SeaTunnel Community Meeting about seatunnel HOT 6 OPEN

@hailin0

1. 2.3.4 Release(https://github.com/apache/seatunnel/releases/tag/2.3.4)
2. new type converter(#5872)
3. event listener (#6419)

from seatunnel.

@EricJoy2048
2024 roadmap

from seatunnel.

is zeta support write and read kafka with kerberos ,how to set the conf

from seatunnel.

Whether to consider allowing paimon to support concurrent writes? At present, the Writer will write to different buckets at the same time, which will cause write conflicts during the final commit.
In fact, in the official Paimon Api, it is possible to detect which Bucket the Row is in, build a Writer for the corresponding bucket, and then write to it, so that there will be no write conflicts at the end of the commit.
In fact, the write-only parameter of the Paimon table can be used to solve the write conflict problem, but it will have some drawbacks, such as too many small files, skipping compression and deleting expired data.
I have tried that the java api can be implemented, and I have tried to modify it in the existing Paimon Connector. The test in the example is normal, but there are still problems in testing actual tables in production environment. It is considered that the modified write process cannot completely avoid this problem, and it has to be calculated and implemented at the upper level of the flow

from seatunnel.

Do you consider implementing the process control of the transform? The current transform can only be executed in order, and you cannot perform different transform logic dynamically based on the data.

from seatunnel.

How about publish the docker image when release, the image can include the all connector jar and other lib we need but can't publish to apache package.
Then uses can eaily use or quich start. Also it will help the deployment on k8s.

from seatunnel.