Hi All,

I am a newer for Flume NG system, and I have seen that Flume NG could
support Avro data file output to HDFS, but seems that no official Source
could read records from Avro data files. So should I implement a custom
Source to do this? If so, do you know any examples to implement this?
Thanks in advance.

Best Regards,

Eric

Hi,

Just noticed this while cleaning my inbox. Sorry for the late reply.
Do you mean how should you read the files once they are written into
HDFS?

Brock

On Thu, Sep 27, 2012 at 8:58 AM, Eric Jiang <[EMAIL PROTECTED]> wrote:
> Hi All,
>
>
>
> I am a newer for Flume NG system, and I have seen that Flume NG could
> support Avro data file output to HDFS, but seems that no official Source
> could read records from Avro data files. So should I implement a custom
> Source to do this? If so, do you know any examples to implement this? Thanks
> in advance.
>
>
>
> Best Regards,
>
> Eric
>
>

--
Apache MRUnit - Unit testing MapReduce - http://incubator.apache.org/mrunit/

If this is for streaming Avro files off the local filesystem, then it's in
scope for https://issues.apache.org/jira/browse/FLUME-1633

Regards
Mike

On Wed, Oct 17, 2012 at 7:46 AM, Brock Noland <[EMAIL PROTECTED]> wrote:

> Hi,
>
> Just noticed this while cleaning my inbox. Sorry for the late reply.
> Do you mean how should you read the files once they are written into
> HDFS?
>
> Brock
>
> On Thu, Sep 27, 2012 at 8:58 AM, Eric Jiang <[EMAIL PROTECTED]>
> wrote:
> > Hi All,
> >
> >
> >
> > I am a newer for Flume NG system, and I have seen that Flume NG could
> > support Avro data file output to HDFS, but seems that no official Source
> > could read records from Avro data files. So should I implement a custom
> > Source to do this? If so, do you know any examples to implement this?
> Thanks
> > in advance.
> >
> >
> >
> > Best Regards,
> >
> > Eric
> >
> >
>
>
>
> --
> Apache MRUnit - Unit testing MapReduce -
> http://incubator.apache.org/mrunit/
>