c2c-oerpscenario team mailing list archive

[John B <justbe.john@xxxxxxxxx>]

Sorry, to expand on the above:

The issue occurs when upgrading from a version of trunk prior to the
refactoring of the sequences. Previously, ir_sequence didn't have the
field/column 'implementation'. The default for that field is 'standard',
so when migrating it adds the column as 'standard'. The issue is that
this causes it to look in another table, ir_sequence_xxx (where xxx is
the sequence id) when generating the next in sequence. Because the
sequences were created prior to the refactoring, the table doesn't
exist.

The workaround is to manually update all sequences to implementation =
no_gap.

-- 
You received this bug notification because you are a member of C2C
OERPScenario, which is subscribed to OpenERP Project Group.
https://bugs.launchpad.net/bugs/868495

Title:
  [trunk] Sequence refactoring mucks up existing sequences

Status in OpenERP Server:
  Incomplete

Bug description:
  Submitting as a bug, though it may be better as a question re:
  migrating existing sequences.

  We're testing the latest updates and it looks the the sequence
  refactoring has caused an issue with existing sequences. Specifically,
  it looks like all sequence 'implementation' are set to 'standard'. For
  existing sequences, this causes an error when it looks for the table
  ir_sequence_XXX, e.g:

   Programming error: relation "ir_sequence_030" does not exist
   LINE 1: SELECT nextval('ir_sequence_030')

  As a workaround, I've set all existing sequences to no_gap instead of
  standard. I'm not sure this is correct, though.

  Thanks.

To manage notifications about this bug go to:
https://bugs.launchpad.net/openobject-server/+bug/868495/+subscriptions