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 <millst...@gmail.com>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 <sro...@gmail.com> 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 <millst...@gmail.com>
>> 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
>> >
>> >
>> >
>>
>
>
