confusion about Re: Write operations in parallel mode's update part. - Mailing list pgsql-hackers
hi,
i found a discuss about parallel dml, it wrote as follow,
Make updates and deletes parallel-restricted rather than
parallel-unsafe - i.e. allow updates and deletes but only in the
leader. This similarly would allow Update -> Gather -> whatever and
Delete -> Gather -> whatever. For this, you'd need a shared combo CID
hash so that workers can learn about new combo CIDs created by the
leader.
parallel-unsafe - i.e. allow updates and deletes but only in the
leader. This similarly would allow Update -> Gather -> whatever and
Delete -> Gather -> whatever. For this, you'd need a shared combo CID
hash so that workers can learn about new combo CIDs created by the
leader.
i have some questions about this,
when do update => gather => whatever, all update jobs done by leader , thus it know itself combo cid mapping,
and only other workers can not learn about that, so why those workers must know leader's combo cids? why those worker
need see leader's updated tuples, could you give me some example cases or Unusual scenes for for this parallel update?
pgsql-hackers by date: