Damon P. Cortesi
Damon P. Cortesi
The Streaming API actually delivers results using chunked encoding, but due to a quirk in their load balancer (so I hear), twitter-stream actually parses everything OK. Other streams (Gnip Power...
The driver currently syncs all databases and all tables - this may not be desirable behavior.
Currently, when queries are built, we'll use the `user.timezone` setting. This could lead to weirdness because timestamps in Athena are UTC. For example, a simple query that tries to filter...
TODO: - [ ] See if there's a way to add a test :)
Federated data sources can, sometimes silently, fail the sync step as seen in #55 and #63. We should have better error handling for those.
We need a way to add a test suite. Unfortunately this requires uploading sample data to S3 and executing Athena queries so it could be a little complex.
Queries cancel fine, but there's still a `WARN` message in the Metabase logs. ``` 05-09 08:42:44 DEBUG async.util :: Request canceled, canceling future. 05-09 08:42:44 WARN middleware.async :: Unhandled exception,...
Queries execute successfully, but there's a `WARN` log message regarding the driver response. ``` 05-09 08:39:13 WARN middleware.async :: Unhandled exception, exepected `catch-exceptions` middleware to handle it. java.lang.Exception: Invalid response...
I left and returned to AWS. Since that time, Glue has: - Added support for defining table names with crawlers - Added support for schema evolution - Upgraded to Python...
Crawlers can now [use existing tables](https://docs.aws.amazon.com/glue/latest/dg/define-crawler.html#crawler-source-type) as a crawler source, which may give us the ability to deprecate our custom partitioning code that searches S3 for new partitions. In combination...