FIX: Stream DB Results to Parquet Files #183
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
rymarczy
commented
Nov 13, 2023
Comment on lines
+104
to
+105
| if not running_in_docker() and not running_in_aws(): | ||
| db_host = "127.0.0.1" |
There was a problem hiding this comment.
pulled this over from dmap_import
rymarczy
commented
Nov 13, 2023
| @@ -16,6 +17,9 @@ def validate_environment( | |||
| process_logger = ProcessLogger("validate_env") | |||
| process_logger.log_start() | |||
|
|
|||
| if private_variables is None: | |||
| private_variables = [] | |||
|
|
|||
There was a problem hiding this comment.
Had to re-structure this function to allow for a private_variable parameter and avoid a pylint too-many-branches flag
rymarczy
commented
Nov 13, 2023
Comment on lines
-59
to
-60
| self.db_manager = DatabaseManager() | ||
|
|
There was a problem hiding this comment.
dropped self.db_manager from being created for all hyper jobs. This would have led to issues with the Hyper file writing ecs that won't be able to connect to our rds.
db_manager is now passed directly into create_parquet and update_parquet methods, as they are the only portions of the class that require db access.
mzappitello
reviewed
Nov 13, 2023
Comment on lines
306
to
316
| with self.session.begin() as cursor: | ||
| result = cursor.execute(select_query).yield_per(batch_size) | ||
| with pq.ParquetWriter(write_path, schema=schema) as pq_writer: | ||
| for part in result.partitions(): | ||
| pq_writer.write_batch( | ||
| pyarrow.RecordBatch.from_pylist( | ||
| [row._asdict() for row in part], schema=schema | ||
| ) | ||
| ) | ||
|
|
||
| return write_path |
There was a problem hiding this comment.
i don't think we need to return the write path since it was provided as part of the input.
mzappitello
suggested changes
Nov 13, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The current Parquet -> Tableau pipeline process is flawed in the amount of memory required to create parquet files from DB SELECT queries. This change is meant to result in a fixed amount of memory usage, no matter what the number of results are returned from a DB query, when creating a parquet file.
This fixed memory usage is achieved by utilizing the
yield_permethod of the SQLAlchemyResultobject, as well as theRecordBatchobject of the pyarrow library.In testing, memory usage for the creation of a parquet file from the
static_stop_timestable maxes out at approximately 5-6GB.If memory usage needs to be further limited, the
write_to_parquetfunction ofDatabaseManageroffers abatch_sizeparameter to limit the number for records flowing into a parquet file per partition.Asana Task: https://app.asana.com/0/1205827492903547/1205940053804614