cancel
Showing results forย 
Search instead forย 
Did you mean:ย 
Generative AI
Explore discussions on generative artificial intelligence techniques and applications within the Databricks Community. Share ideas, challenges, and breakthroughs in this cutting-edge field.
cancel
Showing results forย 
Search instead forย 
Did you mean:ย 

ai_parse_document time out setting

TX-Aggie-00
New Contributor III

We are using ai_parse_document on ~4k pdf files.  We noticed the load is taking quite a bit of time and when we looked into our intermediate table, we are seeing several errors like this:

[
  {
    "error_message":"pdf rendering timed out after 1800 seconds.",
    "page_id":0
  }
]

30 minutes is quite a long time for this and larger files than the ones that failed were successfully parsed.  My question is: is there a way to catch these timeout errors earlier or change the default timeout setting?  

Thanks

1 ACCEPTED SOLUTION

Accepted Solutions

Louis_Frolio
Databricks Employee
Databricks Employee

Greetings @TX-Aggie-00 , 

Thanks for the detailed description and for including the exact error message โ€” that helps.

The โ€œpdf rendering timed out after 1800 secondsโ€ message reflects an internal hard timeout during the document rendering step. The 1800-second (30-minute) limit is fixed on the service side and is not currently exposed as a user-configurable setting.

In practical terms, there isnโ€™t a way to raise or lower that threshold from your end.

Hope this clarifies things.

Cheers, Lou.

View solution in original post

2 REPLIES 2

Louis_Frolio
Databricks Employee
Databricks Employee

Greetings @TX-Aggie-00 , 

Thanks for the detailed description and for including the exact error message โ€” that helps.

The โ€œpdf rendering timed out after 1800 secondsโ€ message reflects an internal hard timeout during the document rendering step. The 1800-second (30-minute) limit is fixed on the service side and is not currently exposed as a user-configurable setting.

In practical terms, there isnโ€™t a way to raise or lower that threshold from your end.

Hope this clarifies things.

Cheers, Lou.

Appreciate the quick response Lou!  It looks like these all occurred in the same batch, so I am guessing something else caused these.

Thanks, Scott