/routine-restrictions.html
8
9
10
11



• MySQL does not support multi-instance installations. This makes migration from one
version to the next more difficult and prevents isolation of one application from other
applications. It also prevents using multiple instances for load balancing.
• MySQL does not have an integrated mechanism to notify clients of database events
• MySQL does not have incremental backup. Instead, you must start a new binary log then
copy the old log files to the backup location. 12
• MySQL does not support user-defined data types
• MySQL includes over 100 system variables plus numerous configuration parameters that
control its operation; it is not self-tuning.
• MySQL has a query cache, key cache, table cache, InnoDB buffer and InnoDB log buffer.
This architecture burdens the developer with allocating and managing memory among
these five caches to get the best performance.

FIREBIRD 2.1
Firebird is a popular open source database that was originally based on the InterBase source
base, but was forked and has since been developed in the open source community. Compared
to InterBase is has limited scalability because it does not provide built-in symmetric multi-
processing (SMP) support. To take advantage of multiple CPUs, you must use the “classic”
version, which spawns a separate process for each user. However, the classic architecture does
not scale well. Having a separate Firebird server process and database cache for each user
consumes resources rapidly as the number of users grows.

A more serious limitation is that Firebird provides limited data recovery in case of media failure
or any other event that damages the database. You can restore from your last backup but all
changes since the last backup will be lost. The only way to provide recovery with little or no data
is to purchase a third-party replication tool and implement replication to another server. This
will protect your data but it also increases development time and deployment cost and
degrades performance.

Firebird has no performance monitoring tools. Diagnosing performance problems at remote
sites is very difficult. In addition, Firebird does not support temporary tables. Tasks that can be
done easily with temporary tables require additional coding when using Firebird.

Firebird also has no strong encryption or security features for protecting data. InterBase SMP
2009 has several strong encryption features for protecting data in secure application
environments.

Like other royalty-free, open source databases, support for Firebird is from the volunteer
development community and a limited number of third-party organizations. However, third-
party support providers do not have access to the core engineers developing the product; this
limits both the level of support you can get as well as the speed with which you can get answers
to questions and issues.

12
MySQL manual page 359-360

Embarcadero Technologies - 10 -


POSTGRESQL 8.1
PostgreSQL is another open source database that uses versioning architecture. However,
PostgreSQL is neither self-tuning nor low maintenance. PostgreSQL has over 170 configuration
options that must be set to control performance. In addition, it does not automatically perform
garbage collection as data is accessed. Instead, you must perform a vacuum operation
frequently to remove old row versions from the database. 13

PostgreSQL logs information to a log file. “However, the log output tends to be voluminous
(especially at higher debug levels) and you won’t want to save it indefinitely. You need to
“rotate” the log files so that new log files are started and old ones removed after a reasonable
period of time.” 14
Like Firebird and other open source databases, support for PostgreSQL is only available from
third-party providers. Consider carefully whether you can get the level of support you need
before you commit to any open source database.

GET THE FEATURES YOU NEED
Minimizing the TCO of your database means getting the features you need to minimize
maintenance, support, hardware, operating system, training, development and troubleshooting
costs. The feature matrix below shows why InterBase is truly the ideal database for the smaller
enterprise.

Feature InterBase Oracle 10g SQLServer MySQL Firebird PostgreSQL
SMP 2008 5.1 15 2.1 8.1
2009
Zero ✔ ✔
maintenance
Requires ✔ ✔
certified DBA
Support ✔ ✔ ✔ ✔
contracts
available from
vendor
Transaction ✔ ✔ ✔ ✔ ✔ ✔
support
SQL support ✔ ✔ ✔ ✔ ✔ ✔
Multi-user and ✔ ✔
single user
versions
User-defined ✔ ✔ ✔ ✔ ✔
functions
✔ ✔ ✔ ✔ ✔ ✔
Cost-based

13
PostgreSQL manual 22.1.2
14
PostgreSQL manual 22.3
15
MySQL features assume you are using Oracle Corporation’s InnoDB storage engine



optimizer
Provides a ✔ ✔ ✔ ✔ ✔ ✔
consistent
snapshot of
data without
blocking
updates

Row-level ✔ ✔ ✔ 16 ✔ ✔ ✔
locking
No lock ✔ ✔ ✔ ✔
escalation
No conversion ✔ ✔ ✔
deadlocks
Snapshot ✔ ✔ ✔ ✔ ✔
transaction
isolation by
default

Two phase ✔ ✔ ✔ ✔ ✔ ✔
commit
SQL-based ✔ ✔ ✔
performance
monitoring

Graphical ✔ ✔ ✔
performance
monitoring

SMP support ✔ ✔
at no
additional cost
Multi-core ✔ ✔ ✔ ✔ ✔
support
Hyperthreadin ✔ ✔ ✔ ✔ ✔
g support
Executable ✔ ✔ ✔ ✔ ✔ ✔
stored
procedures
Selectable ✔ ✔
stored
procedures
Before triggers ✔ ✔ ✔ ✔ ✔
After triggers ✔ ✔ ✔ ✔ ✔ ✔
Old & new ✔ ✔ ✔ ✔
values
available in

16
Row locks can escalate to table locks without warning



before & after
triggers

Multiple ✔ ✔ ✔ ✔
triggers on
same event

Control of ✔ ✔ ✔
trigger firing
order

Triggers fire ✔ ✔ ✔ ✔ ✔
for cascaded
updates &
deletes

Views ✔ ✔ ✔ ✔ ✔ ✔
Before triggers ✔ ✔
on views
Temporary ✔ ✔ ✔ ✔ ✔
tables
Fire client ✔ ✔ ✔ ✔
events
Automatic key ✔ ✔ ✔ ✔ ✔ ✔
generation
Unicode ✔ ✔ ✔ ✔ ✔ ✔
support
Supports the ✔ ✔ ✔ ✔ ✔
null state for
all data types
Declarative ✔ ✔ ✔ ✔ ✔ ✔
referential
integrity
Check ✔ ✔ ✔ ✔ ✔ ✔
constraints
Default values ✔ ✔ ✔ ✔ ✔ ✔
Computed ✔ ✔ ✔ ✔ ✔ ✔
columns
User-defined ✔ ✔ ✔ ✔
data types
Online backup ✔ ✔ ✔ ✔ 17 ✔ ✔
Online ✔ ✔ ✔ ✔ ✔
incremental
backup
Backups ✔
maintain a

17
Requires InnoDB Hot Backup available at extra cost from Oracle Corporation



backup
database on
another server
Online ✔ ✔ ✔ ✔ ✔ ✔
metadata
changes
Journaling for ✔ ✔
point in-time
recovery

Archive journal ✔ ✔ ✔ ✔ 18 ✔
files to another
server
Role-based ✔ ✔ ✔ ✔ ✔
access control
Multiple ✔ ✔ ✔ ✔ 19
instances
running on the
same server
simultaneously
Runs on ✔ ✔ ✔ ✔ ✔
Windows,
Linux & Solaris

Low licensing ✔ ✔ ✔ ✔
cost
Memory 32MB Not 512MB Not 32MB Not
required Published Published Published
Disk footprint 20 MB >1.1 GB 350MB; 140 MB 15.5 MB 77.5 MB
Without 775MB
examples with
or manuals
manuals. and
35MB samples
with.

CONCLUSIONS
InterBase SMP 2009 is a proven, low-cost database system with many unique benefits for small
to medium enterprise (SME) applications. This paper has covered many of the benefits of
InterBase SMP for SME applications and explained why it is a superior choice relative to other
databases designed for large enterprise deployments. InterBase SMP gives you easy
installation, high reliability, built-in disaster recovery, zero database maintenance, and a host of
features that make application development and database deployment easier, faster, and lower
cost.

18
Replication can be used to maintain a backup on another server. MySQL Manual page 947
19
Requires compilation from source if instances are different versions


Embarcadero Technologies, Inc. is a leading provider of award-winning tools for application
developers and database professionals so they can design systems right, build them faster and
run them better, regardless of their platform or programming language. Ninety of the Fortune
100 and an active community of more than three million users worldwide rely on Embarcadero
products to increase productivity, reduce costs, simplify change management and compliance
and accelerate innovation. The company’s flagship tools include: Embarcadero® Change
Manager™, CodeGear™ RAD Studio, DBArtisan®, Delphi®, ER/Studio®, JBuilder® and Rapid
SQL®. Founded in 1993, Embarcadero is headquartered in San Francisco, with offices located
around the world. Embarcadero is online at www.embarcadero.com.

Proven Low-Cost Database for Your Business

More Related Content

What's hot (20)

Similar to Proven Low-Cost Database for Your Business (20)

More from Embarcadero Technologies (20)

Recently uploaded (20)

Proven Low-Cost Database for Your Business