Receiving Null values from Eventhub streaming.

[vkumar]

Hi, I am new to PySpark, and facing an issue while consuming data from the Azure eventhub. I am unable to deserialize the consumed data. I see only null values upon deserializing data using the schema. Please find the below schema, eventhub message, and code that I am using to consume data and let me know how can I resolve this issue. Thanks in advance.

Sample Eventhub message:
EventHubOverrideMessage(gtin13=00******0010, sourceId=0******5, lastUpdateTimestamp=2024-07-09T12:45:00.009805, lastUpdatedUser=null, inStore=ModalityOverride(reason=Other, override=true, startDate=2023-10-23T00:00, endDate=null), pickup=ModalityOverride(reason=Other, override=true, startDate=2023-10-23T00:00, endDate=null), delivery=ModalityOverride(reason=Other, override=true, startDate=2023-10-23T00:00, endDate=null), ship=null)

Deserialization Schema: (Tried replacingTimestampType with stringtype)

StructType([

StructField("gtin13", StringType(), True),

StructField("sourceId", StringType(), True),

StructField("lastUpdateTimestamp", TimestampType(), True),

StructField("lastUpdatedUser", StringType(), True),

StructField("inStore", StructType([

StructField("reason", StringType(), True),

StructField("override", BooleanType(), True),

StructField("startDate", TimestampType(), True),

StructField("endDate", TimestampType(), True)

]), True),

StructField("pickup", StructType([

StructField("reason", StringType(), True),

StructField("override", BooleanType(), True),

StructField("startDate", TimestampType(), True),

StructField("endDate", TimestampType(), True)

]), True),

StructField("delivery", StructType([

StructField("reason", StringType(), True),

StructField("override", BooleanType(), True),

StructField("startDate", TimestampType(), True),

StructField("endDate", TimestampType(), True)

]), True),

StructField("ship", StructType([

StructField("reason", StringType(), True),

StructField("override", BooleanType(), True),

StructField("startDate", TimestampType(), True),

StructField("endDate", TimestampType(), True)

]), True)

])

Actual Schema:

{

"title": "schema",

"type": "object",

"properties": {

"gtin13": {

"type": "string",

"pattern": "^[0-9]{13}$"

},

"sourceId": {

"type": "string",

"pattern": "^[0-9]{8}$"

},

"lastUpdateTimestamp": {

"type": "string",

"format": "date-time"

},

"inStore": {

"type": [

"object",

"null"

],

"properties": {

"reason": {

"type": [

"string",

"null"

]

},

"override": {

"type": [

"boolean",

"null"

]

},

"startDate": {

"type": [

"string",

"null"

],

"format": "date-time"

},

"endDate": {

"type": [

"string",

"null"

],

"format": "date-time"

}

},

"required": [

"reason",

"override",

"startDate",

"endDate"

]

},

"pickup": {

"type": [

"object",

"null"

],

"properties": {

"reason": {

"type": "string"

},

"override": {

"type": [

"boolean",

"null"

]

},

"startDate": {

"type": [

"string",

"null"

],

"format": "date-time"

},

"endDate": {

"type": [

"string",

"null"

],

"format": "date-time"

}

},

"required": [

"reason",

"override",

"startDate",

"endDate"

]

},

"delivery": {

"type": [

"object",

"null"

],

"properties": {

"reason": {

"type": [

"string",

"null"

]

},

"override": {

"type": [

"boolean",

"null"

]

},

"startDate": {

"type": [

"string",

"null"

],

"format": "date-time"

},

"endDate": {

"type": [

"string",

"null"

],

"format": "date-time"

}

},

"required": [

"reason",

"override",

"startDate",

"endDate"

]

},

"ship": {

"type": [

"object",

"null"

],

"properties": {

"reason": {

"type": [

"string",

"null"

]

},

"override": {

"type": [

"boolean",

"null"

]

},

"startDate": {

"type": [

"string",

"null"

],

"format": "date-time"

},

"endDate": {

"type": [

"string",

"null"

],

"format": "date-time"

}

},

"required": [

"reason",

"override",

"startDate",

"endDate"

]

}

},

"required": [

"gtin13",

"sourceId",

"lastUpdateTimestamp"

]

}

CODE:

kafkaDF = (

spark.readStream.format("kafka")

.option("subscribe", topic)

.option(

"kafka.bootstrap.servers",

bootstrap_server,

)

.option("kafka.sasl.mechanism", "PLAIN")

.option("kafka.security.protocol", "SASL_SSL")

.option("kafka.sasl.jaas.config",EVENTHUB_CONNECTION_STRING)

.option("kafka.request.timeout.ms", "60000")

.option("kafka.session.timeout.ms", "30000")

.option("startingOffsets", "earliest")

.option("mode", "PERMISSIVE")

.load()

.withColumn("value", col("value").cast(StringType()))

)

print("Schema of Kafka DataFrame:")

kafkaDF.printSchema()

# Deserialize JSON data

deserialized_stream = kafkaDF.withColumn(

"data", F.from_json(col("value").cast("string"), json_schema)

)

parsed_df = deserialized_stream.select("data.*")

# Display the parsed DataFrame

display(deserialized_stream)

PRINT SCHEMA OUTPUT:
Schema of Kafka DataFrame: root |-- key: binary (nullable = true) |-- value: string (nullable = true) |-- topic: string (nullable = true) |-- partition: integer (nullable = true) |-- offset: long (nullable = true) |-- timestamp: timestamp (nullable = true) |-- timestampType: integer (nullable = true)

Deserialized Output:
{"gtin13":null,"sourceId":null,"lastUpdateTimestamp":null,"inStore":null,"pickup":null,"delivery":null,"ship":null}

[Kaniz_Fatma]

Hi @vkumar, The issue you are facing is likely due to the way the data is being serialized and deserialized in the Azure Event Hub.

• The sample Event Hub message you provided is in a custom format, which is not the typical JSON format. The message appears to be a custom class called EventHubOverrideMessage with nested objects. This means that the data is not in a simple JSON format, and the default JSON deserialization may not work.
• Since the data is not in a standard JSON format, you will need to implement custom deserialization to handle the specific structure of the EventHubOverrideMessage class and its nested objects.
• The sample Event Hub message you provided has some fields that can be null. Make sure your schema and deserialization logic can handle these null values correctly.
• After implementing the custom deserialization, validate the output of the deserialized_stream DataFrame to ensure that the data is being parsed correctly and that there are no null values or other issues.