有没有办法返回加载到BigQuery的坏记录数？

Question

当我加载到BigQuery数据集时，有没有办法使用Python API从作业中获取错误记录的数量？

Answer 1

以下是可用于加载作业的统计信息：

statistics.load nested          object  [Output-only] Statistics for a load job.
statistics.load.inputFileBytes  long    [Output-only] Number of bytes of source data in a load job.
statistics.load.inputFiles      long    [Output-only] Number of source files in a load job.
statistics.load.outputBytes     long    [Output-only] Size of the loaded data in bytes. Note that while a load job is in the running state, this value may change.
statistics.load.outputRows      long    [Output-only] Number of rows imported in a load job. Note that while an import job is in the running state, this value may change.

如果您知道预期的行数，则可以使用outputRows

找出错误的行数

同时 - 您可以控制加载作业中允许的错误记录数量：

configuration.load.allowJaggedRows
configuration.load.ignoreUnknownValues
configuration.load.maxBadRecords

所有这些都可以在中找到 https://cloud.google.com/bigquery/docs/reference/v2/jobs#configuration.load

和
https://cloud.google.com/bigquery/docs/reference/v2/jobs#statistics.load