Hi Pavel,
On Fri, Oct 15, 2021 at 5:02 PM Pavel Borisov <pashkin.elfe@gmail.com> wrote:
>> The lock taken on the parent is either ShareUpdateExclusiveLock or
>> AccessExclusiveLock depending on whether CONCURRENTLY is specified or
>> not. Maybe that should be considered also when locking the children.
>>
>> I've updated the patch that way. (Also, reintroduced the slightly
>> longer commit message that I had added in v3. :))
>
>
> Thanks Amit, for your work!
>
> I am little bit reluctant to the change you made in v5. As per
https://www.postgresql.org/docs/14/sql-altertable.html:
>
> > If CONCURRENTLY is specified, ... the second transaction acquires SHARE UPDATE EXCLUSIVE on the partitioned table
andACCESS EXCLUSIVE on the partition, and the detach process completes.
>
> In comment to find_all_inheritors():
>
> > The specified lock type is acquired on all child relations (but not on the given rel; caller should already have
lockedit)
>
> So I conclude that it is done in a right way in v3 with ACCESS_EXCLUSIVE lock.
Oops, you're right. I had failed to notice when reading the code that
the second transaction takes an AccessExclusiveLock on the target
partition. Reverted back to how this was in v3.
> Also I'd recommend removing the link to a discussion from the test. Anyway we have link in a commit message.
> -- Report: https://postgr.es/m/OS3PR01MB5718DA1C4609A25186D1FBF194089%40OS3PR01MB5718.jpnprd01.prod.outlook.com
Yeah, maybe the link is unnecessary in the test comment, so removed.
Though, I do occasionally see one of those in the test files (try `git
grep https src/test`).
Thanks again.
--
Amit Langote
EDB: http://www.enterprisedb.com