Can you provide a reproducible test case? There are really too many
moving parts here to know what is going on without being able to
reproduce it.
On Jul 1, 2006, at 1:24 PM, Luiz Fernando B. Ribeiro wrote:
Hello,
In a python script I got a structure like this:
# Many request checks
# A zSQl method deleting some old data
cotainer.doDelete(key=key)
errors = [ ]
for data in list:
# Other checks
if error:
errors.append(error)
continue
# Inserts in mysql database
cotainer.doInsert(data=data, other=stuff)
if errors:
raise "An Error with list of error: %s" % ', '.join(errors)
# --- cut ---
The problem is that the raise is not rolling back the delete,
looking in mysql log the following sequence was found:
--Begin transaction
[delete statement]
[the insert statements]
--Rollback
--Begin transaction !!
[delete statement again] !!
--Commit !!!
We tried it many times and revised our code. The only solution was
to put the delete statement in weird block like this:
if True:
cotainer.doDelete(key=key)
As if the rollback was only effective against a "2nd level"
instruction. This is a really odd behavior but with this trick the
rollback was done correctly.
Has anyone ever seen this?
Luiz Fernando B. Ribeiro
_______________________________________________
Zope maillist - Zope@zope.org
http://mail.zope.org/mailman/listinfo/zope
** No cross posts or HTML encoding! **
(Related lists - http://mail.zope.org/mailman/listinfo/zope-announce
http://mail.zope.org/mailman/listinfo/zope-dev )
_______________________________________________
Zope maillist - Zope@zope.org
http://mail.zope.org/mailman/listinfo/zope
** No cross posts or HTML encoding! **
(Related lists -
http://mail.zope.org/mailman/listinfo/zope-announce
http://mail.zope.org/mailman/listinfo/zope-dev )
