Hi Nick
Sorry for the late jump in.
Just wondering why you call putAll of RocksDBMapState and has
RocksDBMapState#clear() followed. seems the state will always be empty
after the process.
Best,
Congxian
Yun Tang 于2020年6月16日周二 下午7:42写道：
> Hi Nick
>
> From my experience, it's not easy to tune
Hi
Sorry to jump in late.
After read the previous email. I have such assumptions, and please correct
me if I'm wrong:
- RocksDBStateBackend with incremental checkpoint
- at least once mode
- the parallelism for the stateful operator is 8
- checkpoint may take too long to complete
- has fix rate in
Bhaskar,
> Glad to know some progress.
Yeah, some progress. Yet overnight run didn't look as good as I hoped.
The throttling required to not crash during snapshots seems to be quite
different from the throttling required to crash not during snapshots.
So the lowest common denominator is quit
Inline comments for your questions.
Is it similar to the first? How the second is different?
The first leverages the K8s operator to make standalone job cluster running
on K8s easier. However,
the second is more Flink native. We have an embedded K8s client in Flink
and could use "kubernetes-sessi
