Support writing to BQ via Storage Write API with SchemaTransforms #23988
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
ahmedabu98
force-pushed
the
bq_write_schematransforms
branch
from
acde5b7
to
7f41639
Compare
ahmedabu98
force-pushed
the
bq_write_schematransforms
branch
from
6e509be
to
8177bca
Compare
|
R: @johnjcasey |
|
Stopping reviewer notifications for this pull request: review requested by someone other than the bot, ceding control |
ahmedabu98
force-pushed
the
bq_write_schematransforms
branch
from
613c463
to
633ef7d
Compare
ahmedabu98
force-pushed
the
bq_write_schematransforms
branch
from
633ef7d
to
8aa51d4
Compare
pabloem
reviewed
Dec 7, 2022
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
pabloem
reviewed
Dec 7, 2022
Comment on lines
226
to
228
| public abstract Builder setNumStorageWriteApiStreams(Integer numStorageWriteApiStreams); | ||
|
|
||
| public abstract Builder setNumFileShards(Integer numFileShards); |
There was a problem hiding this comment.
I think we should have default values for these (file shards are not relevant for storage, right?) - and avoid exposing them if possible. In fact, we should use auto sharding if it's supported.
There was a problem hiding this comment.
will do, that sounds better. fyi support was added in #16795
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
Comment on lines
348
to
352
| if (configuration.getTriggeringFrequencySeconds() != null) { | ||
| write = | ||
| write.withTriggeringFrequency( | ||
| Duration.standardSeconds(configuration.getTriggeringFrequencySeconds())); | ||
| } |
There was a problem hiding this comment.
I think we should probably have a default value for this. Something like:
if input.isUnbounded():
write = write.withTriggeringFrequency(config.getTriggeringFrequency() == null ? DEFAULT_VALUE : conf.getTriggeringFrequency())
so that this parameter can be optional even for streaming pipelines
There was a problem hiding this comment.
SGTM, what default do you think is good here? I've often seen 60s
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
|
retest this please |
|
Thank you @pabloem, PTAL |
|
Friendly ping :) needs a final review |
|
R: @johnjcasey |
3 tasks
pabloem
reviewed
Dec 19, 2022
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
pabloem
reviewed
Dec 19, 2022
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
pabloem
reviewed
Dec 19, 2022
...pache/beam/sdk/io/gcp/bigquery/providers/BigQueryStorageWriteApiSchemaTransformProvider.java
Outdated
Show resolved
Hide resolved
|
Comments addressed, PTAL :) |
|
lgtm thanks! |
lostluck
pushed a commit
to lostluck/beam
that referenced
this pull request
Dec 22, 2022
…ache#23988) * storage write api provider class skeleton * storage write schematransform configuration with validation * small fixes, more validation checks, and setting up the storage write transform * PCollectionRowTuple with two outputs: failed rows and (errorrors+failed rows) * spotless * bq services does not belong as a configuration field * add AT_LEAST_ONCE semantics. also set test bq services * test config validation and successful writes * test for failed rows * test for failed rows * some fixes * experiment * use autoSharding, set default triggering frequency, use beameam schema * use Long instead of Integer * address comments: use autosharding, 5s commit interval, change input and output tags
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.
Adding SchemaTransform capability to BigQueryIO Storage Write API.