DataflowTemplates
DataflowTemplates copied to clipboard
[Bug]: GSC to PubSub losing messages
Related Template(s)
GCS_Text_to_Cloud_PubSub
Template Version
2024-09-19-00_rc00
What happened?
I'm using de GCS to PubSub batch template to publish and consume messages from JSONL files.
I've noticed I'm losing messages in the process, I'm not sure the ratio of missing messages but I have to run the job two to three times to successfully receive a few thousands messages (something like 10k to 20k)
I'm just worried because I'll have to do this process for billions of messages, and I definitely would not like to process this amount of data many time to ensure I wont be losing anything
Is there some know bug about it? Or something I cant do to to enforce the job to be more precise?
Relevant log output
No response