voducdan opened a new issue, #13370:
URL: https://github.com/apache/iceberg/issues/13370
### Query engine
Spark
### Question
I'm using spark to read data from GCS into DataFrame and write to Iceberg
table. My table's DDL looks like this
```
CREATE TABLE catalog.database.table(
row ARRAY<STRUCT<granularity: ARRAY<STRUCT<avgCPT: STRUCT<amount: STRING,
currency: STRING>>>, metadata: STRUCT<campaignId: BIGINT, orgId: BIGINT>>>,
etl_date STRING)
USING iceberg
PARTITIONED BY (etl_date)
LOCATION 'gs://bucket/database/table'
TBLPROPERTIES (
'compression' = 'snappy',
'format' = 'iceberg/parquet',
'format-version' = '2',
'mergeSchema' = 'true',
'write.distribution-mode' = 'none',
'write.format.default' = 'parquet',
'write.location-provider.impl' =
'com.idz.iceberg.FileOnlyLocationProvider',
'write.object-storage.partitioned-paths' = 'false',
'write.parquet.compression-codec' = 'zstd',
'write.spark.accept-any-schema' = 'true')
```
Later, data sources in GCS have changed their structure, with schema looks
like
```
root
|-- row: array (nullable = true)
| |-- element: struct (containsNull = true)
| | |-- granularity: array (nullable = true)
| | | |-- element: struct (containsNull = true)
| | | | |-- avgCPT: struct (nullable = true)
| | | | | |-- amount: string (nullable = true)
| | | | | |-- currency: string (nullable = true)
| | | | |-- date: string (nullable = true)
| | | | |-- impressions: long (nullable = true)
| | | | |-- localSpend: struct (nullable = true)
| | | | | |-- amount: string (nullable = true)
| | | | | |-- currency: string (nullable = true)
| | | | |-- tapInstallCPI: struct (nullable = true)
| | | | | |-- amount: string (nullable = true)
| | | | | |-- currency: string (nullable = true)
| | | | |-- tapInstallRate: double (nullable = true)
| | | | |-- tapInstalls: long (nullable = true)
| | | | |-- tapNewDownloads: long (nullable = true)
| | | | |-- tapRedownloads: long (nullable = true)
| | | | |-- taps: long (nullable = true)
| | | | |-- totalAvgCPI: struct (nullable = true)
| | | | | |-- amount: string (nullable = true)
| | | | | |-- currency: string (nullable = true)
| | | | |-- totalInstallRate: double (nullable = true)
| | | | |-- totalInstalls: long (nullable = true)
| | | | |-- totalNewDownloads: long (nullable = true)
| | | | |-- totalRedownloads: long (nullable = true)
| | | | |-- ttr: double (nullable = true)
| | | | |-- viewInstalls: long (nullable = true)
| | | | |-- viewNewDownloads: long (nullable = true)
| | | | |-- viewRedownloads: long (nullable = true)
| | |-- insights: struct (nullable = true)
| | | |-- bidRecommendation: struct (nullable = true)
| | | | |-- suggestedBidAmount: struct (nullable = true)
| | | | | |-- amount: string (nullable = true)
| | | | | |-- currency: string (nullable = true)
| | |-- metadata: struct (nullable = true)
| | | |-- adGroupDeleted: boolean (nullable = true)
| | | |-- adGroupId: long (nullable = true)
| | | |-- adGroupName: string (nullable = true)
| | | |-- bidAmount: struct (nullable = true)
| | | | |-- amount: string (nullable = true)
| | | | |-- currency: string (nullable = true)
| | | |-- campaignId: long (nullable = true)
| | | |-- countryOrRegion: string (nullable = true)
| | | |-- deleted: boolean (nullable = true)
| | | |-- keyword: string (nullable = true)
| | | |-- keywordDisplayStatus: string (nullable = true)
| | | |-- keywordId: long (nullable = true)
| | | |-- keywordStatus: string (nullable = true)
| | | |-- matchType: string (nullable = true)
| | | |-- modificationTime: string (nullable = true)
| | | |-- orgId: long (nullable = true)
| | |-- other: boolean (nullable = true)
```
At that time, when I run `df.writeTo(output_path).overwritePartitions()` I
got error `pyspark.errors.exceptions.captured.IllegalArgumentException: Field
date not found in source schema`
What am I missing here? Is this a known bug?
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
