nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Oracle Communities

Oracle Analytics and AI

Child Item

Self Replacing Data in Data Flow

User_GBD0K

Hi ,

I am looking to implement a manual tracker for one of ongoing project.

There is a time-series information that I would be manually upload to the system on a frequency basis

Basically Ive created a dataflow that look like below

The idea is that I run the query on the frequency basis to update the table. However Ive found it wont execute if I try to replace "Dataset A" as a save file and while using it as one of the source.

Unfortunately I wont be able to setup proper DB for this tracker to be use as a connection.

Seeking advise.

Thank you,

Regards

Find more posts tagged with

Data Modelling and Datasets

OAS

Data Flow

Accepted answers

GabriellePrichard-Oracle

Hello! Data flows will not allow you to read from and overwrite the same “Dataset A” within one data flow. Here's a pattern you can use instead that involves two data flows placed in a single sequence that can be run on a schedule -

Data Flow 1

Read Dataset A
Write it out immediately as Dataset A_backup (in case you need to keep a copy)
Union your “new portion of A” uploads with Dataset A
Write the combined result to Dataset A_temp

Data Flow 2

Read Dataset A_temp
Overwrite Dataset A with this refreshed dataset

Place these two data flows in 1 sequence.

Let me know if this works for you.

All comments

GabriellePrichard-Oracle

Data Flow 1

Read Dataset A
Write it out immediately as Dataset A_backup (in case you need to keep a copy)
Union your “new portion of A” uploads with Dataset A
Write the combined result to Dataset A_temp

Data Flow 2

Read Dataset A_temp
Overwrite Dataset A with this refreshed dataset

Place these two data flows in 1 sequence.

Let me know if this works for you.

User_GBD0K

Hi Prichard,

Thank you for the input, while the individual dataflow did work, when I placed them into a sequence it seem it would not execute properly.

It seem to failed on the second Data Flow step.

Below is the failed log for Sequence : Work A



************************************************************

*	2025-05-21 11:52:33 'XXXX'.'Work A'

************************************************************

74  INFO  2025-05-21 11:52:33.375 SGT  Starting to process request, run id: 711F5E10946E19AC71E1282BBAEB2E after being in queue for 21 second(s), 366 millisecond(s).

230  INFO  2025-05-21 11:52:39.792 SGT  Will run up to 5 task(s) in parallel

234  INFO  2025-05-21 11:52:39.826 SGT  Creating necessary connections.

238  INFO  2025-05-21 11:52:39.835 SGT  Sources created.

239  INFO  2025-05-21 11:52:39.841 SGT  143.156 (GB) of 2.50 (TB) used (5.592 %).

248  INFO  2025-05-21 11:52:39.868 SGT  Loading tasks for a new run.

251  FINER  2025-05-21 11:52:40.03 SGT  Loading of tasks completed

256  INFO  2025-05-21 11:52:40.128 SGT  A total of 3 task(s) need to be executed for this run.

259  INFO  2025-05-21 11:52:40.144 SGT  Executing tasks.

306  INFO  2025-05-21 11:52:45.626 SGT  Task "DFLW : Job 1" completed

343  SEVERE  2025-05-21 11:52:51.027 SGT  Execution of "dataflow : 'XXXX'.'Job 2' ('XXXX'.'Work A')" on BIJDBC failed. OBIS job id 40000000-5383-1a0b-8029-ac1f4c100000

[nQSError: 46236] Dataset Service error during Bulk Finalize - {"prefix":"DSS","code":50000,"message":"DSS understood the request, but was unable to perform the operation due to an internal error","subMessage":""}

[nQSError: 43224] The Dataflow "Job 2" failed during the execution.

[nQSError: 43204] Asynchronous Job Manager failed to execute the asynchronous job.

346  SEVERE  2025-05-21 11:52:51.061 SGT  An error occurred, which could not be interpreted.

352  SEVERE  2025-05-21 11:52:51.262 SGT  An error occurred, which could not be interpreted.

358  SEVERE  2025-05-21 11:52:51.362 SGT  An error occurred, which could not be interpreted.

362  SEVERE  2025-05-21 11:52:51.369 SGT  An error occurred, which could not be interpreted.

363  SEVERE  2025-05-21 11:52:51.377 SGT  Task "DFLW : Job 2 - Work A" failed.  Check the tasks log for more information

386  INFO  2025-05-21 11:52:51.563 SGT  Finished.

390  INFO  2025-05-21 11:52:51.623 SGT  

************************************************************

*	Summary

************************************************************


392  INFO  2025-05-21 11:52:51.629 SGT  Job Run Summary

	Total Number Of Tasks:	3

	Total Number Of Successful Rows across all tasks:	0

	Total Data Size across all tasks:	.00 (BYTE)


394  INFO  2025-05-21 11:52:51.635 SGT  Job finishing up....

398  ALL  2025-05-21 11:52:51.75 SGT  The process took 18 Second(s) 216 Milli-second(s) 

405  INFO  2025-05-21 11:52:51.769 SGT  Closing connections...

410  ALL  2025-05-21 11:52:51.781 SGT  The process failed.

Below is the failed log for Dataflow : Job 2



**************************************************************************

*	2025-05-21 11:52:45 DFLW : Job 2 - Work A

**************************************************************************

339  INFO  2025-05-21 11:52:45.974 SGT  "dataflow : 'XXXX'.'Job 2' ('XXXX'.'Work A')" on BIJDBC submitted with job id 40000000-5383-1a0b-8029-ac1f4c100000

341  INFO  2025-05-21 11:52:51.01 SGT  Finished execution with status Failed

342  SEVERE  2025-05-21 11:52:51.014 SGT  Execution failed. 

[nQSError: 46236] Dataset Service error during Bulk Finalize - {"prefix":"DSS","code":50000,"message":"DSS understood the request, but was unable to perform the operation due to an internal error","subMessage":""}

[nQSError: 43224] The Dataflow "Job 2" failed during the execution.

[nQSError: 43204] Asynchronous Job Manager failed to execute the asynchronous job.

365  INFO  2025-05-21 11:52:51.386 SGT  Execution time :18 second(s), 377 millisecond(s)

367  INFO  2025-05-21 11:52:51.39 SGT  

***********************************************************************************

*	Historical load summary for dataflow : 'XXXX'.'Job 2'

***********************************************************************************


	Number Of Successful Loads:	1

	Number Of Failed Loads:	0

	Load Duration:	14 second(s), 0 millisecond(s)

	Load Date:	2025-05-21 10:55:15.0

	Average Load Duration:	14 second(s), 0 millisecond(s)

	Minimum Load Duration:	14 second(s), 0 millisecond(s)

	Maximum Load Duration:	14 second(s), 0 millisecond(s)

Any way I can set this as a sequence?

GabriellePrichard-Oracle

You should be allowed to use a data flow output as another data flow input in a sequence. Do the task-level logs provide any more information?