tajo-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Hyunsik Choi <hyun...@apache.org>
Subject Re: starting difficulties
Date Sun, 27 Jul 2014 04:06:22 GMT
Hi Christian,

In order to clarify your problem, could you share your environments,
such as Hadoop version and JVM version? For some hadoop version, you
need to give '-Dhadoop.version=2.4.0' to maven build command.
Otherwise, incorrect class sources may be built for your binary dist.

Best regards,
Hyunsik

On Sat, Jul 26, 2014 at 6:11 AM, Christian Schwabe
<Christian.Schwabe@gmx.com> wrote:
> Hello Hyunsik,
>
> thanks for your quick response.
> I clone the actual github repository and compile the Version by my own.
> So actual I use the version 0.9.0-Snapshot.
> After your answer it seems als the bugs exist in the current version? Can
> you confirm this?
>
> Kind regards,
> Chris
>
> On 25. Juli 2014 18:39:37 MESZ, Hyunsik Choi <hyunsik@apache.org> wrote:
>>
>> Hi Christian,
>>
>> Above all, thank you for your interest in Tajo.
>>
>> The problem seems to be an old bug.
>> Are you using Tajo 0.9.0? Did you get Tajo via git?
>>
>> Best regards,
>> Hyunsik
>>
>> On Fri, Jul 25, 2014 at 9:23 PM, Christian Schwabe
>> <Christian.Schwabe@gmx.com> wrote:
>>>
>>>  Hello guys,
>>>
>>>  this is the first time for me to write to an email list. Sorry for any
>>> wrong
>>>  formatting.
>>>  I am a student of computer science and use the framework Apache Tajo as
>>> part
>>>  of my thesis. However, I have initial teething problems and
>>> unfortunately do
>>>  not know how to remedy them.
>>>
>>>  First, I want to describe what steps do I do to start Tajo:
>>>  1) Start Tajo with tajo-start.sh
>>>  2) Create a new Table
>>>  3) Start bash tsql
>>>
>>>  Here details to my table contents which is composed from a csv file:
>>>
>>>  default> SELECT * from table1;
>>>  id,  name,  score,  type
>>> ________________________________
>>>
>>>  1,  abc,  1.1,  a
>>>  2,  def,  2.3,  b
>>>  3,  ghi,  3.4,  c
>>>  4,  jkl,  4.5,  d
>>>  5,  mno,  5.6,  e
>>>  (5 rows, 0.275 sec, 59 B selected)
>>>
>>>  Everything seems to work up here. Also substr () or concat () function.
>>> Do I
>>>  use now but the count (*) or sum () I get the following exception:
>>>
>>>  default> SELECT count(*) FROM table1;
>>>  ERROR: 1
>>>  java.lang.ArrayIndexOutOfBoundsException: 1
>>>  at org.apache.tajo.util.BitArray.fromByteBuffer(BitArray.java:84)
>>>  at org.apache.tajo.storage.RawFile$RawFileScanner.next(RawFile.java:243)
>>>  at
>>>
>>> org.apache.tajo.engine.planner.physical.SeqScanExec.next(SeqScanExec.java:254)
>>>  at
>>>
>>> org.apache.tajo.engine.planner.physical.HashAggregateExec.compute(HashAggregateExec.java:51)
>>>  at
>>>
>>> org.apache.tajo.engine.planner.physical.HashAggregateExec.next(HashAggregateExec.java:87)
>>>  at
>>>
>>> org.apache.tajo.engine.planner.physical.StoreTableExec.next(StoreTableExec.java:77)
>>>  at org.apache.tajo.worker.Task.run(Task.java:425)
>>>  at org.apache.tajo.worker.TaskRunner$1.run(TaskRunner.java:406)
>>>  at java.lang.Thread.run(Thread.java:744)
>>>
>>>
>>>  Can you tell me what I'm doing wrong at this point that I have neither
>>> the
>>>  count (*) still can use sum ()?
>>>  Is this a known problem? Is it fixed? When will it be fixed? How can I
>>> fix
>>>  it myself? I suppose that I have a wrong configuration, although I have
>>> not
>>>  had much konfikuriert. Because I can hardly imagine that such trivial
>>>  functionalities not in my used version 0.9.0 are available.
>>>
>>>  Kind regards,
>>>  Christian Schwabe

Mime
View raw message