It sounds like the actual test failure does not occur in HEAD, so sounds like whatever it is is already fixed or rendered moot.
The issue about Hadoop 0.20.203.0 is a Hadoop issue at best, and looks like they're already tracking it. I think it should be fixed of course, but it's not a Mahout thing. I am not seeing a Mahout issue here then. On Mon, Jul 11, 2011 at 8:43 PM, Marc Millstone <[email protected]> wrote: > Good News: The offending unit tests passes with the combination of > > IBM JVM + Mahout Trunk + Hadoop 0.20.2 > > > To recap: > > IBM JVM + Mahout .5 + Hadoop 0.20.2 fails with the error > > testStartParallelFPGrowth(org.apache.mahout.fpm.pfpgrowth.PFPGrowthTest): > expected:<{[D=0, E=1, A=0, B=0, C]=1}> but was:<{[A=0, B=0, C=1, D=0, > E]=1}> > > IBM JVM + Mahout trunk + Hadoop 0.20.203.0 fails with user initialization > problems due to IBM JVM and the sun com.sun.security.auth.UnixPrincipal > class. > Is this dependency going to be updated in the future? It does seem to > limit the use of non-Sun JVM's. > > Are there any tickets that should be filed? > > > > On Mon, Jul 11, 2011 at 12:22 PM, Marc Millstone <[email protected] > >wrote: > > > Ok. Let me play with this some more. I appreciate your help. > > > > In .5, under the IBM JVM we still get the Unit test failure in PFPGrowth > > though. It seems > > to be a true failure too, not an initialization issue. > > > > I am now going to install the Sun JVM locally on my cluster and link > > against > > that to see if things pass for Mahout .5. > > > > Feel free to contact me off list if this is getting to be too detail > > oriented, etc for this list. > > > > On Mon, Jul 11, 2011 at 12:18 PM, Sean Owen <[email protected]> wrote: > > > >> I strongly suspect it is because we updated to Hadoop 0.20.203.0. 0.5 > >> depended on 0.20.2. You could locally go back to depending on 0.20.2; it > >> ought to still work fine with the previous version. > >> > >> On Mon, Jul 11, 2011 at 8:14 PM, Marc Millstone <[email protected]> > >> wrote: > >> > >> > Ok. I am starting to think this could be a IBM JVM thing > >> > after finding this JIRA: > >> https://issues.apache.org/jira/browse/HADOOP-7211 > >> > > >> > > >> > Could it be that something changed between trunk and .5 that caused > this > >> > issue to manifest itself in different ways? > >> > In .5, we just got the wrong answer and in trunk, we are getting this > >> > permission issue? > >> > > >> > marc > >> > > >> > > >> > > >> > > > > >
