Skip to end of banner
Go to start of banner

Hive Account

Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 34 Next »

Overview

This account is used by the Snaps in the Hive Snap Pack. The account can be configured with and without Kerberos and supports SSL connection.

You can create an account from Designer or Manager. In Designer, when building pipelines, every Snap that needs an account prompts you to create a new account or use an existing account. The accounts can be created in or used from:

  • Your private project folder: This folder contains the pipelines that will use the account.
  • Your Project Space’s shared folder: This folder is accessible to all the users that belong to the Project Space.
  • The global shared folder: This folder is accessible to all the users within an organization in the SnapLogic instance.

Account Configuration

In Manager, you can navigate to the required folder and create an account (see Configuring Hive Accounts). To create an account for a generic JDBC driver: 

  1. If not already done, upload the JDBC driver for this database as a file for the specific project.
  2. Click Create, then select Hive > Hive Database Account or Generic Hive Database Account (as required).
  3. Supply an account label.
  4. Supply the necessary properties for your database. 
  5. Supply the necessary JDBC driver jars for your driver.
  6. (Optional) Supply additional information on this account in the Notes field of the Info tab.
  7. Click Apply. 

Avoid changing account credentials while pipelines using them are in progress. This may lead to unexpected results, including locking the account.

Account Types 

Hive Database Account

 Hive Database Account

Account Settings

Label

Required. User provided label for the account instance.

Account properties


Hostname

Required. The server address to connect to. 

Default value: None.

Port number

Required. The database server's port to connect to.

Default value: 10000

Database name

Required. The name of the database which the account is to be connected to.
Default value: None.

Username


The username that is allowed to connect to the database.

Example: Snapuser 

Default value: None.

Password

Password used to connect to the data source. Password will be used as the default password when retrieving connections. The password must be valid in order to set up the data source.

Example: Snapuser 

Default value: None.

JDBC jars

List of JDBC JAR files to be loaded. A different driver binary for a driver must have a different name, the same name can not be reused for a different driver. If this property is left blank, a default JDBC driver will be loaded.

Enter the following JDBC jars to configure the Hive Database account for the respective cluster.

For HDP

  • Hive-jdbc-2.0.0.2.3.5.0-81-standalone.jar

  • Zookeeper-3.4.6.jar (Use this for setting up Hive with Zookeeper)

For CDH

  • hive_metastore.jar
  • hive_service.jar
  • hiveJDBC4.jar
  • libfb303-0.9.0.jar
  • libthrift-0.9.0.jar
  • TCLIServiceClient.jar
  • Zookeeper-3.3.6.jar (Use this for setting up Hive with Zookeeper)

Default value: None

  • The JDBC driver can be uploaded through Designer or Manager and it is stored on a per-project basis. That is, only users with access to that project will see JDBC drivers uploaded. To provide access to all users of your org, place the driver in the /shared project.
  • See Advanced Configurations: Configuring Hive with Kerberos section below for a list of JAR files to be uploaded when configuring Hive with Kerberos.

JDBC Driver Class

Required. The JDBC Driver class name. 

For HDP Clusters

Enter the following value: org.apache.hive.jdbc.HiveDriver

For CDH Clusters

Enter the following value: com.cloudera.hive.jdbc4.HS2Driver

Default value: None.

Advanced Properties


           Auto commit

If selected, then batches are immediately committed after they execute. Therefore, only the current executing batch will be rolled back if the Snap fails.

If not selected, then a transaction is started for the Snap run and committed upon run success. The transaction will be rolled back if the Snap fails.

Default value: Selected


For a DB Execute Snap, assume that a stream of documents enter the input view of the Snap and the SQL statement property has JSON paths in the WHERE clause. If the number of documents are large, the Snap executes in more than one batches rather than executing one per each document. Each batch would contain a certain number of WHERE clause values. If Auto commit is turned on, a failure would only roll back the records in the current batch. If Auto commit is turned off, the entire operation would be rolled back. For a single execute statement (with no input view), the setting has no practical effect.

Batch size


Required. Number of statements to execute at a time.
Using a large batch size could use up the JDBC placeholder limit of 2100.

Example: 10

Default value: 50

Fetch size


Required. Number of rows to fetch at a time when executing a query.
Large values could cause the server to run out of memory.

Example: 100

Default value: 100

Max pool size


Required. Maximum number of idle connections a pool will maintain at a time.

Example: 10

Default value: 50

Max life time

Required. Maximum lifetime of a connection in the pool. Ensure that the value you enter is a few seconds shorter than any database or infrastructure-imposed connection time limit. A value of 0 indicates an infinite lifetime, subject to the Idle Timeout value. An in-use connection is never retired. Connections are removed only after they are closed.

Default value: 30

Idle Timeout

Required. The maximum amount of time a connection is allowed to sit idle in the pool. A value of 0 indicates that idle connections are never removed from the pool.

Default value: 5

Checkout timeout

Required. Number of milliseconds to wait for a connection to be available in the pool. Zero waits forever. After set time, then an exception will be thrown and the pipeline will fail.

Example: 10000

Default value: 10000

Url Properties

Properties to use in JDBC Url. These properties will need to be configured when setting up SSL connection. See Advanced Configurations: Configuring Hive with SSL section below for details.

Example: maxAllowedPacket | 1000

Default value: None.

Hadoop properties

Authentication method

Required. Authentication method to use when connecting to the Hadoop service.  

  • None: Allows connection even without the Username and Password
  • Kerberos: Allows connection with Kerberos details such as Client Principal, Keytab file, and Service principal
  • User ID: Allows connection with Username only
  • User ID and Password: Allows connection with Username and Password
  • User ID and Password with SSL: Allows SSL connections with Username and Password. Ensure that you have installed SSL certificates in the JCC node. 

Default value: None

Use Zookeeper 


Specifies if Zookeeper be used to locate the Hadoop service instead of a specific hostname.

If the checkbox is selected, use Zookeeper to resolve the location of the database instead of using the hostname field in the standard block.

Default value: Not selected

When using Zookeeper in combination with a Hive account, add the Zookeeper JAR package file on the Groundplex associated with that Hive account. The version of Zookeeper on the Groundplex should be the same as the version your Hive account uses.

For HDP users, in addition to the zookeeper.jar package, you might also require the curator-client-X.X.X.jar and curator-framework-X.X.X.jar package files on the Groundplex.

Zookeeper URL

If you intend to use Zookeeper, then you must provide the following details:

You must provide the URL of the Zookeeper service. Zookeeper URL formats are different for CDH and HDP.

Default value: None

This is NOT the URL for the Hadoop service being sought.

Hive properties

JDBC Subprotocol


Conditional. This is required when the Authentication method is Kerberos. JDBC Subprotocol to be used. The options available are Hive and Impala.

Default value: Hive

Kerberos properties

Configuration information required for the Kerberos authentication. These properties must be configured if you select Kerberos in the Authentication method property.

Client Principal

Used to authenticate to Kerberos KDC (Kerberos Key Distribution Center - Network service used by the clients and servers for authentication). 

Default value: None.

Keytab file

Keytab file(file used to store encryption keys) used to authenticate to Kerberos KDC.

Default value: None.

Service principal


Principal used by an instance of a service.

Examples: 

  • If you are connecting to a specific server:
    hive/host@REALM or impala/host@REALM
  • If you are connecting(more common for the Snap) to any compliant host (see Use Zookeeper property's description) in which case the principal is:
    'hive/_HOST@REALM' or 'impala/_HOST@REALM'. 

Default value: None.

Generic Hive Database Account

 Generic Hive Database Account

Account Settings

Label

Required. User provided label for the account instance.

Account properties


Username


Username that is allowed to connect to the database. Username will be used as the default username when retrieving connections. The username must be valid in order to set up the data source.


Example: Snapuser 

Default value: None.

Password

Password used to connect to the data source. Password will be used as the default password when retrieving connections. The password must be valid in order to set up the data source.


Example: Snapuser 

Default value: None.

JDBC URL


The URL of the JDBC database.

Example: jdbc:hive://hostname/dbname;sasl.qop=auth-int

Default value: None.

JDBC jars

List of JDBC JAR files to be loaded. A different driver binary for a driver must have a different name, the same name can not be reused for a different driver. If this property is left blank, a default JDBC driver will be loaded.

Enter the following JDBC jars to configure the Generic Hive Database account for the concerned cluster.

For HDP

  • Hive-jdbc-2.0.0.2.3.5.0-81-standalone.jar

  • Zookeeper-3.4.6.jar (Use this for setting up Hive with Zookeeper)

For CDH

  • hive_metastore.jar
  • hive_service.jar
  • hiveJDBC4.jar
  • libfb303-0.9.0.jar
  • libthrift-0.9.0.jar
  • TCLIServiceClient.jar
  • Zookeeper-3.3.6.jar (Use this for setting up Hive with Zookeeper)

Default value: None

  • The JDBC driver can be uploaded through Designer or Manager and it is stored on a per-project basis. That is, only users with access to that project will see JDBC drivers uploaded. To provide access to all users of your org, place the driver in the /shared project.
  • See Advanced Configurations: Configuring Hive with Kerberos section below for a list of JAR files to be uploaded when configuring Hive with Kerberos.

JDBC Driver Class

Required. The JDBC Driver class name. 

For HDP Clusters

Enter the following value: org.apache.hive.jdbc.HiveDriver

For CDH Clusters

Enter the following value: com.cloudera.hive.jdbc4.HS2Driver

Default value: None.

Advanced Properties


           Auto commit

If true (selected), then batches are immediately committed after they execute. Therefore, only the current executing batch will be rolled back if the Snap fails.

If false (not selected), then a transaction is started for the Snap run and committed upon run success. The transaction will be rolled back if the Snap fails.

Default value: Selected


For a DB Execute Snap, assume that a stream of documents enter the input view of the Snap and the SQL statement property has JSON paths in the WHERE clause. If the number of documents are large, the Snap executes in more than one batches rather than executing one per each document. Each batch would contain a certain number of WHERE clause values. If Auto commit is turned on, a failure would only roll back the records in the current batch. If Auto commit is turned off, the entire operation would be rolled back. For a single execute statement (with no input view), the setting has no practical effect.

Batch size


Required. Number of statements to execute at a time.
Using a large batch size could use up the JDBC placeholder limit of 2100.

Example: 10

Default value: 50

Fetch size


Required. Number of rows to fetch at a time when executing a query.
Large values could cause the server to run out of memory.

Example: 100

Default value: 100

Max pool size


Required. Maximum number of idle connections a pool will maintain at a time.

Example: 10

Default value: 50

Max idle time

Required. Minutes a connection can exist in the pool before it is destroyed.

Example: 30

Default value: 30

Idle connection Test period

Required. Number of minutes for a connection to remain idle before a test query is run. This helps keep database connections from timing out.
Default value: 5

Checkout timeout

Required. Number of milliseconds to wait for a connection to be available in the pool. Zero waits forever. After set time, then an exception will be thrown and the pipeline will fail.

Example: 10000

Default value: 10000

Url Properties

Properties to use in JDBC Url. These properties will need to be configured when setting up SSL connection. See Advanced Configurations: Configuring Hive with SSL section below for details.

Example: maxAllowedPacket | 1000

Default value: None.

Hadoop Properties

Authentication method

Required. Authentication method to use when connecting to the Hadoop service.  

  • None: Allows connection even without the Username and Password
  • Kerberos: Allows connection with Kerberos details such as Client Principal, Keytab file, and Service principal
  • User ID: Allows connection with Username only
  • User ID and Password: Allows connection with Username and Password

Default value: None

Use Zookeeper 


Specifies if Zookeeper be used to locate the Hadoop service instead of a specific hostname.

If the checkbox is selected, use Zookeeper to resolve the location of the database instead of using the hostname field in the standard block.

Default value: Not selected

Zookeeper Versions

When using Zookeeper in combination with a Hive account, add the Zookeeper JAR package file on the Groundplex associated with that Hive account. The version of Zookeeper on the Groundplex should be the same as the version your Hive account uses.

For HDP users, in addition to the zookeeper.jar package, you might also require the curator-client-X.X.X.jar and curator-framework-X.X.X.jar package files on the Groundplex.

Zoo Keeper URL

If you intend to use Zookeeper, then you must provide the following details:

You must provide the URL of the Zookeeper service. Zookeeper URL formats are different for CDH and HDP.

Default value: None

This is NOT the URL for the Hadoop service being sought.

Hive properties

JDBC Subprotocol


Conditional. This is required when the Authentication method is Kerberos. JDBC Subprotocol to be used. The options available are Hive and Impala.

Default value: Hive

Kerberos properties

Configuration information required for the Kerberos authentication. These properties must be configured if you select Kerberos in the Authentication method property.

Client Principal

Used to authenticate to Kerberos KDC (Kerberos Key Distribution Center - Network service used by the clients and servers for authentication). 

Default value: None.

Keytab file

Keytab file(file used to store encryption keys) used to authenticate to Kerberos KDC.

Default value: None.

Service principal


Principal used by an instance of a service.

Examples: 

  • If you are connecting to a specific server:
    hive/host@REALM or impala/host@REALM
  • If you are connecting(more common for the Snap) to any compliant host (see Use Zookeeper property's description) in which case the principal is:
    'hive/_HOST@REALM' or 'impala/_HOST@REALM'. 

Default value: None.

Additional Configurations

Configuring Hive with SSL

The following URL properties have to be configured:

URL Property NameURL Property Value
sslRequired. Binary value to denote that SSL is enabled. This value must always be 1.
sslTrustStoreRequired. The path of the SSL trust store in the SLDB.
sslTrustStorePasswordRequired. Password configured for the SSL trust store.
AllowSelfSignedCertsBinary value to denote that server is allowed to use self-signed SSL certificates by the driver. 
CAIssuedCertNamesMismatchBinary value to denote that the CA issued SSL certificate's name is required by the driver to match the host name of the Hive server. 


The above list is specific to Hive with or without Kerberos enabled. With Kerberos enabled, the properties such as Client Principal, Key tab file and Service principal have to be additionally provided.

Limitations or known issues

  • "Method not supported" error while validating Apache Hive JDBC v1.2.1 or earlier
    The Hive Snap Pack does not validate with Apache Hive JDBC v1.2.1 jars or earlier because of a defect in Hive. HDP 2.6.3 and HDP 2.6.1 run on Apache Hive JDBC v1.2.1 jars.
    To validate Snaps that must work with HDP 2.6.3 and HDP 2.6.1, use JDBC v2.0.0 jars.

Testing Environment

  • Hive Version: Hive 1.1.0, Hive 1.2.0

  • Hive with Kerberos works only on Hive JDBC4 driver 2.5.12 and above 

  • Hive with Kerberos is tested on Hadooplex only

  • Hive with SSL works only on Hive JDBC4 driver 2.5.12 and above.

  • Cluster Versions: CDH 5.16.1, CDH 5.10, HDP 2.6.1, HDP 2.6.3 

Account History

4.17 (

  • Added a new authentication method, User ID and Password with SSL, for Hive SSL Accounts which allows SSL connections for valid user name and password credentials.

4.16 (snapsmrc508)

  • No updates made. Automatic rebuild with a platform release.

4.15 (snapsmrc500)

  • No updates made. Automatic rebuild with a platform release.

4.14 (snapsmrc490)

  • Added a new account type: Generic Hive Database Account, this enables connecting to different types of clusters using JDBC URL.

4.8.0 

  • Info tab added to accounts.

  • Database accounts now invalidate connection pools if account properties are modified and login attempts fail.

4.7.0

  • Enabled the account with Kerberos authentication.

  • No labels

Have feedback? Email documentation@snaplogic.com | Ask a question in the SnapLogic Community
© 2017-2024 SnapLogic, Inc.