I am using HIVE to load data into different partitions.
I am creating a table
CREATE TABLE X IF NOT EXISTS ... USING PARQUET PARTITIONED BY (Year, Month,Day)
LOCATION '...'
Afterwards I am performing a full load:
INSERT OVERWRITE TABLE ... PARTITION (Year, Month, Day)
SELECT ... FROM Y
Show partitions shows me all partitions correctly.
and after the full load, I just want to reload always the current year dynamically:
INSERT OVERWRITE TABLE ... PARTITION (Year, Month, Day)
SELECT ... FROM Y WHERE Year = YEAR(CURRENT_DATE())
The issue I have is that HIVE deletes all PREVIOUS partitions i.e. 2017, 2018 and just 2019 persists. I was supposed that HIVE ONLY overwrites the partition for 2019 but not all.
I suppose I do something wrong - any idea is welcome.
Try using "Insert into table" instead of "Insert overwrite table". It should solve your problem. :)
Okay, I got the solution once I studied the official databricks guide more carefully.
Here is the answer:
The semantics are different based on the type of the target table.
Hive SerDe tables: INSERT OVERWRITE doesn’t delete partitions ahead, and only overwrite those partitions that have data written into it at runtime. This matches Apache Hive semantics. For Hive SerDe tables, Spark SQL respects the Hive-related configuration, including hive.exec.dynamic.partition and hive.exec.dynamic.partition.mode.
Native data source tables: INSERT OVERWRITE first deletes all the partitions that match the partition specification (e.g., PARTITION(a=1, b)) and then inserts all the remaining values. Since Databricks Runtime 3.2, the behavior of native data source tables can be changed to be consistent with Hive SerDe tables by changing the session-specific configuration spark.sql.sources.partitionOverwriteMode to DYNAMIC. The default mode is STATIC.
Related
Is it even possible to add a partition to an existing table in Athena that currently is without partitions? If so, please also write syntax for doing so in the answer.
For example:
ALTER TABLE table1 ADD PARTITION (ourDateStringCol = '2021-01-01')
The above command will give the following error:
FAILED: SemanticException table is not partitioned but partition spec exists
Note: I have done a web-search, and variants exist for SQL server, or adding a partition to an already partitioned table. However, I personally could not find a case where one could successfully add a partition to an existing non-partitioned table.
This is extremely similar to:
SemanticException adding partiton Hive table
However, the answer given there requires re-creating the table.
I want to do so without re-creating the table.
Partitions in Athena are based on folder structure in S3. Unlike standard RDBMS that are loading the data into their disks or memory, Athena is based on scanning data in S3. This is how you enjoy the scale and low cost of the service.
What it means is that you have to have your data in different folders in a meaningful structure such as year=2019, year=2020, and make sure that the data for each year is all and only in that folder.
The simple solution is to run a CREATE TABLE AS SELECT (CTAS) query that will copy the data and create a new table that can be optimized for your analytical queries. You can choose the table format (Parquet, for example), the compression (SNAPPY, for example), and also the partition schema (per year, for example).
I have a use case where I have to replace the data in a partition of a table in BigQuery every time 15 mins. Are there any functions available in Bigquery similar to partition exchange in Bigquery or any provision to truncate data of a partition.
Regarding your requirement to load new data every fifteen minutes into a partitioned table you could use Data Manipulation Language (DML).
In order to update rows in a partitioned table you could use the UPDATE statement as stated in the documentation.
Also, in case that you wanted to overwrite partitions you could also use a load job using the CLI as stated here. Using --noreplace or --replace you can specify if you want to append or truncate the given partition.
Note: this is nearly a duplicate of this question with the distinction that in this case, the source table is date partitioned and the destination table does not yet exist. Also, the accepted solution to that question didn't work in this case.
I'm trying to copy a single day's worth of data from one date partitioned table into a new date partitoined table that I have not yet created. My hope is that BigQuery would simply create the date-partitioned destination table for me like it usually does for the non-date-partitioned case.
Using BigQuery CLI, here's my command:
bq cp mydataset.sourcetable\$20161231 mydataset.desttable\$20161231
Here's the output of that command:
BigQuery error in cp operation: Error processing job
'myproject:bqjob_bqjobid': Partitioning specification must be provided
in order to create partitioned table
I've tried doing something similar using the python SDK: running a select command on a date partitioned table (which selects data from only one date partition) and saving the results into a new destination table (which I hope would also be date partitioned). The job fails with the same error:
{u'message': u'Partitioning specification must be provided in order to
create partitioned table', u'reason': u'invalid'}
Clearly I need to add a partitioning specification, but I couldn't find any documentation on how to do so.
You need to create the partitioned destination table first (as per the docs):
If you want to copy a partitioned table into another partitioned
table, the partition specifications for the source and destination
tables must match.
So, just create the destination partitioned table before you start copying. If you can't be bothered specifying the schema, you can create the destination partitioned table like so:
bq mk --time_partitioning_type=DAY mydataset.temps
Then, use a query instead of a copy to write to the destination table. The schema will be copied with it:
bq query --allow_large_results --replace --destination_table 'mydataset.temps$20160101''SELECT * from `source`'
Can anyone please suggest how to create partition table in Big Query ?.
Example: Suppose I have one log data in google storage for the year of 2016. I stored all data in one bucket partitioned by year , month and date wise. Here I want create table with partitioned by date.
Thanks in Advance
Documentation for partitioned tables is here:
https://cloud.google.com/bigquery/docs/creating-partitioned-tables
In this case, you'd create a partitioned table and populate the partitions with the data. You can run a query job that reads from GCS (and filters data for the specific date) and writes to the corresponding partition of a table. For example, to load data for May 1st, 2016 -- you'd specify the destination_table as table$20160501.
Currently, you'll have to run several query jobs to achieve this process. Please note that you'll be charged for each query job based on bytes processed.
Please see this post for some more details:
Migrating from non-partitioned to Partitioned tables
There are two options:
Option 1
You can load each daily file into separate respective table with name as YourLogs_YYYYMMDD
See details on how to Load Data from Cloud Storage
After tables created, you can access them either using Table wildcard functions (Legacy SQL) or using Wildcard Table (Standar SQL). See also Querying Multiple Tables Using a Wildcard Table for more examples
Option 2
You can create Date-Partitioned Table (just one table - YourLogs) - but you still will need to load each daily file into respective partition - see Creating and Updating Date-Partitioned Tables
After table is loaded you can easily Query Date-Partitioned Tables
Having partitions for an External Table is not allowed as for now. There is a Feature Request for it:
https://issuetracker.google.com/issues/62993684
(please vote for it if you're interested in it!)
Google says that they are considering it.
I have a lot of data in a Parquet based Hive table (Hive version 0.10). I have to add a few new columns to the table. I want the new columns to have data going forward. If the value is NULL for already loaded data, that is fine with me.
If I add the new columns and not update the old Parquet files, it gives an error and it looks strange as I am adding String columns only.
Error getting row data with exception java.lang.UnsupportedOperationException: Cannot inspect java.util.ArrayList
Can you please tell me how to add new fields to Parquet Hive without affecting the already existing data in the table ?
I use Hive version 0.10.
Thanks.
1)
Hive starting with version 0.13 has parquet schema evoultion built in.
https://issues.apache.org/jira/browse/HIVE-6456
https://github.com/Parquet/parquet-mr/pull/297
ps. Notice that out-of-the-box support for schema evolution might take a toll on performance. For example, Spark has a knob to turn parquet schema evolution on and off. After one of the recent Spark releases, it's now off by default because of performance hit (epscially when there are a lot of parquet files). Not sure if Hive 0.13+ has such a setting too.
2)
Also wanted to suggest to try creating views in Hive on top of such parquet tables where you expect often schema changes, and use views everywhere but not tables directly.
For example, if you have two tables - A and B with compatible schemas, but table B has two more columns, you could workaround this by
CREATE VIEW view_1 AS
SELECT col1,col2,col3,null as col4,null as col5 FROM tableA
UNION ALL
SELECT col1,col2,col3,col4,col5 FROM tableB
;
So you don't actually have to recreate any tables like #miljanm has suggested, you can just recreate the view. It'll help with the agility of your projects.
Create a new table with the two new columns. Insert data by issuing:
insert into new_table select old_table.col1, old_table.col2,...,null,null from old_table;
The last two nulls are for the two new columns. That's it.
If you have too many columns, it may be easier for you to write a program that reads the old files and writes the new ones.
Hive 0.10 does not have support for schema evolution in parquet as far as I know. Hive 0.13 does have it, so you may try to upgrade hive.