[Hawkular-dev] Logging ...

[Lucas Ponce]

And what about the logging packages ?

> import org.jboss.logging.*;

vs

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

I personally vote for org.slf4j.

I prefer a mix on option 1) and 2), a common numeration for codes, but each component can still using a specific acronym.

That it can help better to follow traces/logs.

Also, this point can be discussed later, but is there any convention for debug, like this one: 

        if (LOG.isDebugEnabled()) {
            LOG.debug("Message received: [{}]", msg);
        }

I mean, I like to use DEBUG for verbosity, but to reduce performance penalty, we can discuss a common approach.

Above one is just a suggestion.

Thanks,
Lucas


----- Original Message -----
> From: "Heiko W.Rupp" <hrupp at redhat.com>
> To: hawkular-dev at lists.jboss.org
> Sent: Wednesday, January 28, 2015 12:48:26 PM
> Subject: [Hawkular-dev] Logging ...
> 
> Hey,
> 
> We should / need to start using logging codes just like Wildfly does.
> 
> 11:19:27,714 INFO  [org.wildfly.extension.io] (ServerService Thread Pool --
> 31) WFLYIO001: Worker 'default' has auto-configured to 8 core threads with
> 64 task threads based on your 4 available processors
> 11:19:27,716 INFO  [org.jboss.as.clustering.infinispan] (ServerService Thread
> Pool -- 32) JBAS010280: Activating Infinispan subsystem.
> 
> We have two options here
> 
> 1)
> We  use "HAWK" as our prefix and the following  code ranges initially
> 
> 0-999  Hawkular (integration)
> 1000-1999 Hawkular-Metrics
> 2000-2999 Hawkular-Alerts (notifiers should probably be broken out in
> separate blocks)
> 3000-3999 Hawkular-Inventory
> 4000-4999 Hawkular-Bus
> 
> 2)
> We use a prefix per component (e.g. HAWK, HKMET, HKAL, ...) and each
> component has its own code range starting
> at 0
> 
> Personally I prefer option 1
> 
> There is already documentation on this available
> 
> 	• https://developer.jboss.org/wiki/HowToLogInJBossProjects
> 	• https://developer.jboss.org/wiki/InternationalizedLoggingAndExceptions
> 	• https://developer.jboss.org/wiki/JBossLoggingTooling
> 
> While the documentation and examples look a bit complicated, I think we can
> "reduce" them for
> our purpose for Warnings and Errors and do not need them on debug/trace
> level.
> For info level I guess one should personal judgement (and then if it is not
> important enough for
> its own code, it is probably not really something at info, but rather at
> debug level).
> 
> In practice we would probably end up with something like:
> 
> import org.jboss.logging.*;
> public interface RestApiLogger {
> 
>     RestApiLogger LOGGER =
>     Logger.getMessageLogger(RestApiLogger.class,"org.hawkular.inventory.rest");
> 
>     @LogMessage(level = Logger.Level.WARN)
>     @Message(id = 3000, value = "Something bad has happened")
>     void warn(@Cause Throwable t);
> 
>   < plus much more refined messages here >
> }
> 
> and then in code
> 
> } catch (Exception e) {
>     RestApiLogger.LOGGER.warn(e);
>     return Response.serverError().entity(e).build();
> }
> 
> 
> 
> --
> Reg. Adresse: Red Hat GmbH, Technopark II, Haus C,
> Werner-von-Siemens-Ring 14, D-85630 Grasbrunn
> Handelsregister: Amtsgericht München HRB 153243
> Geschäftsführer: Charles Cachera, Michael Cunningham, Paul Hickey, Charlie
> Peters
> 
> 
> _______________________________________________
> hawkular-dev mailing list
> hawkular-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hawkular-dev
>