incubator-crunch-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Josh Wills <jwi...@cloudera.com>
Subject Re: Correct usage Thrift PTypes
Date Thu, 20 Dec 2012 19:08:55 GMT
Should be fixed now-- sorry about that. Issue is CRUNCH-137 if you need to
re-open it.


On Thu, Dec 20, 2012 at 11:02 AM, Josh Wills <jwills@cloudera.com> wrote:

> It looks to me like we shouldn't be referencing getConfiguration() in that
> context-- the TBase instance we create doesn't implement Configurable, so
> the Configuration object has no effect. I'll file a bug for it and then
> submit the one-line fix.
>
>
> On Thu, Dec 20, 2012 at 10:58 AM, Josh Wills <jwills@cloudera.com> wrote:
>
>> Hrm, that sounds like a regression. You're saying the Thrift PTypes
>> worked fine in 0.4.0-incubating?
>>
>>
>> On Thu, Dec 20, 2012 at 10:51 AM, Whitacre,Micah <
>> MICAH.WHITACRE@cerner.com> wrote:
>>
>>>  We have some code that is running based off of 0.4.0-incubating.  It
>>> uses the following:
>>>
>>>   PTable <ImmutableBytesWritable, Result> raw = pipeline.read(new
>>> HBaseSourceTarget(…));
>>>  PCollection<PersonRecord> records = raw.parallelDo(newSomeFn(), PTypes.thrifts(PersonRecord.
>>> class, raw.getTypeFamily()));
>>>
>>>  When upgrading to 0.5.0-incubating-SNAPSHOT (technically a fork
>>> because we want it to work off CDH 4.1.1 with MRv1) we now get the
>>> following exception:
>>>
>>> java.lang.NullPointerException
>>> 	at org.apache.crunch.DoFn.getConfiguration(DoFn.java:128)
>>> 	at org.apache.crunch.types.PTypes$ThriftInputMapFn.initialize(PTypes.java:167)
>>> 	at org.apache.crunch.DoFn.setContext(DoFn.java:106)
>>> 	at org.apache.crunch.fn.CompositeMapFn.initialize(CompositeMapFn.java:37)
>>> 	at org.apache.crunch.types.writable.WritableType.initialize(WritableType.java:109)
>>> 	at org.apache.crunch.types.writable.WritableTableType.initialize(WritableTableType.java:107)
>>> 	at org.apache.crunch.impl.mr.emit.IntermediateEmitter.<init>(IntermediateEmitter.java:47)
>>> 	at org.apache.crunch.impl.mr.run.RTNode.initialize(RTNode.java:78)
>>> 	at org.apache.crunch.impl.mr.run.RTNode.initialize(RTNode.java:68)
>>> 	at org.apache.crunch.impl.mr.run.RTNode.initialize(RTNode.java:68)
>>> 	at org.apache.crunch.impl.mr.run.CrunchTaskContext.getNodes(CrunchTaskContext.java:57)
>>> 	at org.apache.crunch.impl.mr.run.CrunchMapper.setup(CrunchMapper.java:41)
>>> 	at org.apache.hadoop.mapreduce.Mapper.run(Mapper.java:138)
>>> 	at org.apache.hadoop.mapred.M
>>>
>>>
>>>  This is odd because tracing through the code it essentially means that
>>> on the call to setContext(…) it is passing a null context into the
>>> setContext(…).  Other forms of our code that don't use PTypes but instead
>>> use Writable seems to work fine.  Is it invalid we are passing in Writable
>>> PTypeFamily into the PTypes methods?  The documentation is not clear about
>>> what can/can't be passed in.  Before I started creating test for this
>>> wanted to make sure we just weren't using it incorrectly.
>>>
>>>  Thanks.
>>>  CONFIDENTIALITY NOTICE This message and any included attachments are
>>> from Cerner Corporation and are intended only for the addressee. The
>>> information contained in this message is confidential and may constitute
>>> inside or non-public information under international, federal, or state
>>> securities laws. Unauthorized forwarding, printing, copying, distribution,
>>> or use of such information is strictly prohibited and may be unlawful. If
>>> you are not the addressee, please promptly delete this message and notify
>>> the sender of the delivery error by e-mail or you may call Cerner's
>>> corporate offices in Kansas City, Missouri, U.S.A at (+1) (816)221-1024.
>>>
>>
>>
>>
>> --
>> Director of Data Science
>> Cloudera <http://www.cloudera.com>
>> Twitter: @josh_wills <http://twitter.com/josh_wills>
>>
>>
>
>
> --
> Director of Data Science
> Cloudera <http://www.cloudera.com>
> Twitter: @josh_wills <http://twitter.com/josh_wills>
>
>


-- 
Director of Data Science
Cloudera <http://www.cloudera.com>
Twitter: @josh_wills <http://twitter.com/josh_wills>

Mime
View raw message