[
https://issues.apache.org/jira/browse/BEAM-10475?focusedWorklogId=513692&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-513692
]
ASF GitHub Bot logged work on BEAM-10475:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 18/Nov/20 18:25
Start Date: 18/Nov/20 18:25
Worklog Time Spent: 10m
Work Description: aaltay commented on a change in pull request #13292:
URL: https://github.com/apache/beam/pull/13292#discussion_r526323192
##########
File path: sdks/python/apache_beam/transforms/util.py
##########
@@ -780,6 +783,48 @@ def expand(self, pcoll):
self.max_buffering_duration_secs,
self.clock))
+ @typehints.with_input_types(Tuple[K, V])
+ @typehints.with_output_types(Tuple[K, Iterable[V]])
+ class WithShardedKey(PTransform):
+ """A GroupIntoBatches transform that outputs batched elements associated
+ with sharded input keys.
+
+ The sharding is determined by the runner to balance the load during the
+ execution time. By default, it spreads the input elements with the same key
+ to all available threads executing the transform.
+ """
+ def __init__(self, batch_size, max_buffering_duration_secs=None):
+ """Create a new GroupIntoBatches.WithShardedKey.
+
+ Arguments:
+ batch_size: (required) How many elements should be in a batch
+ max_buffering_duration_secs: (optional) How long in seconds at most an
+ incomplete batch of elements is allowed to be buffered in the states.
+ The duration must be a positive second duration and should be given
as
+ an int or float.
+ """
+ self.batch_size = batch_size
+
+ if max_buffering_duration_secs is not None:
+ assert max_buffering_duration_secs > 0, (
+ 'max buffering duration should be a positive value')
+ self.max_buffering_duration_secs = max_buffering_duration_secs
+
+ _pid = os.getpid()
Review comment:
Yes both UUID and making it analogous to Java are good ideas. For my
understanding what is the purpose of this key selection? Are we trying to by
default to keep key space equal to the size of whole set of threads across all
workers?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 513692)
Time Spent: 18h 50m (was: 18h 40m)
> GroupIntoBatches with Runner-determined Sharding
> ------------------------------------------------
>
> Key: BEAM-10475
> URL: https://issues.apache.org/jira/browse/BEAM-10475
> Project: Beam
> Issue Type: Improvement
> Components: runner-dataflow
> Reporter: Siyuan Chen
> Assignee: Siyuan Chen
> Priority: P2
> Labels: GCP, performance
> Time Spent: 18h 50m
> Remaining Estimate: 0h
>
> [https://s.apache.org/sharded-group-into-batches|https://s.apache.org/sharded-group-into-batches__]
> Improve the existing Beam transform, GroupIntoBatches, to allow runners to
> choose different sharding strategies depending on how the data needs to be
> grouped. The goal is to help with the situation where the elements to process
> need to be co-located to reduce the overhead that would otherwise be incurred
> per element, while not losing the ability to scale the parallelism. The
> essential idea is to build a stateful DoFn with shardable states.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
