Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
HBase, mail # dev - deprecating (old) metrics in favor of metrics2 framework


+
Ted Yu 2012-07-07, 14:23
+
Jonathan Hsieh 2012-07-09, 16:28
+
Ted Yu 2012-07-09, 16:35
+
Ted Yu 2012-07-09, 19:28
+
Elliott Clark 2012-07-09, 21:34
+
lars hofhansl 2012-07-09, 23:47
+
Ted Yu 2012-07-09, 23:52
+
lars hofhansl 2012-07-09, 23:56
+
Otis Gospodnetic 2012-07-10, 04:05
+
Ted Yu 2012-07-10, 08:45
+
Gary Helmling 2012-07-10, 18:26
+
Ted Yu 2012-07-10, 19:02
+
Gary Helmling 2012-07-10, 20:29
+
Ted Yu 2012-07-10, 20:35
+
Gary Helmling 2012-07-10, 20:55
+
Ted Yu 2012-07-10, 21:03
+
Alex Baranau 2012-07-10, 22:24
+
Elliott Clark 2012-07-10, 22:33
+
Ted Yu 2012-07-10, 23:14
+
Andrew Purtell 2012-07-10, 23:54
+
Elliott Clark 2012-07-10, 23:58
+
Alex Baranau 2012-07-11, 16:29
+
Elliott Clark 2012-07-11, 17:50
+
Andrew Purtell 2012-07-11, 18:57
Copy link to this message
-
Re: deprecating (old) metrics in favor of metrics2 framework
Elliott Clark 2012-07-12, 02:07
Yep.  This is pretty darn close to DI.  I would be willing to change the
patch that I just put up to use Guice if that's what people wanted.  I know
some of our classes could use some DI loving to help with testability.

On Wed, Jul 11, 2012 at 11:57 AM, Andrew Purtell <[EMAIL PROTECTED]>wrote:

> On Wed, Jul 11, 2012 at 10:50 AM, Elliott Clark <[EMAIL PROTECTED]>
> wrote:
> [...]
> >
> > The reflection option seems like it would be a big perf hit.  Reflection
> > basically means that nothing ever gets inlined so all function calls into
> > metrics2 code would be very expensive.  Since it seems like we are adding
> > more and more instrumentation this perf impact would only grow.
> >  In addition as more hadoop versions come out all of our reflection code
> > would get much more complicated and brittle.
> >
> > The conditionally loaded jar would be nice in that the JIT would only see
> > one version of the factory classes on the classpath and everything could
> be
> > optimized just like any other jvm run jit'd code. In addition there are
> > other places that we use reflection to do things conditionally and a
> > conditionally loaded jar would be nice.
>
> Coprocessors are a solution close to conditionally loaded JARs: there
> is not reflection, very frequently executed code would be JITed, the
> only thing missing is direct inlining.
>
> More and more (useful) metrics are going into hot paths, we want all
> three I think.
>
> We keep inching toward DI.
>
> Best regards,
>
>    - Andy
>
> Problems worthy of attack prove their worth by hitting back. - Piet
> Hein (via Tom White)
>
+
Stack 2012-07-12, 08:23
+
Elliott Clark 2012-07-13, 00:11
+
Elliott Clark 2012-07-10, 21:01