> Could you list some needs from your point of view ?
I know the typical requirements about extending the native flink sql for
iceberg tables are:
1. There are many table management operations in apache iceberg which can
be friendly if we can express in SQL for end users. such as: expire
iceber
Hi, Openlnx, FG and Jingsong, thanks for your attention about this issue.
I am sorry for the late reply, I’ m working on the further job about the
implement
of merge statement, and now we support this in our internal for the basic SQL
syntax.
About: Implement our flink sql plugin/extensions whi
Hi zoucao,
Thanks for your proposal. I believe this discussion will take us one
step further for iceberg/hudi integration.
## `MERGE` for streaming
I feel that `MERGE` is very good for stream computing. And, this is
where our Flink can have an advantage over other computation systems.
MERGE INT
> In the theory aspect, incremental data should be carefully considered for
streaming data. In this situation, the data flow from target_table to
target_table
will be a loop, and the incremental data with one key will keep going
through
the loop. It looks very strange.
This is the same concern I
I'm currently maintaining the iceberg flink modules from apache iceberg
community.
Currently, the spark has a great integration experience with iceberg format
v2 in batch mode. In this document [1],
The merge into syntax from spark sql extensions does really help a lot when
people want to change
Hi, Martijn Visser, thanks for your reply. Firstly, I am sorry for posting the
discussion twice. I sent the message to the dev mail group from an unsub-
scribed account, but the message was not shown for a while, and I guessed that
the dev mail group would not post an email coming from an unsubscr
Hi zoucao,
I see that this message was posted twice, so I choose to only reply to the
latest one (this one). Thanks for bringing this up for discussion.
I agree that support for a merge statement would be a welcome addition to
Flink SQL for those that are using it for bounded jobs. How would you
