OK, that sounds reasonable.
In the code below
#Aggregation code in Alarm call, which uses withWatermark
def computeCount(df_processedAlarm, df_totalAlarm):
processedAlarmCnt = None
if df_processedAlarm.count() > 0:
processedAlarmCnt =
df_processedAlarm.withWatermark("timestamp", "10 seconds")\
.groupBy(
window(col("timestamp"), "1 minutes").alias("window")
).count()
It is more efficient to use
* if(len(df_processedAlarm.take(1)) > 0:*
processedAlarmCnt =
df_processedAlarm.withWatermark("timestamp", "10 seconds")\
.groupBy(
window(col("timestamp"), "1 minutes").alias("window")
).count()
else:
print("DataFrame is empty")
HTH
view my Linkedin profile
<https://www.linkedin.com/in/mich-talebzadeh-ph-d-5205b2/>
https://en.everybodywiki.com/Mich_Talebzadeh
*Disclaimer:* Use it at your own risk. Any and all responsibility for any
loss, damage or destruction of data or any other property which may arise
from relying on this email's technical content is explicitly disclaimed.
The author will in no case be liable for any monetary damages arising from
such loss, damage or destruction.
On Thu, 17 Feb 2022 at 06:33, karan alang <karan.al...@gmail.com> wrote:
> Hi Mich,
> the issue was related to incorrect, which is resolved.
>
> However, wrt your comment - 'OK sounds like your watermark is done
> outside of your processing.'
>
> In my use-case which primarily deals with syslogs, syslog is a string
> which needs to be parsed (with defensive coding built in to ensure records
> are in correct format), before it is fed to
> 3 different classes (AlarmProc being one of them) - where there is
> additional parsing + aggregation for specific types of logs.
> The way I'm handling this is by using -- foreachBatch(convertToDict) in
> the writeStream method, and the parsing + aggregation happens for the
> microbatch.
> foreachBatch - will wait for the parsing and aggregation to complete for
> the microbatch, and then proceed to do the same with the next microbatch.
>
> Since it involves a lot of parsing + aggregation, it requires more than a
> df.select() - hence the approach above is taken.
> From what I understand, the watermark is done within the processing ..
> since it is done per microbatch pulled with each trigger.
>
> Pls let me know if you have comments/suggestions on this approach.
>
> thanks,
> Karan Alang
>
>
> On Wed, Feb 16, 2022 at 12:52 AM Mich Talebzadeh <
> mich.talebza...@gmail.com> wrote:
>
>> OK sounds like your watermark is done outside of your processing.
>>
>> Check this
>>
>> # construct a streaming dataframe streamingDataFrame that
>> subscribes to topic temperature
>> streamingDataFrame = self.spark \
>> .readStream \
>> .format("kafka") \
>> .option("kafka.bootstrap.servers",
>> config['MDVariables']['bootstrapServers'],) \
>> .option("schema.registry.url",
>> config['MDVariables']['schemaRegistryURL']) \
>> .option("group.id", config['common']['appName']) \
>> .option("zookeeper.connection.timeout.ms",
>> config['MDVariables']['zookeeperConnectionTimeoutMs']) \
>> .option("rebalance.backoff.ms",
>> config['MDVariables']['rebalanceBackoffMS']) \
>> .option("zookeeper.session.timeout.ms",
>> config['MDVariables']['zookeeperSessionTimeOutMs']) \
>> .option("auto.commit.interval.ms",
>> config['MDVariables']['autoCommitIntervalMS']) \
>> .option("subscribe", "temperature") \
>> .option("failOnDataLoss", "false") \
>> .option("includeHeaders", "true") \
>> .option("startingOffsets", "latest") \
>> .load() \
>> .select(from_json(col("value").cast("string"),
>> schema).alias("parsed_value"))
>>
>>
>> resultM = streamingDataFrame.select( \
>> col("parsed_value.rowkey").alias("rowkey") \
>> , col("parsed_value.timestamp").alias("timestamp") \
>> , col("parsed_value.temperature").alias("temperature"))
>> result = resultM. \
>> withWatermark("timestamp", "5 minutes"). \
>> groupBy(window(resultM.timestamp, "5 minutes", "5
>> minutes")). \
>> avg('temperature'). \
>> writeStream. \
>> outputMode('complete'). \
>> option("numRows", 1000). \
>> option("truncate", "false"). \
>> format('console'). \
>> option('checkpointLocation', checkpoint_path). \
>> queryName("temperature"). \
>> start()
>>
>> HTH
>>
>>
>>
>> view my Linkedin profile
>> <https://www.linkedin.com/in/mich-talebzadeh-ph-d-5205b2/>
>>
>>
>> https://en.everybodywiki.com/Mich_Talebzadeh
>>
>>
>>
>> *Disclaimer:* Use it at your own risk. Any and all responsibility for
>> any loss, damage or destruction of data or any other property which may
>> arise from relying on this email's technical content is explicitly
>> disclaimed. The author will in no case be liable for any monetary damages
>> arising from such loss, damage or destruction.
>>
>>
>>
>>
>> On Wed, 16 Feb 2022 at 06:37, karan alang <karan.al...@gmail.com> wrote:
>>
>>>
>>> Hello All,
>>>
>>> I have a Structured Streaming pyspark program running on GCP Dataproc,
>>> which reads data from Kafka, and does some data massaging, and aggregation.
>>> I'm trying to use withWatermark(), and it is giving error.
>>>
>>> py4j.Py4JException: An exception was raised by the Python Proxy. Return
>>> Message: Traceback (most recent call last):
>>>
>>> File
>>> "/usr/lib/spark/python/lib/py4j-0.10.9-src.zip/py4j/java_gateway.py", line
>>> 2442, in _call_proxy
>>>
>>> return_value = getattr(self.pool[obj_id], method)(*params)
>>>
>>> File "/usr/lib/spark/python/lib/pyspark.zip/pyspark/sql/utils.py",
>>> line 196, in call
>>>
>>> raise e
>>>
>>> File "/usr/lib/spark/python/lib/pyspark.zip/pyspark/sql/utils.py",
>>> line 193, in call
>>>
>>> self.func(DataFrame(jdf, self.sql_ctx), batch_id)
>>>
>>> File
>>> "/tmp/178d0ac9c82e42a09942f7f9cdc76bb7/StructuredStreaming_GCP_Versa_Sase_gcloud.py",
>>> line 444, in convertToDictForEachBatch
>>>
>>> ap = Alarm(tdict, spark)
>>>
>>> File
>>> "/tmp/178d0ac9c82e42a09942f7f9cdc76bb7/StructuredStreaming_GCP_Versa_Sase_gcloud.py",
>>> line 356, in __init__
>>>
>>> computeCount(l_alarm_df, l_alarm1_df)
>>>
>>> File
>>> "/tmp/178d0ac9c82e42a09942f7f9cdc76bb7/StructuredStreaming_GCP_Versa_Sase_gcloud.py",
>>> line 262, in computeCount
>>>
>>> window(col("timestamp"), "10 minutes").alias("window")
>>>
>>> TypeError: 'module' object is not callable
>>>
>>> Details are in stackoverflow below :
>>>
>>> https://stackoverflow.com/questions/71137296/structuredstreaming-withwatermark-typeerror-module-object-is-not-callable
>>>
>>> Any ideas on how to debug/fix this ?
>>> tia !
>>>
>>
