I have found slf4j is generally no harder to use than any other framework and it does have significant advantages when trying to integrate with other environments (because it is just a facade). If find that the slf4j strategy of picking a jar file to match the environment is simpler for me than the commons logging magic approach and more flexible than the log4j "we are the world" approach. This is probably an artifact of the scale of service I build and on a larger, more containerized level the world might not be so simple.
For ZK on the server side, I doubt the choice much matters. For the client side, I would think that the flexibility offered by slf4j would be good. On Fri, Dec 4, 2009 at 12:56 PM, Patrick Hunt <ph...@apache.org> wrote: > I pinged hadoop general@ a few months ago and got lukewarm reception > (generally "having multiple logger backends sux"). But no one really seemed > too interested one way or the other: http://bit.ly/71M6tK > > I'm not sure if anyone is thinking about this wrt hadoop as a whole... > > For example, what's the plan when integrating Avro into Hadoop M/R & HDFS? > > Patrick > > > Yonik Seeley wrote: > >> Any thoughts about migrating to SLF4J for java logging? >> Solr now uses it, as does Avro I believe, and other parts of Hadoop. >> >> -Yonik >> http://www.lucidimagination.com >> > -- Ted Dunning, CTO DeepDyve