Changes between Initial Version and Version 1 of Ticket #16484, comment 1


Ignore:
Timestamp:
Aug 6, 2011, 2:55:30 PM (13 years ago)
Author:
Aymeric Augustin

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #16484, comment 1

    initial v1  
    1 From code inspection, I don't see how this can happen: it the call to `obj.save(force_insert=must_create, using=using)`, `must_create` is `False`, so the session should just be updated if it already exists. Here, you're hitting a naked `raise`, which means that the original author thought this should never happen.
     1From code inspection, I don't see how this can happen: in the call to `obj.save(force_insert=must_create, using=using)`, `must_create` is `False`, so the session should just be updated if it already exists. Here, you're hitting a naked `raise`, which means that the original author thought this should never happen.
    22
    33I can only imagine one cause for a race condition: if two requests running in two separate processes open a transaction, then commit: under some isolation levels, this may fail.
Back to Top