Hi,
can you share your code ? I will fix that with you. I suggest to check the
expand() method in the read PTransform and the way you use generic there.
Any plan to donate this IO: I would be happy to review the PR !
Do you leverage some InfluxDB feature (like splitting/sharding) ?
Regards
JB
On 06/30/2017 07:26 AM, P. Ramanjaneya Reddy wrote:
Hi Beam Dev,
We have developed our own sdk io functions for read/write InfluxDBIO
operations in apache BEAM. it is works with default coder, which is
StringUtf8Coder.of().
PCollection<String> output = pipeline.apply(
InfluxDbIO.<String>read()
.withUri("http://localhost:8086";)
.withDatabase("beam"));
With reference mongoDB and JDBC, implemented the read function with
setcoder() options in InfluxDB also, but it is not working.
PCollection<String> output = pipeline.apply(
InfluxDbIO.<String>read()
.withParser(new InfluxDbIO.Parser<String>() {
@Override
public void parse(String input,
InfluxDbIO.ParserCallback<String>
callback) throws IOException {
callback.output(input);
}
})
.withUri("http://localhost:8086";)
.withDatabase("beam")
.withCoder(StringUtf8Coder.of()));----> with coder
getting error as
java.lang.ClassCastException: org.apache.beam.sdk.values.PBegin cannot be
cast to org.apache.beam.sdk.values.PCollection
Thanks & Regards,
Ramanjaneya
--
Jean-Baptiste Onofré
[email protected]
http://blog.nanthrax.net
Talend - http://www.talend.com
