Hi Tim,
Reference a blog comes from Ververica:
"When you choose RocksDB as your state backend, your state lives as a
serialized byte-string in either the off-heap memory or the local disk."
It also contains many tune config options you can consider.[1]
Best,
Vino
[1]: https://www.ververica.com
Hi Anji,
Actually, I am not familiar with how to partition via timestamp. Flink's
streaming BucketingSink provides this feature.[1] You may refer to this
link and customize your sink.
I can ping a professional committer who knows more detail of FS connector
than me, @kklou...@gmail.com may give
Hi Polarisary,
The fields of your `UserTableFunction` maybe not serializable like
`Connection` and `PreparedStatement`. So you can make them `transient` and
let them not participate in the serialization.
Hope this helps.
Polarisary 于2019年12月26日周四 下午4:47写道：
> Hi all
> When I use udf, it throws
For Streaming Jobs that use RocksDB my understanding is that state is
allocated off-year via RocksDB.
If this is true then does it still make sense to leave 70% (default
taskmanager.memory.fraction) of the heap for Flink Manged memory given that
it is likely not being used for state?Or am I mi
Thanks Vino.
I am able to write data in parquet now. But now the issue is how to write a
dataset to multiple output path as per timestamp partition.
I want to partition data on date wise.
I am writing like this currently that will write to single output path.
DataSet> df = allEvents.flatMap(new
Hi all
When I use udf, it throws Unable to serialize Exception as follows:
Exception in thread "main" org.apache.flink.table.api.ValidationException:
Unable to serialize object 'UserTableFunction' of class
‘udtf.UserTableFunction'.
at
org.apache.flink.table.utils.EncodingUtils.encod
