Andy-- sorry, I accidentally left in "--thread 2.0C" when I shouldn't have. I
did get the errors with it, but I also got them without it, just "mvn
-Dmaven.repo.local=/tmp/maven clean install". I just tried it again, no
threading trickiness, starting from an empty Maven repo, and got the same
problems, and the run finished up with dozens of gigs free on disc.
Maybe someone else who uses a Mac can try this? It's so weird and I've never
seen it before. I really suspect that it's something odd about me and not Jena.
{grin} But I thought I should report it.
---
A. Soroka
The University of Virginia Library
> On May 11, 2016, at 12:14 PM, Andy Seaborne <[email protected]> wrote:
>
> As far as I know, nothing has changed in the area of jena-jdbc and you've got
> through TDB testing to get to jena-jdbc.
>
> Did it ever work at any version? Especially with/without "--threads 2.0C"
>
> I just tried "--thread 2.0C" with an empty repo and got weird things
> happening. maven 3.3.3 fell apart.
>
> (e.g. "Failed to install project artifacts: MavenProject:
> org.apache.jena:jena-text:3.1.1-SNAPSHOT:: Artifact is not fully assembled")
> see below.
>
> "--threads 2" worked.
>
> (I have a quad core, hyper threaded CPU)
>
> Lots of things happen out of order so if disk is shared between tests things
> will go wrong.
>
>
> There are two kinds of exceptions:
>
> case 1:
>
> java.lang.Exception: Unexpected exception
>
> expected<java.sql.SQLFeatureNotSupportedException> but
> was<org.apache.jena.tdb.base.file.FileException>
>
> case 2:
>
> org.apache.jena.tdb.base.file.FileException: BlockMgrMapped.segmentAllocate:
> Segment = 0
>
>
> What's the disk setup?
> (have you run out of disk space? Macs report at least all the space for
> sparse files, maybe they use that space)
>
> If a file area is shared in some way, parallel might mess up.
>
> Andy
>
> "--thread 2.0C" ==>
> [INFO]
> ------------------------------------------------------------------------
> [INFO] Reactor Summary:
> [INFO]
>
> (edited to fit)
>
> Parent ............................... SUCCESS [ 1.284 s]
> Shadowed external libraries .......... SUCCESS [ 2.187 s]
> IRI .................................. SUCCESS [ 4.612 s]
> Base Common Environment .............. SUCCESS [ 7.816 s]
> Core ................................. SUCCESS [ 35.050 s]
> ARQ (SPARQL 1.1 Query Engine) ........ SUCCESS [ 38.207 s]
> TDB (Native Triple Store) ............ SUCCESS [01:01 min]
> Libraries POM ........................ SUCCESS [ 0.185 s]
> Command line tools ................... SUCCESS [ 18.849 s]
> SPARQL Text Search ................... SKIPPED
> SPARQL Spatial Search ................ SKIPPED
> Data Tables for RDF and SPARQL ....... SUCCESS [ 6.451 s]
> SDB (SQL based triple store) ......... SKIPPED
> Fuseki1 .............................. SKIPPED
> Fuseki - A SPARQL 1.1 Server ......... SUCCESS [ 0.183 s]
> Fuseki Server Engine ................. SKIPPED
> Fuseki WAR File ...................... SKIPPED
> Fuseki Server Standalone Jar ......... SKIPPED
> Fuseki Binary Distribution ........... SKIPPED
> Security Permissions ................. SUCCESS [ 18.584 s]
> JDBC Parent .......................... SUCCESS [ 0.732 s]
> JDBC Core API ........................ SUCCESS [ 34.060 s]
> JDBC Remote Endpoint Driver .......... SKIPPED
> JDBC In-Memory Driver ................ SUCCESS [ 12.506 s]
> JDBC TDB Driver ...................... SKIPPED
> JDBC Driver Bundle ................... SKIPPED
> Maven Plugins, including schemagen ... FAILURE [ 10.110 s]
> Elephas .............................. SUCCESS [ 0.422 s]
> Elephas - Common API ................. SUCCESS [ 3.943 s]
> Elephas - I/O ........................ SKIPPED
> Elephas - Map/Reduce ................. SUCCESS [ 18.601 s]
> Elephas - Statistics Demo App ........ SKIPPED
> Distribution ......................... SUCCESS [ 7.141 s]
> OSGi ................................. SUCCESS [ 0.291 s]
> OSGi bundle .......................... SUCCESS [ 13.043 s]
> OSGi Karaf features .................. SUCCESS [ 0.892 s]
> Extras ............................... SUCCESS [ 0.301 s]
> Extras - Query Builder ............... SUCCESS [ 14.210 s]
> [INFO] Apache Jena ........................................ SUCCESS [ 0.948 s]
>
> On 11/05/16 16:14, A. Soroka wrote:
>> For Mac OS X, using
>>
>> wget --no-check-certificate
>> https://dist.apache.org/repos/dist/dev/jena/source/jena-3.1.0-source-release.zip
>> unzip jena-3.1.0-source-release.zip
>> cd jena-3.1.0
>> mvn --threads 2.0C -Dmaven.repo.local=/tmp/maven clean install
>>
>> I got errors (not failures) in jena-jdbc-driver-tdb, specifically for
>> TestTdbDiskResultSets. I've put it up here:
>>
>> https://gist.github.com/ajs6f/fded4a65dba235f3ad614c7f89444d99
>>
>> Lots and lots of org.apache.jena.tdb.base.file.FileExceptions. Does that
>> look like a real problem or just something weird about my setup?
>>
>> ---
>> A. Soroka
>> The University of Virginia Library
>>
>>> On May 9, 2016, at 8:56 AM, Andy Seaborne <[email protected]> wrote:
>>>
>>> On 09/05/16 13:44, Stian Soiland-Reyes wrote:
>>>>> OSGi didn't work at 3.0.1 either as I understand PR#134
>>>> No, 3.0.1 jena-osgi was also broken because of the ServiceLoader issue.
>>>>
>>>
>>> That's what I meant - jena-osgi did not work whatever the reason.
>>>
>>> As it took so long to show up, we can conclude the usage of the
>>> project-produced OSGi bundle is low.
>>>
>>>> Due to the Karaf issue you mention, then I suggest adding
>>>> org.apache.servicemix.bundles.xerces as an
>>>> informational <optional>true</optional> dependency.
>>>>
>>>> See https://github.com/apache/jena/pull/140
>>>
>>> Great - I hope OSGi users will test that out.
>>>
>>> Andy
>>>
>>
>
