Taking info from log files Failed indexing process with V15.0.504 #99550

https://thebrain.zendesk.com/agent/tickets/99550

Hello Xavier,

Thank you for the updated log file and for testing this on version 15.0.507. The additional details are very helpful.

From what you’ve confirmed so far, this behavior appears to be specific to your largest Brains:

• Indexing consistently stalls around ~71–72% on this Brain
• Indexing completes successfully on smaller Brains
• Indexing also works as expected on a newly created Brain

That distinction is important and gives us a clearer signal on where to focus.

More broadly, to address your question about large PDFs and indexing:

TheBrain is capable of storing and indexing large PDF files, but indexing performance is generally better when content is distributed across smaller, interconnected Brains rather than a single Brain containing thousands of large attachments. Very large attachment sets can significantly increase indexing time and resource usage, which aligns with what we’re seeing in this case.

That said, this guidance is about performance characteristics and best practices. It does not fully explain why indexing is stalling at a specific point, which is why we’re continuing to investigate this behavior rather than treating it as expected or acceptable.

At this stage, the next step is to review the new log you provided under 15.0.507 and correlate it with the attachment processing behavior we’re seeing. We’ll follow up once we’ve had a chance to analyze it more closely and determine whether this is something we can improve on our side.

Thank you again for your patience and for working through this with us.

Best regards,

John