nicor88

For `take` operations will be nice to handle both amount of record and `within timeout`, to avoid to stale processing when the amount of values is not reached.

@vertexclique check the implementation of take on faust. It supports an amount of records, in combination with a parameter called `within`. Imaging to have a `stream.take(100)`. > The problem with...

the issue can be solved by installing a specific version of markup safe `markupsafe==2.0.1 `

@shlomiken this should be fixed. I removed the provider and add a couple of make commands: make infra-init make infra-plan Can you also try and let me know if it...

@aaronsteers I like the idea of having modules in terraform. This repo contains the really first iteration. I need to have some time to take a look into this.

@cicerojmm thanks for your work, I'm not a big fan of using EFS to mount the dags. Instead, I prefer to bake new images and release the changes. Overall it...

Not sure if totally related to this, but I'm also getting some weird behavior with the k8s executor and pod_overwrites. See details [here](https://apache-airflow.slack.com/archives/CCV3FV9KL/p1663609356775399). Then when I'm clearing the task from...

@Tomme could you have a look??

@rumbin dropping a table in the current adaptor setup means: looking the table location path, cleaning up via S3 apis, then finally dropping the table in athena. See implementation [here](https://github.com/Tomme/dbt-athena/blob/master/dbt/adapters/athena/impl.py#L83)....

@rumbin well, in athena world a drop statement of a table is not like Postres/Redshift/Snowflake. But now with Iceberg should get closer to it. Anyhow, the iceberg helper add a...