Ingesting RDBMS Data as New Tables Arrive in Hive
Learn about loading all relational database data into new Hive tables automatically, or — dare we say it — autoMAGICally.
Join the DZone community and get the full member experience.Join For Free
Let's say that a company wants to know when new tables are added to a JDBC source (say, an RDBMS). Using the
ListDatabaseTables processor, we can get a list of
TABLE s, and also views, system tables, and other database objects, but for our purposes, we want tables with data. I have used the
ngdbc.jar from SAP HANA to connect and query tables with ease.
For today's example, I am connecting to MySQL, as I have a MySQL database available for use and modification.
mysql -u root -p test < person.sql CREATE USER 'nifi'@'%' IDENTIFIED BY 'reallylongDifficultPassDF&^D&F^Dwird'; GRANT ALL PRIVILEGES ON *.* TO 'nifi'@'%' WITH GRANT OPTION; COMMIT; mysql> show tables; +----------------+ | Tables_in_test | +----------------+ | MOCK_DATA | | mock2 | | personReg | | visitor | +----------------+ 4 rows in set (0.00 sec)
I created a user to use for my JDBC Connection Pool in NiFi to read the metadata and data.
These table names will show up in NiFi in the
ListDatabaseTables: Let's get a list of all the tables in MySQL for the database we have chosen.
After it starts running, you can check its state, see what tables were ingested, and see the most recent timestamp (Value).
We will get back what catalog we read from, how many tables there are, and each table name and it's full name.
HDF NiFi supports generic JDBC drivers and specific coding for Oracle, MS SQL Server 2008, and MS SQL Server 2012+.
GenerateTableFetch using the table name returned from the list returned by the database control.
extract text to get the SQL statement created by
generate table fetch.
We add a new attribute,
Execute SQL with that
Convert AVRO files produced by
ExecuteSQL into performant Apache ORC files:
PutHDFS to store these ORC files in Hadoop.
I added the table name as part of the directory structure, so a new directory is created for each transferred table. Now, we have dynamic HDFS directory structure creation.
Replace the text to build a SQL statement that will generate an external Hive table on our new ORC directory.
PutHiveQL to execute the statement that was just dynamically created for our new table.
We now have instantly queryable Hadoop data available to Hive, SparkSQL, Zeppelin, ODBC, JDBC, and a ton of BI tools and interfaces.
Finally, we can look at the data that we have ingested from MySQL into new Hive tables.
That was easy! The best part is that as new tables are added to MySQL, they will be auto-ingested into HDFS and Hive tables.
Use Hive merge capabilities to update changed data. We can also ingest to Phoenix/HBase and use the upsert DML.
Test with other databases. Tested with MySQL.
Quick tip (HANA): In NiFi, refer to tables with their full name in quotes:
Published at DZone with permission of Tim Spann, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.