Hi Pranjul,
First of all, you adopted on-heap state backend: HashMapStateBackend, which
would not use native off-heap memory. Moreover, JobManager would not initialize
any keyed state backend instance. And if not enable high availability,
JobManagerCheckpointStorage would also not use direct me
Hi,
We are running the JobManager container with 1024GB out of which 512MB is
allocated to the heap. We observe that the JobManager container's resident
memory is always increasing and it never comes down. Heap usage remains to be
constant and not rising abruptly. Can anyone help here where els
Hi!
if I use parallelism of 2 or 4 - it takes the same time.
>
It might be that there is no data in some parallelisms. You can click on
the nodes in Flink web UI and see if it is the case for each parallelism,
or you can check out the metrics of each operator.
if I don't increase parallelism and
Hi Matthias,
How often do you register the event-time timer?
It is registered per input record, or re-registered a new timer after an
event-time timer is triggered?
Would you please provide your test case code, it would be very helpful for
troubleshooting.
Best wishes,
JING ZHANG
Matthias Broeche
Hi Yu,
The type of field
`com.twosigma.research.options.optticks.core.types.Event.askPrice` in the
original POJO class is Double, right?
If it's right, then next step we should find out why source.readUTF() is
`java.lang.Integer` instead of Double.
Nathan Yu 于2021年8月13日周五 下午9:05写道：
> When the
Hi Wouter,
I suspect that it’s transferring the file venv.zip and so it may take some
time. Does it stuck there forever? Could you share some log file?
Regards,
Dian
> 2021年8月14日 下午4:47，Wouter Zorgdrager 写道：
>
> Hi all,
>
> I'm still dealing with the PyFlink deployment issue as described bel
Hi Bogdan,
There is still no TestHarness for Python ProcessFunction. However it seems a
good idea to provide a TestHarness for Python functions such as
ProcessFunction. I have created
https://issues.apache.org/jira/browse/FLINK-23787 as a following-up.
Regards,
Dian
> 2021年8月13日 下午7:11，Matthi
Hi Yuval,
>From the exception stack, it seems the original sql union all two
LookupTableSource and write to a sink, right?
I think it is reasonable to throw an exception, maybe we could make the
error message more clear to understand.
`LookupableTableSource` is used to look up data from external s
hello all , I read some article , I think many company using flink operator
is using separate cluster for each job and this can be achieved using
flinkk8soperator ? Please can you help me sharing some pointer video /git link
which can help me for installing on AWS- EKS and I have zookeeper/
Hi,
I'm trying to run a UNION ALL query on two LookupTableSource tables defined
with JDBC. When attempting this Flink complains that this is an unsupported
feature:
Caused by: org.apache.calcite.plan.RelOptPlanner$CannotPlanException: There
are not enough rules to produce a node with desired prop
Yes! That did it. Changed to localhost and all works fine now.
I was wrong thinking it would like to connect to Beam SDK worker from my
client machine, hence i added the load balancer.
Thank you Jan!
On Sun, 15 Aug 2021 at 16:45, Jan Lukavský wrote:
> Hi Gorjan,
>
> the address of localhost is
Hi Gorjan,
the address of localhost is hard-coded in the python worker pool (see
[1]). There should be no need to setup a load-balancer for the
worker_pool, if you have it as another container in each TM pod, it
should suffice to replace {beam_sdk_url} with 'localhost'. Each TM will
then have
Hi,
I am not getting anything in the metric
"flink.operator.KafkaConsumer.records-lag-max". Tried checking the value of
Mbean using JMX for
"org.apache.flink.taskmanager.job.task.operator.KafkaConsumer.records-lag-max"
but I am always getting either NaN or 0.
I have checked other Mbeans su
13 matches
Mail list logo
