I have a copy data activity for on-premise SQL Server as source and ADLS Gen2 as sink. There is a control table to pickup tableName, watermarkDateColumn and the watermarkDatetime to pull incremental data from the source database.
After data is pulled/loaded in sink, I want to get the max of the watermarkDateColumn in my dataset. Can it be obtained from #activity('copyActivity1').output?
I'm not allowed to use one extra lookup activity to query the source table for getting the max(watermarkDateColumn) in pipeline.
Copy activity only could be used for data transmission,not for any other aggregation feature. So #activity('copyActivity1').output won't help. Since you said you can't use lookup activity, i'm afraid your requirement is not available so far.
If you prefer not using additional activities, I suggest you using Data Flow Activity instead which is more flexible.There is built-in aggregation feature in the Data Flow Activity.
Related
Please help me with copying data from Google BigQuery to Azure Data Lake Storage Gen2 with Serverless SQL Pool.
I am using Azure Synapse's Copy data pipeline. The issue is I cannot figure out how to handle source table from the BigQuery with hierarchical schema. This result in missing columns and inaccurate datetime value at the sink.
The source is a Google BigQuery table, it is made of Google Cloud Billing export of a project's standard usage cost. The source table's schema is hierarchical with nested columns, such as service.id; service.description; sku.id; sku.description; Project.labels.key; Project.labels.value, etc.
When I click on Preview data from the Source tab of the Copy data pipeline, it only gives me the top of the column hierarchy, for example: It would only show the column name of [service] and with value of {\v":{"f":[{"v":"[service.id]"},{"v":"[service.descrpition]"}]}}
image description: Source with nested columns result in issues with Synapse Copy Data Pipline
I have tried to configure the Copy Pipline with the following:
Source Tab:
Use query - I think the solution lays in here, but I cannot figure out the syntax of selecting the proper columns. I watched a Youtube video from TechBrothersIT How to Pass Parameters to SQL query in Azure Data Factory - ADF Tutorial 2021, but still unable to do it.
Sink Tab:
1.Sink dataset in various format of csv, json and parquet - with csv and parquet getting similar result, and json format failed
2.Sink dataset to Azure SQL Database - failed because it is not supported with Serverless SQL Pool
3.Mapping Tab: note: edited on Jan22 with screenshot to show issue.
Tried with Import schemas, with Sink Tab copy behavior of None, Flatten Hierarchy and Preserve Hierarchy, but still unable to get source column to be recognized as Hierarchical. Unable to get the Collection reference nor the Advanced Editor configurations to show up. Ref: Screenshot of Source columns not detected as Hierarchical MS Doc on Schema and data type mapping in copy activity
I have also tried with the Data flow pipeline, but it does not support Google BigQueryData Flow Pipe Source do not support BigQuery yet
Here are the steps to reproduce / get to my situation:
Register Google cloud, setup billing export (of standard usage cost) to BigQuery.
At Azure Synapse Analytics, create a Linked service with user authentication. Please follow Data Tech's Youtube video
"Google BigQuery connection (or linked service) in Azure Synapse analytics"
At Azure Synapse Analytics, Integrate, click on the "+" sign -> Copy Data Tool
I believe the answer is at the Source tab with Query and Functions, please help me figure this out, or point me to the right direction.
Looking forward to your input. Thanks in advance!
ADF allows you to write the query in google bigquery source dataset. Therefore write the query to unnest the nested columns using unnest operator and then map it to the sink.
I tried to repro this with sample nested table.
img:1 nested table
img:2 sample data of nested table
Script to flatten the nested table:
select
user_id,
a.post_id,
a.creation_date
from `ds1.stackoverflow_nested`
cross join unnest(comments) a
img:3 flattened table.
Use this query in copy activity source dataset.
img:4 Source settings of copy activity.
Then take the sink dataset, do the mapping and execute the ADF pipeline.
Reference:
MS document on google bigquery as a source - ADF
GC document on unnest operator
When using a data flow in azure data factory to move data, I've noticed that the data (at the sink) is missing columns that contains NULL values. When using the copy activity to copy the same data, the columns are present in the sink with their NULL values.
Record after a copy activity:
Record after a data flow:
Source is parquet, sink is azure cosmos db. My goal is to avoid defining any schemas, as I simply want to copy all of the data "as is". I've used the "allow schema drift" option on the source and sink.
I would just use the copy activity, but it doesn't appear to have the ability to define a maximum speed (RU consumption) like the data flow does, so the copy activity ends up consuming all of the cosmos db's RUs very quickly (as described here)
EDIT:
sink data preview shows all columns
sink inspect tab shows all columns
Dataflows always skip writing JSON tags with NULLs. There is no workaround currently other than copy activity.
This is really not a good design or behavior on Microsoft's part because you can't Standardize in Cosmos weather to "Keep" or "Remove" null fields in your JSON.
Querying Cosmos
Where field1 = NULL is completely different than where NOT IS_DEFINED (field1) and will yield an entirely different result set.
And if your users don't know if the ADF developer used a Dataflow with a Sink vs a Copy Activity in a Pipeline then the may get erroneous results in a query. The only way to ensure you get all the data is to always use:
Where field1 = NULL or where NOT IS_DEFINED (field1)
Users should not have to depend on knowing what kind of ADF functionality was chosen for a specific JSON document in a Cosmos NoSQL collection to do a query. Plus you can't standardize that you will "Keep" null across all Cosmos documents or you will "Remove" nulls across all Cosmos documents. Unless you force everyone to use Pipelines only or Dataflows only. Depending on the complexity using Pipeline only is not always possible. But using Dataflow only is also not always needed.
I have two tables in snowflake named table1 and table2. Table1 is the source table which contains incremental data and table2 is the target table.
So my usecase is I have to take data from table1 and update the data into table2 but this process has to be done using Azure Data Factory.
I tried to create a data flow in ADF but it didn't allowed me to connect with the snowflake directly as it is not in the supported sources list. The native snowflake connector only supports the Copy Data Activity. So as a work around I first created a copy activity which copy the data from snowflake to azure blob. Then used the Azure Blob as source for Data Flow to create my scd1 implementation and saved the output in csv files.
Now My question is how should I update the data in target table2. Because If I directly use the copy activity to copy the csv files into snowflake then it will result in the duplicate records at snowflake side. For instance lets say table2 contains a row
id,name,age,data
1234,kristopher,24,somedata
and table1 contains
id,name,age,data
1234,kristopher,24,some-new-data
So now I have table1 data in csv which has to be loaded in snowflake. If I am loading directly then the resultant looks something like this.
id,name,age,data
1234,kristopher,24,somedata
1234,kristopher,24,some-new-data
But I only need
1234,kristopher,24,some-new-data
Let me know if some more explanation is required. I am new to Azure Data Factory and Snowflake as well.
Thanks
As you have observed, the ADF Data Flows currently don't support Snowflake datasets as a source.
You could theoretically follow this design pattern but it seems like alot of work for the requirement you have described. An alternative would be to go down the Azure Function route, but again I would trade off the requirement vs. effort required.
If it didn't have to be in ADF, then a quick approach would be to use a Snowflake Task to schedule some SQL to manage the SCD behavior for you.
I hope this helps.
Best regards,
Dan.
you can put your login in a snowflake stored procedure, then execute your stored proc in ADF
I want to implement the scd2 in the snowflake tables. My source and target tables are present in snowflake only. The entire process has to be done using Azure Data Factory.
I went through the documentation given by azure for implementing the scd2 using data flows but when I tried to create a dataset for snowflake connection its showing as disabled.
Is there any way or any documentation where I can see the steps to create SCD2 in adf with snowflake tables.
Thanks
vipendra
SCD2 in ADF can be built and managed graphically via data flows. The Snowflake connector for ADF today does not work directly with data flows, yet. So for now, you will need to use the Copy Activity in an ADF pipeline and stage the dimension data in Blob or ADLS, then build your SCD2 logic in data flows using the staged data.
Your pipeline will look something like this:
[Copy Activity Snowflake-to-Blob] -> [Data Flow SCD2 logic Blob-to-Blob] -> [Copy Activity Blob-to-Snowkflake]
We are working on direct connectivity to Snowflake from data flows and hope to land that soon.
If your source and target tables are both in Snowflake, you could use Snowflake Streams to do this. There's a blog post covering this in more detail at https://community.snowflake.com/s/article/Building-a-Type-2-Slowly-Changing-Dimension-in-Snowflake-Using-Streams-and-Tasks-Part-1
However, in short, if you have a source table source, you can put a stream on it like so:
create or replace stream source_changes on table source;
This will capture all the changes that are made to the source table. You can then build a view on that stream that establishes how you want to feed those changes into the SCD table. (The blog post uses case statements to put start and end dates in on each row in the view).
From there, you can use a Snowflake Task to automate the process of loading from the stream into the SCD only when the Stream actually has changes.
I am accessing Databricks Delta tables from Azure Data Factory, which does not have a native connector to Databricks tables. So, as a workaround, I create the tables with the LOCATION keyword to store them in Azure Data Lake. Then, since I know the table file location, I just read the underlying Parquet files from Data Factory. This works fine.
But... what if there is cached information in the Delta transaction log that has not yet been written to disk? Say, an application updated a row in the table, and the disk does not yet reflect this fact. Then my read from Data Factory will be wrong.
So, two questions...
Could this happen? Are changes held in the log for a while before being written out?
Can I force a transaction log flush, so I know the disk copy is updated?
Azure Data Factory has built in delta lake support (this was not the case at the time the question was raised).
Delta is available as an inline dataset in a Azure Data Factory data flow activity. To get column metadata, click the Import schema button in the Projection tab. This will allow you to reference the column names and data types specified by the corpus (see also the docs here).
ADF supports Delta Lake format as of July 2020:
https://techcommunity.microsoft.com/t5/azure-data-factory/adf-adds-connectors-for-delta-lake-and-excel/ba-p/1515793
The Microsoft Azure Data Factory team is enabling .. and a data flow connector for data transformation using Delta Lake
Delta is currently available in ADF as a public preview in data flows as an inline dataset.