Hi Stanislav , On Pyspark DF can you the following
df.printSchema() and send the output please HTH Mich Talebzadeh, Dad | Technologist | Solutions Architect | Engineer London United Kingdom 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, 28 Dec 2023 at 12:31, Поротиков Станислав Вячеславович < s.poroti...@skbkontur.ru> wrote: > Ok. Thank you very much! > > > > Best regards, > > Stanislav Porotikov > > > > *From:* Mich Talebzadeh <mich.talebza...@gmail.com> > *Sent:* Thursday, December 28, 2023 5:14 PM > *To:* Hyukjin Kwon <gurwls...@apache.org> > *Cc:* Поротиков Станислав Вячеславович <s.poroti...@skbkontur.ru.invalid>; > user@spark.apache.org > *Subject:* Re: Pyspark UDF as a data source for streaming > > > > You can work around this issue by trying to write your DF to a flat file > and use Kafka to pick it up from the flat file and stream it in. > > > > Bear in mind that Kafa will require a unique identifier as K/V pair. Check > this link how to generate UUID for this purpose > > > https://stackoverflow.com/questions/49785108/spark-streaming-with-python-how-to-add-a-uuid-column > > > > HTH > > > Mich Talebzadeh, > > Dad | Technologist | Solutions Architect | Engineer > > London > > United Kingdom > > > > [image: Рисунок удален отправителем.] 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, 28 Dec 2023 at 09:33, Hyukjin Kwon <gurwls...@apache.org> wrote: > > Just fyi streaming python data source is in progress > > https://github.com/apache/spark/pull/44416 we will likely release this in > spark 4.0 > > > > On Thu, Dec 28, 2023 at 4:53 PM Поротиков Станислав Вячеславович > <s.poroti...@skbkontur.ru.invalid> wrote: > > Yes, it's actual data. > > > > Best regards, > > Stanislav Porotikov > > > > *From:* Mich Talebzadeh <mich.talebza...@gmail.com> > *Sent:* Wednesday, December 27, 2023 9:43 PM > *Cc:* user@spark.apache.org > *Subject:* Re: Pyspark UDF as a data source for streaming > > > > Is this generated data actual data or you are testing the application? > > > > Sounds like a form of Lambda architecture here with some > decision/processing not far from the attached diagram > > > > HTH > > > Mich Talebzadeh, > > Dad | Technologist | Solutions Architect | Engineer > > London > > United Kingdom > > > > [image: Рисунок удален отправителем.] 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, 27 Dec 2023 at 13:26, Поротиков Станислав Вячеславович < > s.poroti...@skbkontur.ru> wrote: > > Actually it's json with specific structure from API server. > > But the task is to check constantly if new data appears on API server and > load it to Kafka. > > Full pipeline can be presented like that: > > REST API -> Kafka -> some processing -> Kafka/Mongo -> … > > > > Best regards, > > Stanislav Porotikov > > > > *From:* Mich Talebzadeh <mich.talebza...@gmail.com> > *Sent:* Wednesday, December 27, 2023 6:17 PM > *To:* Поротиков Станислав Вячеславович <s.poroti...@skbkontur.ru.invalid> > *Cc:* user@spark.apache.org > *Subject:* Re: Pyspark UDF as a data source for streaming > > > > Ok so you want to generate some random data and load it into Kafka on a > regular interval and the rest? > > > > HTH > > Mich Talebzadeh, > > Dad | Technologist | Solutions Architect | Engineer > > London > > United Kingdom > > > > [image: Рисунок удален отправителем.] 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, 27 Dec 2023 at 12:16, Поротиков Станислав Вячеславович > <s.poroti...@skbkontur.ru.invalid> wrote: > > Hello! > > Is it possible to write pyspark UDF, generated data to streaming dataframe? > > I want to get some data from REST API requests in real time and consider > to save this data to dataframe. > > And then put it to Kafka. > > I can't realise how to create streaming dataframe from generated data. > > > > I am new in spark streaming. > > Could you give me some hints? > > > > Best regards, > > Stanislav Porotikov > > > >
