Navigation
Home
SourceForge Home
Articles on Java
Getting Started
Essence Documentation
Download
Contact
Get rewarded for helping out
Email A Question
Weblog
|
Welcome to the home of Essence Java Framework.
As a simple Java framework, Essence delivers simple and novel solutions to
your complex problems. It draws from such products as Nano, Spring, Hibernate,
JDO, ODMG, JMX and JMS.
Essence 0.70 is 234 KB in size and around 330 KB including all dependent
jars.
Why use Essence?
Essence contains a number of innovative features that simplify the
configuration and management of components and data within your application.
The purpose of Essence is to use top down design and manage the complexity of
the application's configuration. Many frameworks start with the simplest
objects and attempt to assemble these together to make more complex high level
components. Essence assumes you have a high level design for your system and
attempts to hide the details of how each component is implemented.
These include:
- High performance highly available clustering. Essence 0.68
has been tested with 2-way and 4-way mastering (were each update is
committed to both or all 4 servers) and persisted to disk in near real time.
On a Linux blade, Essence 0.68 has been tested with 2-way and 4-way mastering (were each update is
committed to both or all 4 servers) and persisted to disk in near real time. On
a Linux blade, the tests performed 9,700 to 120,000 updates per second depending on the transaction size.
The tests performed 60,000 updates per second and 6 million
lookups per second concurrently in 2- way durable mastering mode. In
4-way durable mastering mode, the tests performed 24,000 updates per second and
2.4 million lookups per second.
Performance test results
- Explicit
declaration of all dependencies. A central configuration which defines
all dependencies between components. This is enforced by relying solely on
Dependency
Injection. Unlike other products, components have no access to the
Container which created them so you will have no container-dependent code
nor can the component have dependencies which were not given to it by the
container.
- Data sharing through
Immutable
data value
objects and
Event
driven processing. Separation between logically stateless
business objects and stateful data value objects. This ensures the
current state of the application is visible and well defined. The data model
supports distributed data and event processing in a thread safe and durable
way. The data model supports replacement of one data storage type with
another by configuration. A JDBC table can be replaced with a tab
separated file for the purpose of development or units tests. This
allows for complex canned data and data driven testing.
- Binding of components to Managed
Logical Process.
Logical process act as if it were in a single threaded application.
Components are bound to a logical process and can ensure it is only called
by one thread. Each logical process has only one thread and components in
different logical process are isolated from each other.
- Distributed durable collections can be accessed as a
Map,
a
Collection or a
BlockingQueue. Collections can be in memory, on a remote server,
embedded in the configuration file, in a tab separated file, a binary file,
or in a JDBC database. An event listener can be added to any collection
locally or remotely to hear changes. Bootstrapping is just an entrySet() or
values() call. Persisted data can be updated synchronously,
semi-synchronously or asynchronously.
- Named parameters for constructors. Reflection based conventions
are used rather than annotations, XML, setters or getters.
- Multiple container unit tests support. Essence supports unit test
with multiple containers to test connection failure, failover conditions and
load balancing. In the JMS broker example, two client container connect to a
broker which is killed and a second broker started. This demonstrated
successful restarting of the broker. This means you can unit failover
and load balancing scenarios even with complex applications.
- JMX support for logical processes and durable collections
- Base types are primitive types, their wrappers,
String, Enum types,
BigDecimal,
BigInteger, Class,
Date.
List<String>,
Map<String, String>,
Set<String>. Also supported are Arrays of base types (but not
collections)
- A package path can be defined so that class names can be shortened to just
the class name.
Essence Version 0.70 released
The latest version is
Essence 0.70, and the online Javadoc is available at
Getting started Javadoc
. Essence 0.70 has around 7,000 lines of tests, whereas 0.68 had 2,500 lines.
What example programs exist?A sample JMS broker is provided:
Essence JMS Sample 0.70. The client and broker code produces a jar which is
39K, all required jars combined are 354K.
What is example of a simple data value object?
public class DestinationEntry extends DataValue {
public final String destination;
public final long messageId;
public DestinationEntry(String destination, long messageId) {
this.destination = destination;
this.messageId = messageId;
}
}
This is used in the JMS broker example. This is all the code you need to
write. The constructor can be generated using an IDE. The container handles the
rest. You don't have to extend DataValue class but it gives you a number of
convenience methods such as equals(), hashCode(), toString(), asMap() to turn
into a map of key/values, project() to convert into another data value and set()
to create another immutable data value object except with one or more value
changed.
What is example of a simple component?The following component has one
property
value, a reference to a Map and one to another component.
public class CompTestWithRef extends CompTest {
private final Map<String, TestEntity> map;
private final CompTest testRef;
public CompTestWithRef(long value, Map<String, TestEntity> map, CompTest testRef) {
super(value);
this.map = map;
this.testRef = testRef;
}
}
What collections features are supported?
| |
ReadMode |
| |
PersistMode |
| |
Object Types |
| |
NONE |
ALL |
MINIMAL |
| |
NONE |
ASYNC |
SEMI_SYNC |
SYNC |
READ_ONLY |
| |
Any |
Serializable |
Datable |
DataValue |
| In Memory |
 |
 |
|
| |
|
|
|
|
|
| |
|
|
|
|
| Tab sep file |
|
|
|
| |
|
|
|
|
|
| |
|
|
|
|
| Single File |
|
|
|
| |
|
|
|
|
|
| |
|
|
|
|
| Directory |
|
|
 |
| |
|
|
|
|
|
| |
|
|
|
|
| JDBC |
|
|
|
| |
|
|
|
|
|
| |
|
|
|
|
| Remote (net) |
|
|
|
| |
|
|
|
|
|
| |
|
|
|
|
Legend
|
|
Feature not planned |
|
|
Feature available |
|
|
Planned for release v1.0 |
Modes
| ReadMode.NONE |
The store is not loaded from a persisted storage |
| ReadMode.ALL |
Read the entire store into memory on start up. |
| ReadMode.MINIMAL |
Read data as required. |
| PersistMode.NONE |
Do not persist any data changes. |
| PersistMode.ASYNC |
Write data asynchronously to disk. Every 0.5 - 5 seconds depending
on the type
|
| PersistMode.SEMI_SYNC |
Write data close to synchronously to disk. Every 25 - 250 ms
depending on the type.
|
| PersistMode.SYNC |
Write data synchronously to disk. |
| PersistMode.READ_ONLY |
Disable writing to the store and throw an exception if a write is
attempted. |
| Object Type (Any) |
Any Object type |
| Object Type (Serializable) |
Java Serializable objects |
| Object Type (Datable) |
An Essence specific serialization based on DataInput and DataOutput
streams. |
| Object Type (Data Value) |
Any object which can be reduced to a Map of String keys and base
type values. |
Suggested Tools
YourKit Review: Visualization is a key attribute of a profiler and
YourKit is an excellent tool. I suggest profiling your application before
releasing in case there is a performance issue which can be easily resolved.
YourKit found a number of CPU and memory consumption issues I wasn't aware off
but definitely worth fixing.
About the Developer
 |
Peter Lawrey has been working with Java since 1999 and has worked
for a number of Investment Banks in the London.
He worked for
Sun Microsystems (Australia) for a number of years before moving to
the UK.
|
|
