- 530 Views
- 2 replies
- 0 kudos
Why driver memory is capped
Hi teamWe are using a job cluster to run spark with MERGE. Somehow it needs a lot driver memory. We allocate 128G+16core node for driver, and specify spark.driver.memory=96000m. We can see it is 96000m from env table of spark UI. The config is like:"...
- 530 Views
- 2 replies
- 0 kudos
- 0 kudos
Thanks for response. We are doubt why driver memory cannot be fully used (only 48G out of 128G is used for driver). Is this related with repartition?
- 0 kudos
- 954 Views
- 5 replies
- 1 kudos
Resolved! Unable to read Unity Catalog schema
Recently bumped into this (first-time) error, without a clear message as to cause.Insights welcomed.Error
- 954 Views
- 5 replies
- 1 kudos
- 1 kudos
@DivyaPasumarthi issue still persists, but found a workaround.Go to SQL Editor module, expand Catalog panel on the left, highlight desired table, right-click > Open in Catalog Explorer.
- 1 kudos
- 1761 Views
- 6 replies
- 0 kudos
Resolved! SELECT from VIEW to CREATE a table or view
Hi; I'm new to Databricks, so apologies if this is a dumb question.I have a notebook with SQL cells that are selecting data from various Delta tables into temporary views. Then I have a query that joins up the data from these temporary views.I'd lik...
- 1761 Views
- 6 replies
- 0 kudos
- 0 kudos
Thanks, FelixIvy. Just to clarify, the reason you can't use temporary views to load a materialized view is because materialized views (like regular views) must be created using a single query that is saved as part of the view definition. So the sol...
- 0 kudos
- 383 Views
- 1 replies
- 1 kudos
OAuth U2M AWS Token Failure
I am attempting to generate a manual OAuth token using the instructions for AWS. When attempting to generate the account level authentication code I run into a localhost error:I have confirmed that all variables and urls are correct and that I am log...
- 383 Views
- 1 replies
- 1 kudos
- 1 kudos
After investigating further, the localhost issue was because I was already logged in and did not need to login again. The returned URL contained the authorization code. I was able to authenticate and run account level API requests with the generated ...
- 1 kudos
- 3489 Views
- 5 replies
- 2 kudos
Resolved! 404 on GET Billing usage data (API)
I'm trying to get my billing usage data from Databricks API (documentation: https://docs.databricks.com/api/gcp/account/billableusage/download) but I keep getting an 404 error.Code:import requestsimport jsontoken = dbutils.notebook.entry_point.getDbu...
- 3489 Views
- 5 replies
- 2 kudos
- 2 kudos
Bumping this to see if there is a solution. Per Databricks basic authentication is no longer allowed. I am unable to authenticate to get access to this endpoint (401 error). Does anyone have a solution to querying this endpoint?
- 2 kudos
- 645 Views
- 4 replies
- 2 kudos
Change datetime format from one to another without changing datatype in databricks sql
Change datetime"2002-01-01T00:00:00.000" to 'MM/dd/yyyy HH:mm:ss' format without changing datatype/ having it in datetime data type
- 645 Views
- 4 replies
- 2 kudos
- 2 kudos
Hi @Mister-Dinky ,As @szymon_dybczak if you have a datetime, then you have a datetime.What you see is just a format of defined in the Databricks UI. Other applications may display it differently depending on the defaults, regional formats etc.If you ...
- 2 kudos
- 971 Views
- 3 replies
- 1 kudos
Autoloader configuration for multiple tables from the same directory
I would like to get a recommendation on how to structure ingestion of lots of tables of data. I am using autoloader currently with the directory searching mode.I have concerns about performance in the future and have a requirement to ensure that data...
- 971 Views
- 3 replies
- 1 kudos
- 1 kudos
there is an easier way to see what has been processed:SELECT * FROM cloud_files_state('path/to/checkpoint'https://docs.databricks.com/en/ingestion/cloud-object-storage/auto-loader/production.html
- 1 kudos
- 463 Views
- 2 replies
- 0 kudos
Autoloader not ingesting all file data into Delta Table from Azure Blob Container
I have done the following, ie. crate a Delta Table where I plan to load the Azure Blob Container files that are .json.gz files: df = spark.read.option("multiline", "true").json(f"{container_location}/*.json.gz") DeltaTable.create(spark) \ .addCol...
- 463 Views
- 2 replies
- 0 kudos
- 0 kudos
If it's streaming data, space it out with 10 seconds trigger .trigger(processingTime="10 seconds") Do all the JSON files have the same schema? As your table creation is dynamic (df.schema), if all JSON doesn't have the same schema they may be skipp...
- 0 kudos
- 486 Views
- 0 replies
- 0 kudos
Adding a message to azure service bus
I am trying to send a message to a service bus in azure. But I get following error:ServiceBusError: Handler failed: DefaultAzureCredential failed to retrieve a token from the included credentials.This is the line that fails: credential = DefaultAzure...
- 486 Views
- 0 replies
- 0 kudos
- 398 Views
- 1 replies
- 0 kudos
How to set file size for MERGE
Hi team,I use MERGE to merge source to target table. Source is incremental reading with checkpoint on delta table. Target is delta table without any partition. If the table is empty, with spark.databricks.delta.optimizeWrite.enabled it can create fil...
- 398 Views
- 1 replies
- 0 kudos
- 0 kudos
Hi @Brad ,There are a couple of considerations here, the main being your runtime version and also whether you are using unit catalog.Check this document:https://docs.databricks.com/en/delta/tune-file-size.html
- 0 kudos
- 445 Views
- 3 replies
- 0 kudos
Will MERGE incur a lot driver memory
Hi team,We have a job to run MERGE on a target table with around 220 million rows. We found it needs a lot driver memory (just for MERGE itself). From the job metrics we can see the MERGE needs at least 46GB memory. Is there some special thing to mak...
- 445 Views
- 3 replies
- 0 kudos
- 0 kudos
Hi @Brad ,Could you try to apply very standard optimization practices and check the outcome:1. If your runtime is greater equal 15.2, could you implement liquid clustering on the source and target tables using JOIN columns?ALTER TABLE <table_name> CL...
- 0 kudos
- 654 Views
- 1 replies
- 2 kudos
Resolved! Trigger a workflow from a different databricks environment
Hello everyone,In the company I work we have a lot of different databricks environments and now we're in need of deeper integration of processes from environment's X and Y. There's a workflow in Y that runs a process that when finished we would like ...
- 654 Views
- 1 replies
- 2 kudos
- 2 kudos
Hi @hcord ,You can use REST API in the last task to trigger a workflow in different workspace
- 2 kudos
- 893 Views
- 1 replies
- 0 kudos
Loading data from spark dataframe directly to Sharepoint
Hi guys!I am trying to load data directly from PySpark dataframe to Sharepoint folder and I cannot find a solution regarding it.I wanted to implement workaround using volumes and logic apps, but there are few issues. I need to partition df in a few f...
- 893 Views
- 1 replies
- 0 kudos
- 0 kudos
One approach could involve using Azure Data Lake as an intermediary. You can partition your PySpark DataFrames and load them into Azure Data Lake, which is optimized for large-scale data storage and integrates well with PySpark. Once the data is in A...
- 0 kudos
- 3218 Views
- 4 replies
- 2 kudos
Resolved! Remove Duplicate rows in tables
HelloI've seen posts that show how to remove duplicates, something like this:MERGE into [deltatable] as targetUSING ( select *, ROW_NUMBER() OVER (Partition By [primary keys] Order By [date] desc) as rn from [deltatable] qualify rn> 1 ) as sourceON ...
- 3218 Views
- 4 replies
- 2 kudos
- 2 kudos
Hi @dpc ,if you like using SQL:1. Test data:# Sample data data = [("1", "A"), ("1", "A"), ("2", "B"), ("2", "B"), ("3", "C")] # Create DataFrame df = spark.createDataFrame(data, ["id", "value"]) # Write to Delta table df.write.format("delta").mode(...
- 2 kudos
- 380 Views
- 1 replies
- 0 kudos
First time to see "Databricks is experiencing heavy load" message. What does it mean really?
Hi, I just went to run a Databricks pyspark notebook and saw this message:This is a notebook I've run before but never saw this. Is it referring to my cluster? The Databricks infrastructure? My notebook ran normally, just wondering though. Google sea...
- 380 Views
- 1 replies
- 0 kudos
- 0 kudos
never saw that message, but my guess it is not your cluster but the Databricks platform in your region.status.databricks.com perhaps has some info.
- 0 kudos