Hi Andrea,
For your first question, I think you are right, but the basis is set by the
default value for `parallelism.default` in flink-conf.yaml. [1]
For your second question, I guess you use `forward` function between
"learn" and "select" methods. Am I right?
That exception is an expected behav
Hi Andrea,
AFAIK, `keyBy` function you used will wrap all keys you selected into
`Tuple`. You can use `Tuple.f0` to get your key, whose type will be
`String`.
If you want the KeyedStream has String Type for its key, you can use
`KeySelector` in keyBy function. [1]
Hope this will help you.
Best Re
Hi all,
I'm trying to implement a time ordering inside a stream using window
function. Then my purposes is to order the element inside a tumbling window.
This is my code (written following the doc):
DataStream LCxAccStream = env
.addSource(new FlinkKafkaConsumer010
Some updates on this:
Aside from reworking how the S3 directory handling is done, we also looked
into supporting S3 different than we currently do. Currently support goes
strictly through Hadoop's S3 file systems, which we need to change, because
we want it to be possible to use Flink without Hado
Hi!
The problem should be fixed in the latest master.
https://github.com/apache/flink/commit/536675b03a5050fda9c3e1fd403818cb50dcc6ff
Release should come soon...
Best,
Stephan
On Wed, Oct 4, 2017 at 12:50 PM, Stephan Ewen wrote:
> Will be opening a PR for file system configuring which should
Hi,
I read the doc about parallelism, parallel execution and job scheduling but
however I have some doubts about parallelism.
1.
In my first try I unset parallelism in my code and commented
parallelism.default key in link-conf file. In this case I supposed the
parallelism was set by Flink automat
We have a data which is broad and slow; hundreds of thousands of
keys, a small number will get an event every few seconds, some get
an event every few days, and the vast majority will get an event
in a few times an hour. Let's say then that keeping this data
Hi!
I am quite new to Flink CEP and try to define a state change pattern
with it. This means that only discrete changes in the event stream
should be detected i.e.
a a b b - triggers a single change from a to b
Considering b the "bad" state, I would like to additionally recognize
the state chang
