After that enchantment, each worker could request multiple assignment-batches, before its previous batches were processed by the Controller. This means that for each batch that was processed, the Controller was deleting from the "assignment" table, all the assignments (-batches) delivered to the Worker that brought that batch, even though the "attempt" and "payload" records for the rest of the batches were not inserted in the DB yet. So in a new assignments-batch request, the same publications that were already under processing, were delivered to the same or other Workers.
Now, for each finished batch, only the assignments of that batch are deleted from the "assignment" table.
One side effect of using the parquet-files, is that the timestamps are now BIGDECIMAL numbers, instead of "Timestamp" objects, but, converting them to such objects is pretty easy, if we ever need to do it.
- Code polishing.
- Fix an NPE, when the "getTestUrls"-endpoint is called. It was thrown because of an absent per-thread initialization of some thread-local variables.
- Fix JdbcTemplate error when querying the "getFileLocationForHashQuery".
- Fix the "S3ObjectStore.isLocationInStore" check.
- Fix not catching/handling some exceptions.
- Fix/improve log-messages.
- Optimize the "getFileLocationForHashQuery" to return only the first row. In the latest change, without this optimization, the query-result would cause non-handling the same-hash cases, because of an exception.
- Optimize the "ImpalaConnector.databaseLock.lock()" positioning.
- Update the "getTestUrls" api-path.
- Optimize list-allocation.
- Re-add the info-message about the successful emptying of the S3-bucket.
- Code cleanup.
Lampros Smyrnaios
Antonis Lempesis