| From: | shihao zhong <zhong950419(at)gmail(dot)com> |
|---|---|
| To: | michael(at)paquier(dot)xyz |
| Cc: | alvherre(at)kurilemu(dot)de, pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: Partitioned index marked valid while a partition is pending detach |
| Date: | 2026-11-08 03:42:22 |
| Message-ID: | CAGRkXqTbOH1EXpO158LyJP8bB+vuyCN_TT89V+tp8eeNTxrV7g@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Fri, Nov 06, 2026 at 10:58:41AM +0900, Michael Paquier wrote:
> Perhaps we should try to work around simplifying
> RelationGetPartitionDesc() instead, and try to get rid of the
> omit_detached idea entirely?
Thanks, that made the fix much simpler. v4-0003 removes the
omit_detached argument of RelationGetPartitionDesc(), so DDL on the
parent now reaches a partition pending detach. FINALIZE then works
after ADD CHECK or RENAME COLUMN on the parent. This holds when
the partition is two levels down too.
I could not delete omit_detached entirely. Queries still need it, and
without it detach-partition-concurrently-1 to 4 fail.
0001 and 0002 work as is on 15 to 19. 0003 changes an API, so master only
Thanks,
Shihao
| Attachment | Content-Type | Size |
|---|---|---|
| v3-0003-Let-DDL-see-partitions-pending-detach.patch | application/octet-stream | 18.5 KB |
| v3-0002-Test-partitioned-index-validity-while-a-partition-is.patch | application/octet-stream | 4.3 KB |
| v3-0001-Count-partitions-pending-detach-when-validating-a-pa.patch | application/octet-stream | 1.2 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Xuneng Zhou | 2026-11-08 04:03:25 | Re: aio tests failing on newer Linux kernels |
| Previous Message | shihao zhong | 2026-11-08 02:56:39 | Re: Fix stale column context after COPY FROM conversion errors |