[
https://issues.apache.org/jira/browse/BEAM-6014?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16679156#comment-16679156
]
Valentyn Tymofieiev commented on BEAM-6014:
-------------------------------------------
I checked the output with
{noformat}
bq show $PROJECT:$BQ_DATASET.$TABLE_NAME{noformat}
However, BigQuery console shows that that streaming buffer is not empty:
{noformat}
Streaming buffer statistics
Estimated size 48 B
Estimated rows 12
{noformat}
And we can indeed get the data if we run:
{noformat}
bq query "select * from $BQ_DATASET.$TABLE_NAME"
Waiting on bqjob_r282d5c07653cf8ac_00000166f0f84ff3_1 ... (0s) Current status:
DONE
+-------+---------------+
| month | tornado_count |
+-------+---------------+
| 7 | 8 |
| 4 | 5 |
| 5 | 6 |
| 11 | 9 |
| 9 | 7 |
| 2 | 7 |
| 12 | 10 |
| 1 | 16 |
| 6 | 5 |
| 3 | 6 |
| 8 | 4 |
| 10 | 10 |
+-------+---------------+
{noformat}
> bigquery_tornadoes example does not work on Python Direct Runner
> ----------------------------------------------------------------
>
> Key: BEAM-6014
> URL: https://issues.apache.org/jira/browse/BEAM-6014
> Project: Beam
> Issue Type: Bug
> Components: sdk-py-core
> Reporter: Valentyn Tymofieiev
> Assignee: Ahmet Altay
> Priority: Major
>
> Steps to reproduce:
>
> {noformat}
> <Make new virtual env, install apache-beam[gcp] >
>
> PROJECT=$(gcloud config get-value project)
> BQ_DATASET=${USER}_bq_dataset
> TABLE_NAME=out
> bq rm -rf --project=$PROJECT $BQ_DATASET
> bq mk --project=$PROJECT $BQ_DATASET
> python -m apache_beam.examples.cookbook.bigquery_tornadoes --output
> $BQ_DATASET.$TABLE_NAME --project=$PROJECT
> bq show $PROJECT:$BQ_DATASET.$TABLE_NAME
> {noformat}
> {noformat}
> Last modified Schema Total Rows Total Bytes Expiration Time Partitioning
> Labels
> ----------------- --------------------------- ------------ -------------
> ------------ ------------------- --------
> 07 Nov 14:45:23 |- month: integer 0 0
> |- tornado_count: integer
> {noformat}
>
> The resulting Database is empty, however it has several rows if we run the
> example with Dataflow runner.
> Older versions of SDK also seem to have this problem.
> cc: [~chamikara], [~charleschen].
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
