Hi! I am getting BulkBatchFailed exception while using pk_chunking. I am using get_batch_list() for getting the batches from job_id and for each batch i am doing is_batch_done(), which is raising the exception.

[kunal4422]

No description provided.

[kunal4422]

Here is the associated code :

job_id = bulk.create_queryall_job("Contact", contentType='CSV',pk_chunking=100000,concurrency='Serial') batch_id = bulk.query(job_id, "select twitter_profile__c, userid__c from contact") batch_list=bulk.get_batch_list(job_id) for item in batch_list:batch_id=item['id'],job_id=item['jobId'] while not bulk.is_batch_done(batch_id=batch_id,job_id=job_id): #print(item) time.sleep(10)
Output : `
[{'id': '75118000004I1CtAAK', 'jobId': '75018000004sUucAAE', 'state': 'NotProcessed', 'createdDate': '2019-12-26T08:47:36.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1CyAAK', 'jobId': '75018000004sUucAAE', 'state': 'Queued', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1D3AAK', 'jobId': '75018000004sUucAAE', 'state': 'Queued', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1D8AAK', 'jobId': '75018000004sUucAAE', 'state': 'InProgress', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1D9AAK', 'jobId': '75018000004sUucAAE', 'state': 'Queued', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1DDAA0', 'jobId': '75018000004sUucAAE', 'state': 'Queued', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}, {'id': '75118000004I1DIAA0', 'jobId': '75018000004sUucAAE', 'state': 'Queued', 'createdDate': '2019-12-26T08:47:37.000Z', 'systemModstamp': '2019-12-26T08:47:37.000Z', 'numberRecordsProcessed': '0', 'numberRecordsFailed': '0', 'totalProcessingTime': '0', 'apiActiveProcessingTime': '0', 'apexProcessingTime': '0'}]
75118000004I1CtAAK
75018000004sUucAAE

---

BulkBatchFailed Traceback (most recent call last)
in
17 for item in batch_list:
18
---> 19 while not bulk.is_batch_done(batch_id=item['id'],job_id=item['jobId']):
20 #print(item)
21 time.sleep(10)

~\AppData\Local\Continuum\anaconda3\lib\site-packages\salesforce_bulk\salesforce_bulk.py in is_batch_done(self, batch_id, job_id)
426 if batch_state in bulk_states.ERROR_STATES:
427 status = self.batch_status(batch_id, job_id)
--> 428 raise BulkBatchFailed(job_id, batch_id, status.get('stateMessage'), batch_state)
429 return batch_state == bulk_states.COMPLETED
430

BulkBatchFailed: Batch 75118000004I1CtAAK of job 75018000004sUucAAE failed: None
`

[roshin8]

I added an example to the README which shows how to use PK Chunking. Let me know if that helps you.
https://github.com/heroku/salesforce-bulk/pull/77/files

[hdao1121]

I actually think this is a bug. is_batch_done throws an error when the return status is NOT_PROCESSED. But according to salesforce doc, NOT_PROCESSED is expected while chunking.

[Doctacon]

hdao1121, do you know if this ever got resolved?

[lambacck]

If you are at the point of using pk_chunking, you are better off not using is_batch_done and instead getting the batch status list to determine completion. The example in the linked PR is on path do that.