spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Suryansh Agnihotri <sagnihotri2...@gmail.com>
Subject beeline spark thrift server issue
Date Thu, 13 May 2021 15:22:49 GMT
Hi
I was trying to access spark sql through jdbc but facing some error. I am
trying to run beeline

! /usr/lib/spark/bin/beeline -u
'jdbc:hive2://host:10016/default;transportMode=binary'  -e '' 2>&1| awk
'{print}'|grep -i -e 'Connection refused' -e 'Invalid URL' -e 'Error: Could
not open'

Error: Could not open client transport with JDBC Uri:
host:10016/default;transportMode=binary: java.net.ConnectException:
Connection refused (Connection refused) (state=08S01,code=0)

hive.server2.thrift.port=10016 and mode is binary.
I verified the process is running on this port.
I checked the spark thrift server logs https://pastebin.com/G5Mwaw7E
It says "java.lang.RuntimeException: Unable to instantiate
org.apache.hadoop.hive.ql.metadata.SessionHiveMetaStoreClient"
(pasted the logs in above link).

*I am using spark version 3.1.0 , hive 3.1.2 and hadoop 3.1.2.*
Following this guide
https://spark.apache.org/docs/latest/sql-data-sources-hive-tables.html#interacting-with-different-versions-of-hive-metastore

I set "spark.sql.hive.metastore.version" to 3.1.2 and set jars to point to
hivemetastore 3.1.2 jars but still getting the same error.

>From logs I also got suspicious about a mismatch in the version of
datanucleus core in hive and spark but both use the same version.
https://github.com/apache/hive/blob/branch-3.1/pom.xml#L129
https://github.com/apache/spark/blob/branch-3.1/pom.xml#L184 (edited)

Is this a known issue , how should this be fixed. Let me know if anything
else is required.
Thanks

Mime
View raw message