Prepared-to-go pattern information pipelines with Dataflow | by Netflix Expertise Weblog | Dec, 2022

by Jasmine Omeke, Obi-Ike Nwoke, Olek Gorajek

This submit is for all information practitioners, who’re fascinated by studying about bootstrapping, standardization and automation of batch information pipelines at Netflix.

Chances are you’ll bear in mind Dataflow from the submit we wrote final yr titled Information pipeline asset administration with Dataflow. That article was a deep dive into one of many extra technical elements of Dataflow and didn’t correctly introduce this device within the first place. This time we’ll attempt to give justice to the intro after which we are going to concentrate on one of many very first options Dataflow got here with. That function is named pattern workflows, however earlier than we begin in let’s have a fast have a look at Dataflow on the whole.



Dataflow is a command line utility constructed to enhance expertise and to streamline the information pipeline improvement at Netflix. Try this excessive degree Dataflow assist command output under:

$ dataflow --help
Utilization: dataflow [OPTIONS] COMMAND [ARGS]...

--docker-image TEXT Url of the docker picture to run in.
--run-in-docker Run dataflow in a docker container.
-v, --verbose Permits verbose mode.
--version Present the model and exit.
--help Present this message and exit.

migration Handle schema migration.
mock Generate or validate mock datasets.
venture Handle a Dataflow venture.
pattern Generate absolutely useful pattern workflows.

As you may see Dataflow CLI is split into 4 fundamental topic areas (or instructions). Probably the most generally used one is dataflow venture, which helps people in managing their information pipeline repositories by creation, testing, deployment and few different actions.

The dataflow migration command is a particular function, developed single handedly by Stephen Huenneke, to totally automate the communication and monitoring of a knowledge warehouse desk modifications. Because of the Netflix inner lineage system (constructed by Girish Lingappa) Dataflow migration can then provide help to establish downstream utilization of the desk in query. And at last it might probably provide help to craft a message to all of the house owners of those dependencies. After your migration has began Dataflow may even preserve monitor of its progress and provide help to talk with the downstream customers.

Dataflow mock command is one other standalone function. It permits you to create YAML formatted mock information recordsdata based mostly on chosen tables, columns and some rows of information from the Netflix information warehouse. Its fundamental function is to allow straightforward unit testing of your information pipelines, however it might probably technically be utilized in every other conditions as a readable information format for small information units.

All of the above instructions are very prone to be described in separate future weblog posts, however proper now let’s concentrate on the dataflow pattern command.

Dataflow pattern workflows is a set of templates anybody can use to bootstrap their information pipeline venture. And by “pattern” we imply “an instance”, like meals samples in your native grocery retailer. One of many fundamental causes this function exists is rather like with meals samples, to present you “a style” of the manufacturing high quality ETL code that you can encounter contained in the Netflix information ecosystem.

All of the code you get with the Dataflow pattern workflows is absolutely useful, adjusted to your atmosphere and remoted from different pattern workflows that others generated. This pipeline is protected to run the second it exhibits up in your listing. It would, not solely, construct a pleasant instance mixture desk and fill it up with actual information, however it can additionally current you with an entire set of really helpful parts:

  • clear DDL code,
  • correct desk metadata settings,
  • transformation job (in a language of selection) wrapped in an non-compulsory WAP (Write, Audit, Publish) sample,
  • pattern set of information audits for the generated information,
  • and a totally useful unit take a look at on your transformation logic.

And final, however not least, these pattern workflows are being examined constantly as a part of the Dataflow code change protocol, so you may make certain that what you get is working. That is one option to construct belief with our inner person base.

Subsequent, let’s take a look on the precise enterprise logic of those pattern workflows.

Enterprise Logic

There are a number of variants of the pattern workflow you will get from Dataflow, however all of them share the identical enterprise logic. This was a acutely aware choice with a view to clearly illustrate the distinction between numerous languages through which your ETL might be written in. Clearly not all instruments are made with the identical use case in thoughts, so we’re planning so as to add extra code samples for different (than classical batch ETL) information processing functions, e.g. Machine Studying mannequin constructing and scoring.

The instance enterprise logic we use in our template computes the highest hundred motion pictures/exhibits in each nation the place Netflix operates each day. This isn’t an precise manufacturing pipeline operating at Netflix, as a result of it’s a extremely simplified code but it surely serves nicely the aim of illustrating a batch ETL job with numerous transformation levels. Let’s evaluation the transformation steps under.

Step 1: each day, incrementally, sum up all viewing time of all motion pictures and exhibits in each nation

, country_code
, SUM(view_hours) AS view_hours
FROM some_db.source_table
WHERE playback_date = CURRENT_DATE
, country_code

Step 2: rank all titles from most watched to least in each county

, country_code
, view_hours
PARTITION BY country_code
ORDER BY view_hours DESC
) AS title_rank

Step 3: filter all titles to the highest 100

, country_code
, view_hours
, title_rank
WHERE title_rank <= 100

Now, utilizing the above easy 3-step transformation we are going to produce information that may be written to the next Iceberg desk:

CREATE TABLE IF NOT EXISTS $TARGET_DB.dataflow_sample_results (
title_id INT COMMENT "Title ID of the film or present."
, country_code STRING COMMENT "Nation code of the playback session."
, title_rank INT COMMENT "Rank of a given title in a given nation."
, view_hours DOUBLE COMMENT "Complete viewing hours of a given title in a given nation."
"Instance dataset dropped at you by Dataflow. For extra data on this
and different examples please go to the Dataflow documentation web page."
date DATE COMMENT "Playback date."

As you may infer from the above desk construction we’re going to load about 19,000 rows into this desk each day. And they’re going to look one thing like this:

 sql> SELECT * FROM foo.dataflow_sample_results 
WHERE date = 20220101 and country_code = 'US'
ORDER BY title_rank LIMIT 5;

title_id | country_code | title_rank | view_hours | date
11111111 | US | 1 | 123 | 20220101
44444444 | US | 2 | 111 | 20220101
33333333 | US | 3 | 98 | 20220101
55555555 | US | 4 | 55 | 20220101
22222222 | US | 5 | 11 | 20220101
(5 rows)

With the enterprise logic out of the best way, we will now begin speaking in regards to the parts, or the boiler-plate, of our pattern workflows.


Let’s take a look at the commonest workflow parts that we use at Netflix. These parts could not match into each ETL use case, however are used usually sufficient to be included in each template (or pattern workflow). The workflow writer, in any case, has the ultimate phrase on whether or not they need to use all of those patterns or preserve just some. Both method they’re right here to start out with, able to go, if wanted.

Workflow Definitions

Under you may see a typical file construction of a pattern workflow package deal written in SparkSQL.

├── backfill.sch.yaml
├── every day.sch.yaml
├── fundamental.sch.yaml
├── ddl
│ └── dataflow_sparksql_sample.sql
└── src
├── mocks
│ ├── dataflow_pyspark_sample.yaml
│ └── some_db.source_table.yaml
├── sparksql_write.sql

Above bolded recordsdata outline a sequence of steps (a.ok.a. jobs) their cadence, dependencies, and the sequence through which they need to be executed.

That is a method we will tie parts collectively right into a cohesive workflow. In each pattern workflow package deal there are three workflow definition recordsdata that work collectively to supply versatile performance. The pattern workflow code assumes a every day execution sample, however it is rather straightforward to regulate them to run at totally different cadence. For the workflow orchestration we use Netflix homegrown Maestro scheduler.

The fundamental workflow definition file holds the logic of a single run, on this case one day-worth of information. This logic consists of the next elements: DDL code, desk metadata data, information transformation and some audit steps. It’s designed to run for a single date, and meant to be referred to as from the every day or backfill workflows. This fundamental workflow may also be referred to as manually throughout improvement with arbitrary run-time parameters to get a really feel for the workflow in motion.

The every day workflow executes the fundamental one each day for the predefined variety of earlier days. That is typically vital for the aim of catching up on some late arriving information. That is the place we outline a set off schedule, notifications schemes, and replace the “high water mark” timestamps on our goal desk.

The backfill workflow executes the fundamental for a specified vary of days. That is helpful for restating information, most frequently due to a metamorphosis logic change, however typically as a response to upstream information updates.


Usually, step one in a knowledge pipeline is to outline the goal desk construction and column metadata through a DDL assertion. We perceive that some people select to have their output schema be an implicit results of the remodel code itself, however the express assertion of the output schema is just not solely helpful for including desk (and column) degree feedback, but additionally serves as one option to validate the remodel logic.

├── backfill.sch.yaml
├── every day.sch.yaml
├── fundamental.sch.yaml
├── ddl
│ └── dataflow_sparksql_sample.sql
└── src
├── mocks
│ ├── dataflow_pyspark_sample.yaml
│ └── some_db.source_table.yaml
├── sparksql_write.sql

Typically, we want to execute DDL instructions as a part of the workflow itself, as an alternative of operating outdoors of the schedule, as a result of it simplifies the event course of. See under instance of hooking the desk creation SQL file into the fundamental workflow definition.

      - job:
id: ddl
sort: Spark
script: $S3./ddl/dataflow_sparksql_sample.sql


The metadata step offers context on the output desk itself in addition to the information contained inside. Attributes are set through Metacat, which is a Netflix inner metadata administration platform. Under is an instance of plugging that metadata step within the fundamental workflow definition

     - job:
id: metadata
sort: Metadata
proprietor: $username
- dataflow
- pattern
lifetime: 123
date: pk
country_code: pk
rank: pk


The transformation step (or steps) might be executed within the developer’s language of selection. The instance under is utilizing SparkSQL.

├── backfill.sch.yaml
├── every day.sch.yaml
├── fundamental.sch.yaml
├── ddl
│ └── dataflow_sparksql_sample.sql
└── src
├── mocks
│ ├── dataflow_pyspark_sample.yaml
│ └── some_db.source_table.yaml
├── sparksql_write.sql

Optionally, this step can use the Write-Audit-Publish pattern to make sure that information is appropriate earlier than it’s made obtainable to the remainder of the corporate. See instance under:

      - template:
id: wap
sort: wap
- job:
id: write
sort: Spark
script: $S3./src/sparksql_write.sql


Audit steps might be outlined to confirm information high quality. If a “blocking” audit fails, the job will halt and the write step is just not dedicated, so invalid information is not going to be uncovered to customers. This step is non-compulsory and configurable, see a partial instance of an audit from the fundamental workflow under.

- perform: columns_should_not_have_nulls
blocking: true
- title_id

Excessive-Water-Mark Timestamp

A profitable write will usually be adopted by a metadata name to set the legitimate time (or high-water mark) of a dataset. This permits different processes, consuming our desk, to be notified and begin their processing. See an instance excessive water mark job from the fundamental workflow definition.

      - job:
id: hwm
sort: HWM
hwm_datetime: $EXECUTION_DATE

Unit Exams

Unit take a look at artifacts are additionally generated as a part of the pattern workflow construction. They consist of information mocks, the precise take a look at code, and a easy execution harness relying on the workflow language. See the bolded file under.

├── backfill.sch.yaml
├── every day.sch.yaml
├── fundamental.sch.yaml
├── ddl
│ └── dataflow_sparksql_sample.sql
└── src
├── mocks
│ ├── dataflow_pyspark_sample.yaml
│ └── some_db.source_table.yaml
├── sparksql_write.sql

These unit checks are meant to check one “unit” of information remodel in isolation. They are often run throughout improvement to rapidly seize code typos and syntax points, or throughout automated testing/deployment section, to guarantee that code modifications haven’t damaged any checks.

We wish unit checks to run rapidly in order that we will have steady suggestions and quick iterations throughout the improvement cycle. Operating code towards a manufacturing database might be sluggish, particularly with the overhead required for distributed information processing programs like Apache Spark. Mocks will let you run checks regionally towards a small pattern of “actual” information to validate your transformation code performance.


Over time, the extraction of information from Netflix’s supply programs has grown to embody a wider vary of end-users, equivalent to engineers, information scientists, analysts, entrepreneurs, and different stakeholders. Specializing in comfort, Dataflow permits for these differing personas to go about their work seamlessly. A lot of our information customers make use of SparkSQL, pyspark, and Scala. A small however rising contingency of information scientists and analytics engineers use R, backed by the Sparklyr interface or different information processing instruments, like Metaflow.

With an understanding that the information panorama and the applied sciences employed by end-users will not be homogenous, Dataflow creates a malleable path towards. It solidifies totally different recipes or repeatable templates for information extraction. Inside this part, we’ll preview a couple of strategies, beginning with sparkSQL and python’s method of making information pipelines with dataflow. Then we’ll segue into the Scala and R use circumstances.

To start, after putting in Dataflow, a person can run the next command to grasp get began.

$ dataflow pattern workflow --help                                                         
Dataflow (0.6.16)

Utilization: dataflow pattern workflow [OPTIONS] RECIPE [TARGET_PATH]

Create a pattern workflow based mostly on chosen RECIPE and land it within the
specified TARGET_PATH.

At present supported workflow RECIPEs are: spark-sql, pyspark,
scala and sparklyr.

- if not specified, present listing is assumed
- factors to a listing, it is going to be used because the goal location

--source-path TEXT Supply path of the pattern workflows.
--workflow-shortname TEXT Workflow quick title.
--workflow-id TEXT Workflow ID.
--skip-info Skip the data in regards to the workflow pattern.
--help Present this message and exit.

As soon as once more, let’s assume now we have a listing referred to as stranger-data through which the person creates workflow templates in all 4 languages that Dataflow presents. To raised illustrate generate the pattern workflows utilizing Dataflow, let’s have a look at the complete command one would use to create one in every of these workflows, e.g:

$ cd stranger-data
$ dataflow pattern workflow spark-sql ./sparksql-workflow

By repeating the above command for every sort of transformation language we will arrive on the following listing construction

├── pyspark-workflow
│ ├── fundamental.sch.yaml
│ ├── every day.sch.yaml
│ ├── backfill.sch.yaml
│ ├── ddl
│ │ └── ...
│ ├── src
│ │ └── ...
│ └── tox.ini
├── scala-workflow
│ ├── construct.gradle
│ └── ...
├── sparklyR-workflow
│ └── ...
└── sparksql-workflow
└── ...

Earlier we talked in regards to the enterprise logic of those pattern workflows and we confirmed the Spark SQL model of that instance information transformation. Now let’s focus on totally different approaches to writing the information in different languages.


This partial pySpark code under may have the identical performance because the SparkSQL instance above, but it surely makes use of Spark dataframes Python interface.

def fundamental(args, spark):

source_table_df = spark.desk(f"some_db.source_table)

viewing_by_title_country = (
source_table_df.choose("title_id", "country_code",
.filter(col("date") == date)
.filter("title_id IS NOT NULL AND view_hours > 0")
.groupBy("title_id", "country_code")

window = Window.partitionBy(

ranked_viewing_by_title_country = viewing_by_title_country.withColumn(
"title_rank", rank().over(window)

col("title_rank") <= 100
"date", lit(int(date))
target_table, overwrite=True


Scala is one other Dataflow supported recipe that gives the identical enterprise logic in a pattern workflow out of the field.

package deal

object ExampleApp
import spark.implicits._

def readSourceTable(sourceDb: String, dataDate: String): DataFrame =
.filter($"playback_start_date" === dataDate)

def viewingByTitleCountry(sourceTableDF: DataFrame): DataFrame =
.choose($"title_id", $"country_code", $"view_hours")
.filter($"view_hours" > 0)
.groupBy($"title_id", $"country_code")

def addTitleRank(viewingDF: DataFrame): DataFrame =
"title_rank", F.rank().over(

def writeViewing(viewingDF: DataFrame, targetTable: String, dataDate: String): Unit =
.choose($"title_id", $"country_code", $"title_rank", $"view_hours")
.filter($"title_rank" <= 100)
.withColumn("date", F.lit(dataDate.toInt))

def fundamental():
sourceTableDF = readSourceTable("some_db", "source_table", 20200101)
viewingDf = viewingByTitleCountry(sourceTableDF)
titleRankedDf = addTitleRank(viewingDF)

R / sparklyR

As Netflix has a rising cohort of R customers, R is the newest recipe obtainable in Dataflow.



fundamental <- perform(args, spark) >
fundamental(args = args, spark = spark)