uot;x=== "+x));
Result:
keys= a1
keys= a0
x=== a0
x=== a2
====x=== a4
x=== a6
x=== a8
But flink on cluster, mapstate not support transfer value, such as :
.proce
s= a0
> ====x=== a0
>x=== a2
>x=== a4
>x=== a6
>x=== a8
>
>
> But flink on cluster, mapstate not support transfer value, such as :
>
> .process(new KeyedProc
Hi Shengjk1,
You should call "mapState.put(key,objectList);" manually after calling
"objectList.add(stringObjectTuple2.f1);" to write it to the state backend. This
is because objectList is just a common Java list object and it will not be
synced to state backend automatic when updated.
Regards
0
x=== a0
x=== a2
x=== a4
====x=== a6
x=== a8
But flink on cluster, mapstate not support transfer value, such as :
.process(new KeyedProcessFunction, String>()
{
...
@Overr
Hiļ¼all
This is my code
.process(new KeyedProcessFunction, String>() {
...
@Override
public void open(Configuration parameters) throws Exception {
super.open(parameters);
StateTtlConfig ttlConfig = StateTtlConfig
.newBuilder(Time.days(7))
.cleanupInBackground()
.setUpdateType(Stat
