is it possible to converta spark dataframe to dynamic frame and then using bookmark feature on the s3 folder used to read data in spark frame

df = spark.read.parquet("s3://folder/")
df = df.withColumn('filename', input_file_name())
AmazonS3_node1697616892615 = DynamicFrame.fromDF(df, glueContext, "s3sparkread")

if this is the code will the folder be bookmarked ? i want to read only new data coming to the folder.

i need to read in spark data frame as i am facing this issue [https://repost.aws/questions/QULSA0zge2QN-_2c_BqIQ_XA/in-aws-glue-not-able-to-add-the-source-file-name-in-the-dynamic-frame]( repost link)

主题

存储分析数据库

标签

亚马逊简单存储服务 (S3)AWS Glue 提取变换和加载数据

语言

English

已提问 1 年前684 查看次数

1 回答

最新
投票最多
评论最多

这些答案有用吗？为正确答案投票，以帮助社区从您的知识中受益。

No, that will just transform the data already read but won't keep track of files.
Bookmarks require a DynamicFrame and a transaction_ctx specified.

专家

Gonzalo Herreros

已回答 1 年前

KG
1 年前
what can i change in this code to do bookmarking , i cannot read the data directly in to dynamic frame as i need the s3 file name in one of the columns.

Do i need to modify this line to have some thing more

AmazonS3_node1697616892615 = DynamicFrame.fromDF(df, glueContext, "s3sparkread")

is it possible to converta spark dataframe to dynamic frame and then using bookmark feature on the s3 folder used to read data in spark frame

相关内容