Terraform中BigQuery作业显示成功但云端未完成，报错求助

核心原因 #

日志报错Cannot set destination table in jobs with DDL statements明确指出问题：当查询语句是DDL（如CREATE TABLE ... AS SELECT）时，不能同时在Terraform的query块中配置destination_table。因为DDL语句本身已经指定了要创建的目标表，额外设置destination_table会导致BigQuery作业冲突，最终作业标记为完成但实际未执行表创建操作。

额外问题 #

原SQL语句存在两处语法错误：

• 字段列表末尾多了逗号：email,FROM应改为email FROM
• 项目名拼写错误：myporject应为myproject

修正后的Terraform配置 #

移除destination_table块，修正SQL语法：

resource "google_bigquery_job" "user_data_job" {
  job_id     = "job_query_${random_id.random_job_id.hex}"
  location   = "US"

  labels = {
    "job_type" = "create_table"
  }

  query {
    query = "CREATE TABLE myproject.shop45_dataset.user_data AS SELECT first_name, last_name, email FROM myproject.shop45_dataset.shop_data"

    allow_large_results = true
    flatten_results     = true
  }

  depends_on = [
    google_bigquery_job.import_job
  ]
}

替代方案（不使用DDL） #

如果希望通过destination_table指定目标表，可改用普通SELECT语句，去掉DDL部分：

resource "google_bigquery_job" "user_data_job" {
  job_id     = "job_query_${random_id.random_job_id.hex}"
  location   = "US"

  labels = {
    "job_type" = "create_table"
  }

  query {
    query = "SELECT first_name, last_name, email FROM myproject.shop45_dataset.shop_data"

    destination_table {
      project_id = "myproject"
      dataset_id = "shop45_dataset"
      table_id   = "user_data"
    }

    create_disposition = "CREATE_IF_NEEDED"
    write_disposition  = "WRITE_EMPTY"
    allow_large_results = true
    flatten_results     = true
  }

  depends_on = [
    google_bigquery_job.import_job
  ]
}

内容的提问来源于stack exchange，提问作者PRABHAT KUMAR