[jira] [Created] (SPARK-29484) Very Poor Performance When Reading SQL Server Tables Using Active Directory Authentication

["Scott Black (Jira)" <ji...@apache.org>]

Scott Black created SPARK-29484:
-----------------------------------

             Summary: Very Poor Performance When Reading SQL Server Tables Using Active Directory Authentication
                 Key: SPARK-29484
                 URL: https://issues.apache.org/jira/browse/SPARK-29484
             Project: Spark
          Issue Type: Bug
          Components: Spark Core
    Affects Versions: 2.4.3
         Environment: Test Case

jars for ms sql client

 

adal4j 1.6.4

oauth2-oidc-sdk 6.16.2

json-smart 2.3

nimbus-jose-jwt 7.9

mssql-jdbc 7.4.0.jre8

 

Slow JDBC URL using AD

"jdbc:sqlserver://<SERVER NAME>:1433;database=<DB NAME>;user=<AD USERNAME>;password=<PASSWORD>;encrypt=true;ServerCertificate=false;trustServerCertificate=true;hostNameInCertificate=*.database.windows.net;loginTimeout=30;authentication=ActiveDirectoryPassword"

 

URL For Expect Performance Using SQL Server Account

 

"jdbc:sqlserver://<SERVER NAME>:1433;database=<DB NAME>;user=<SS USERNAME>;password=<PASSWORD>;encrypt=true;ServerCertificate=false;trustServerCertificate=true;hostNameInCertificate=*.database.windows.net;loginTimeout=30;authentication=sqlPassword"
            Reporter: Scott Black


When creating a dataframe via JDBC from MS SQL Server performance is so bad to be unusable when authentication was performed with Active Directory. When authentication is using SQL Server account performance is as expected. Created a java that that connected to the same Azure SQL database via Active Directory account and performance was the as Sql Server account.  When connecting via Active Directory it could 30 minutes to 1 hour to read a 250 row table compared to 5 seconds using SQL Server account or using a console Java app connecting via AD.

 



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org