|
|
Right Java tools for J2EE, Web Service, Grid and more |
| |Home | Products | Documents | Downloads | Purchase | Support | Company | Partners | Contact | |
The design goals
The explicit and implicit goals are different for these two logging tools.
SuperLog is designed as a complete logging system,
not just an API. It is for J2EE distributed computing
environment. All logging activities are synchronized among all system in the
network.
With the introduction of LimpidLog, which
is for "log for debugging" and AuditLog, which is for "log for
auditing", SuperLog is a full logging system. See related
paper Loging In The J2EE World: log for debugging and for audit.
JDK logging is designed as an API for single JVM only. Logging activities are
not synchronized among different running systems in the network.
The architectures
SuperLog completely separates log request (logging methods) and control (management methods). The request
for logging is invoked
from the application code while the logging behavior is managed in a separate central program. It sets universal log threshold and
registration to check the eligibility for every log request. The management efficiently
controls all instances of logging program in clustering environment and keeps
them synchronized.
JDK logging puts every piece in one memory space. At first glance, java.util.logging.LogManager is the global management. But the word
"global" here is global among all classes within one JVM. Each application class sets (through its java.util.logging.Logger) its
own threshold and eligibility check. There is no central control in
clustering. It is difficult to add a central control program. It is even more
difficult to prevent any one from changing local attributes and break the synchronization
of the whole system, especially when application components are made by
different peoples.
SuperLog supports both LimpidLog and AuditLog
SuperLog supports LimpidLog and AuditLog. SuperLog combines both LimpidLog and AuditLog under one roof.
The usage
SuperLog is very easy to use. All preparation and initialization are transparent and done before you log
automatically. You just call logging methods.
Because it is simple, you have less chance to introduce bugs in your logging logic.
JDK logging is not that simple. You can see it from its examples. JDK logging is much more flexible, but the extra flexibility may not
be needed in most cases.
Other features
SuperLog is a complete system. You can dynamically re-configure all its
attributes using a GUI tool. It has built-in facilities for alert, send email and others. Source code level tracing is built in
as well. This is the default reader of log message. The purpose of logging
is to read log message when needed.
Related articles