Hi,
On Thu, 8 Mar 2018 15:37:13 +0800
Jeff Zhang wrote:
JZ> Is it possible for you to upload to hdfs and consume it via hdfs in
JZ> flink ? AFAIK, hdfs web ui allow user to upload data
yes it is possible to upload via a separate tool.
I was asking because it would be very convenient for the us
Hi,
here is a more concrete description of what I intend to do:
There are some analysis which involve parsing data which is stored in
csv files. Up until now these files have to be uploaded manually to the
server to be able to process them.
It would be great to have a simple file upload form wit
Hi,
where would I need to look regarding documentation on how to implement
upload functionality in a zeppelin notebook.
The basic idea is to pick a file and pipe this either onto a server
directory or directly into an interpreter if possible.
Regards,
Jens
--
CTO, Wegtam GmbH, 07. Lenzing 201
Hi,
On Fri, 17 Nov 2017 13:12:11 +0100
André Schütz wrote:
AS> When using the following import, we get an error:
AS>
AS> import org.apache.flink.streaming.connectors.cassandra._
AS>
AS> error: object connectors is not a member of package
AS> org.apache.flink.streaming import
AS> org.apache.fli
Hi,
On Thu, 9 Nov 2017 14:05:42 +0100
Jens Grassel wrote:
JG> However, after following the documentation it seems that zeppelin
JG> more or less starts its own flink node and does not connect to the
JG> existing cluster. Jobs are executed fine and logged in the
JG> interpreter l
Hi,
we try to use an existing flink cluster with zeppelin following the
official documentation at:
http://zeppelin.apache.org/docs/0.8.0-SNAPSHOT/setup/deployment/flink_and_spark_cluster.html
We are using current zeppelin compiled from source with flink 1.3.2
support and a flink 1.3.2 cluster.
Hi,
On Wed, 08 Nov 2017 14:47:00 +
Jeff Zhang wrote:
JZ> I think you misunderstand the connecting existing process, it means
JZ> connecting to an existing zeppelin interpreter process. Here it is
JZ> flink interpreter process. But seems
JZ> you are configure it to connect to flink cluster. I
On Wed, 08 Nov 2017 13:52:50 +
Jeff Zhang wrote:
JZ> zeppelin-interpreter-flink-whatever.log is what I mean
As mentioned this log file is only created by the embedded flink
interpreter. If the "Connect to existing process" option is used the
file is not created and even if it exist no data
Hi,
On Wed, 08 Nov 2017 13:23:11 +
Jeff Zhang wrote:
JZ> There's another log file which is interpreter log under the logs
JZ> folder, could you check that file ?
there are just two logfiles in that folder:
zeppelin-whatever.log
zeppelin-whatever.out
However if we configure the interpreter
Hi,
we've tried to connect to a flink (1.1.3) cluster (started with
`start-cluster.sh`) from zeppelin 0.7.3 (bin-all).
The cluster is running at localhost:6123 and we configured the
interpreter using:
[X] Connect to existing process
Host : localhost
Port : 6123
But if we try to run flink code fr
10 matches
Mail list logo
