Hi, I'm trying to create a datastream in Pyflink. Reading from Azure ADLS2 and writing to ADLS2. Using version 1.17 as writing to ADLS2 is supported from 1.17 I believe. I'm getting the following error. • Added plugins for azure in azure-fs-hadoop • Tried adding plugin for fs-hadoop-shaded (not sure if this will help) Is there any sample project for streaming to/from Azure ADLS2. Thanks for your help

Could you share the jar file of azure-fs-hadoop used?

@Dian Fu Built this Jar from release1.17 branch. Let me know if I can get compiled jar for 1.17.

@Vishnu Prashaanth Penu Sathiyamoorthy Have you configured

hadoop.security.group.mapping

explicitly to

org.apache.hadoop.security.JniBasedUnixGroupsMappingWithFallback

? If yes, you need to configure it to

<http://org.apache.flink.fs.shaded.hadoop3.org|org.apache.flink.fs.shaded.hadoop3.org>.apache.hadoop.security.JniBasedUnixGroupsMappingWithFallback

.

I've not configured that property. I've just tried to use out of the box configs and property. Do we need to configure in flink_conf.yaml?

It may be configured somewhere else in your environment. See https://lists.apache.org/thread/y2x78zqsmnsnh1wkvgv6pg793ttztw6d which is a similar issue.

@Dian Fu I just tried your suggestion and it does not seem to work