[wildfly-dev] Configuration of statistics gathering

[Andrig Miller]

----- Original Message -----
> From: "Jason Greene" <jason.greene at redhat.com>
> To: "Jesper Pedersen" <jesper.pedersen at jboss.org>
> Cc: wildfly-dev at lists.jboss.org
> Sent: Tuesday, December 17, 2013 1:14:18 PM
> Subject: Re: [wildfly-dev] Configuration of statistics gathering
> 
> 
> On Dec 17, 2013, at 2:06 PM, Jason Greene <jason.greene at redhat.com>
> wrote:
> 
> > 
> > On Dec 17, 2013, at 2:09 AM, Jesper Pedersen
> > <jesper.pedersen at jboss.org> wrote:
> > 
> >> 
> >> I'm ok with the code changes. HOWEVER, the default value should be
> >> true.
> >> 
> >> WildFly is a server which aims at ease for developers, so all
> >> default
> >> settings should reflect that. So, using a default value of false
> >> in this
> >> case is wrong.
> > 
> > I disagree that statistics are a developer concern. IMO they are
> > more of a production debugging feature and you only need them if
> > you plan to look at them. Otherwise it’s really just a bottleneck.
> 
> Slight tangent. We should test out LongAdder for connection pool
> stats at some point. It should make a significant perf difference.
> 

Yes, I would like to bring this in for all the stats, so they have the least amount of impact when they are on.  I understand Infinispan is already using it, but we had to turn off their stats too to get the best second-level cache performance (made a big difference even with LongAdder).

Andy

> --
> Jason T. Greene
> WildFly Lead / JBoss EAP Platform Architect
> JBoss, a division of Red Hat
> 
> 
> _______________________________________________
> wildfly-dev mailing list
> wildfly-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/wildfly-dev
>