| From: | Zheng Li <zhengli10(at)gmail(dot)com> |
|---|---|
| To: | Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> |
| Cc: | Japin Li <japinli(at)hotmail(dot)com>, Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org>, Dilip Kumar <dilipbalaut(at)gmail(dot)com>, rajesh singarapu <rajesh(dot)rs0541(at)gmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org> |
| Subject: | Re: Support logical replication of DDLs |
| Date: | 2022-05-13 21:01:33 |
| Message-ID: | CAAD30U+CRgUgkAg33KzNBKwCbsgiSc5z3NYvxNzEfS0Zg2S1WA@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general pgsql-hackers |
> > 4. The handling related to partition tables seems missing because, on
> > the subscriber-side, it always creates a relation entry in
> > pg_subscription_rel which won't work. Check its interaction with
> > publish_via_partition_root.
>
> I will test it out.
Hi,
patch 0010 properly handles partitioned table creation on the apply
worker. Whether a replicated partitioned table should be added to
pg_subscription_rel catalog depends on the setting of
publish_via_partition_root of the publication. Thus we need to connect
to the source DB and check if the partitioned table is in
pg_catalog.pg_publication_tables after the apply worker creates the
partitioned table.
Thanks to Borui Yang for enabling and testing replication of DDL type
T_RenameStmt in patch 0009.
I've also rebased all the patches. Github branch of the same change
can be found here:
https://github.com/zli236/postgres/commits/ddl_replication
Regards,
Zheng
| Attachment | Content-Type | Size |
|---|---|---|
| 0005-Support-replication-of-CREATE-.-AS-.-and-SELECT-.-IN.patch | application/octet-stream | 8.0 KB |
| 0002-Support-logical-logging-and-decoding-of-DDL-command-.patch | application/octet-stream | 61.4 KB |
| 0003-Logical-replication-of-DDL-messages.patch | application/octet-stream | 41.2 KB |
| 0004-Enable-replication-of-CREATE-MATERIALIZED-VIEW-AS-st.patch | application/octet-stream | 4.1 KB |
| 0006-Add-couple-tests-for-DDL-replication-case.patch | application/octet-stream | 4.5 KB |
| 0009-Support-replication-of-DDL-type-T_RenameStmt-table-r.patch | application/octet-stream | 8.3 KB |
| 0001-Define-DDL-replication-levels-via-the-CREATE-PUBLICA.patch | application/octet-stream | 65.0 KB |
| 0007-Enable-logging-and-replication-of-DDLs-executed-insi.patch | application/octet-stream | 17.5 KB |
| 0008-Fail-replication-worker-on-DDL-command-that-rewrites.patch | application/octet-stream | 4.8 KB |
| 0010-Handle-partitioned-table-creation-on-the-apply-worke.patch | application/octet-stream | 11.5 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bryn Llewellyn | 2022-05-16 19:03:12 | Re: Deferred constraint trigger semantics |
| Previous Message | Niels Jespersen | 2022-05-13 16:25:21 | Logon via GSSAPI from Linux fails, but works from Windows |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Thomas Munro | 2022-05-13 21:25:08 | Re: Backends stunk in wait event IPC/MessageQueueInternal |
| Previous Message | Mikael Kjellström | 2022-05-13 20:35:22 | Re: recovery test failure on morepork with timestamp mystery |