--- Sasha Pachev <sasha@stripped> wrote:
> Very possibly a bug. INNODB deadlocks itself during
> the operation, then detects
> the deadlock and aborts the transaction. If this is
> indeed a bug, I am sure
> Heikki would like to fix it ASAP.
There actually was a "create index" that was being
executed on the web_master table.
Still, that shouldn't result in a deadlock creating
the other table should it? DDL against the web_master
table of course should lock timeout because of the
index creation, but you should be able to create
another table based off that table with no problem,
since you can obviously get a consistent read off the
base table.
Should I create a simple test scenario and post it to
the bug list?
