I started going through DELTA LAKE file format, is hive capable of reading data from this newly introduced delta file format? If so could you please let me know the serde you were using.
Hive support is available with Delta Lake file format. First, step is to add the jars from https://github.com/delta-io/connectors, in our hive path. And then create a table using following format.
CREATE EXTERNAL TABLE test.dl_attempts_stream
(
...
)
STORED BY 'io.delta.hive.DeltaStorageHandler'
LOCATION
Delta Format picks up partition by default, so no need to mention partition while creating a table.
NOTE: If data is being inserted via a Spark job, please provide hive-site.xml, and enableHiveSupport in Spark Job, to create Delta Lake table in Hive.
Related
We are using spark for reading/writing data in delta format stored in HDFS (Databricks Delta table version 0.5.0).
We would like to utilize the power of Hive to interact with the delta tables.
How can we register an existing data in delta format from a path on HDFS to Hive?
Please note that currently we are running spark (2.4.0) on cloudera platform (CDH 6.3.3)
The only way I can do this so far is by registering it as an unmanaged table. The most significant difference, as far as I can tell, is that if you drop an unmanaged table, it does not drop the underlying data.
I am trying to understand the spark DataFrame API method called saveAsTable.
I have following question
If I simply write a dataframe using saveAsTable API
df7.write.saveAsTable("t1"), (assuming t1 did not exist earlier), will the newly created table be a hive table which can be read outside spark using Hive QL ?
Does spark also create some non-hive table (which are created using saveAsTable API but can not be read outside spark using HiveQL)?
How can check if a table is Hive Table or Non-Hive table ?
(I am new to big data processing, so pardon me if question is not phrased properly)
Yes. Newly created table will be hive table and can be queried from Hive CLI(Only if the DataFrame is created from single input HDFS path i.e. from non-partitioned single input HDFS path).
Below is the documentation comment in DataFrameWriter.scala class. Documentation link
When the DataFrame is created from a non-partitioned
HadoopFsRelation with a single input path, and the data source
provider can be mapped to an existing Hive builtin SerDe (i.e. ORC and
Parquet), the table is persisted in a Hive compatible format, which
means other systems like Hive will be able to read this table.
Otherwise, the table is persisted in a Spark SQL specific format.
Yes, you can do. You table can be partitioned by a column, but can not use bucketing (its a problem between spark and hive).
I have tried to access the Hive ORC Transactional table (which has underlying delta files on HDFS) using PySpark but I'm not able to read the transactional table through sparkContext/hiveContext.
/mydim/delta_0117202_0117202
/mydim/delta_0117203_0117203
Officially Spark not yet supported for Hive-ACID table, get a
full dump/incremental dump of acid table to regular hive orc/parquet partitioned table then read the data using spark.
There is a Open Jira saprk-15348 to add support for reading Hive ACID table.
If you run major compaction on Acid table(from hive) then spark able to read base_XXX directories only but not delta directories Spark-16996 addressed in this jira.
There are some workaround to read acid tables using SPARK-LLAP as mentioned in this link.
I think starting from HDP-3.X HiveWareHouseConnector is able to support to read HiveAcid tables.
I am planning to use SparkSQL (not pySpark) on top of data in Amazon S3. So I believe I need to create Hive external table and then can use SparkSQL. But S3 data is partitioned and want to have the partitions reflected in Hive external table also.
What is the best way to manage the hive table on a daily basis. Since
, everyday new partitions can be created or old partitions can be
overwritten and what to do , so as to keep the Hive external table
up-to-date?
Create a intermediate table and load to your hive table with insert overwrite partition on date.
we have 2 clusters one Map R and another our own. We want created new setup in our own hardware using the Map R data.
I have copied all the orc files from the Map R cluster and followed the same folder structure
Created a orc formatted table with location of #1
then executed this command "MSCK REPAIR TABLE <>"
above steps passed without error, but when i query the partitions then job fails with below error
java.lang.IllegalArgumentException: Buffer size too small. size = 262144 needed = 4958903
at org.apache.hadoop.hive.ql.io.orc.InStream$CompressedStream.readHeader(InStream.java:193)
at org.apache.hadoop.hive.ql.io.orc.InStream$CompressedStream.read(InStream.java:238)
Can some one tell me can we create HIVE ORC partition tables directly from the orc files?
My storage is Azure data lake.
According to your description, based on my understanding, I think you want to copy all orc files from a cluster to another and load these orc files as a hive table.
For doing it, please just try to follow the command below to create external table for loading orcfile data.
CREATE EXTERNAL TABLE IF NOT EXSISTS <table name> (<column_name column_type>, ...)
ROW FORMAT SERDE 'org.apache.hadoop.hive.ql.io.orc.OrcSerde'
STORED AS ORC
LOCATION '<orcfile path>'
If not aware of the columns list of an orc file, you can refer to the Hive manual ORC File Dump Utility to print the ORC file metadata in JSON format via hive --orcfiledump -j -p <location-of-orc-file-or-directory>.