I use HDFS as remote storage for dvc artefacts, kerberos authentication. But when I run ‘dvc push’, I run into the error (but HDFS is accessible from terminal, i.e. I correctly authenticate to hdfs by my keytab file):
Picked up JAVA_TOOL_OPTIONS: -Djava.security.krb5.conf=/usr/adh/etc/krb5.conf
loadFileSystems error:
(unable to get root cause for java.lang.NoClassDefFoundError)
(unable to get stack trace for java.lang.NoClassDefFoundError)
hdfsBuilderConnect(forceNewInstance=1, nn=<my host>, port=8020, kerbTicketCachePath=(NULL), userName=(NULL)) error:
(unable to get root cause for java.lang.NoClassDefFoundError)
(unable to get stack trace for java.lang.NoClassDefFoundError)
/arrow/cpp/src/arrow/status.cc:137: Failed to disconnect hdfs client: IOError: HDFS hdfsFS::Disconnect failed. Detail: [errno 9] Bad file descriptor
ERROR: unexpected error - HDFS connection failed
When I specify settings directly, the error remains:
Picked up JAVA_TOOL_OPTIONS: -Djava.security.krb5.conf=/usr/adh/etc/krb5.conf
loadFileSystems error:
(unable to get root cause for java.lang.NoClassDefFoundError)
(unable to get stack trace for java.lang.NoClassDefFoundError)
hdfsBuilderConnect(forceNewInstance=1, nn=<my host>, port=8020, kerbTicketCachePath=FILE:/tmp/krb5cc_298426831_298426831, userName=user1) error:
(unable to get root cause for java.lang.NoClassDefFoundError)
(unable to get stack trace for java.lang.NoClassDefFoundError)
/arrow/cpp/src/arrow/status.cc:137: Failed to disconnect hdfs client: IOError: HDFS hdfsFS::Disconnect failed. Detail: [errno 9] Bad file descriptor
ERROR: unexpected error - HDFS connection failed
dvc version 2.10.2
I really hope for help, thank you in advance