Hi Guillaume,
Guillaume wrote:
Hi all,
I would like to test your tool.
So I downloaded the sources from GitHub then compiled them without any
error.
I deployed the war on Tomcat and loading and querying are functional
throught REST API (using loadrdf and queryrdf).
As loading is quite slow, I would like to test BulkLoad.
My problem is that I cannot use BulkLoad.
Tables are well created into Accumulo instance but remains empty.
There are no specific errors as the MapReduce jobs finish correctly.
How did you verify this? Did you scan the table and find no entries in
the table? If relevant, did you verify that the user scanning the table
has the proper Accumulo authorizations to read the data?
The command I launched is :
hadoop jar accumulo.rya-3.2.10-SNAPSHOT-shaded.jar
mvm.rya.accumulo.mr.fileinput.BulkNtripsInputTool
-Dac.zk=instance1:2181,instance2:2181,instance3:2181 -Dac.instance=accumulo
-Dac.username=root -Dac.pwd=**** -Drdf.tablePrefix=triplestore_
-Dio.sort.mb=64 /path/to/input/file.nt
The only thing I noticed is in Zookeeper logs. I can read down there:
INFO org.apache.zookeeper.server.PrepRequestProcessor: Got user-level
KeeperException when processing sessionid:0x1521c6d09850005 type:create
cxid:0x7f5c8 zxid:0x8600009422 txntype:-1 reqpath:n/a Error
Path:/accumulo/63820edb-74fd-4369-a5f2-b598a12aaafe/fate/tx_0061a9e5da06e3dd
Error:KeeperErrorCode = NodeExists for
/accumulo/63820edb-74fd-4369-a5f2-b598a12aaafe/fate/tx_0061a9e5da06e3dd
This is not indicative of a problem. If you are not seeing an error in
your invocation of Rya (and thus Accumulo), it is very likely ZooKeeper
is being used as intended.
Do you have any idea that would help me to make it work?
Regards,
--
Guillaume DC
