RE: issue with synchronized_standby_slots - Mailing list pgsql-hackers

From Hayato Kuroda (Fujitsu)
Subject RE: issue with synchronized_standby_slots
Date
Msg-id OSCPR01MB149662ED9CDEB4C6D5F212FAAF5F0A@OSCPR01MB14966.jpnprd01.prod.outlook.com
Whole thread Raw
In response to Re: issue with synchronized_standby_slots  (Shlok Kyal <shlok.kyal.oss@gmail.com>)
Responses Re: issue with synchronized_standby_slots
List pgsql-hackers
Dear Shlok,

Thanks for updating the patch! Few comments.

1.
You must separate patch for master and PG18, because ReplicationSlotValidateNameInternal()
does not accept `bool allow_reserved_name` in the version.

2.
Also, test for PG18 should not have the case which rejects the reserved name.

3.
```
-- Parallel worker does not throw error during startup.
SET min_parallel_table_scan_size TO 0;
SET parallel_setup_cost TO 0;
SET parallel_tuple_cost TO 0;
```

According to contrib/pg_stat_statements/sql/parallel.sql, max_parallel_workers_per_gather
should be also set. There is a possiblity that `make installcheck` is used and
it has max_parallel_workers_per_gather=0.

4.
```
    foreach_ptr(char, name, *elemlist)
```

You can add a comment atop here like:
Iterate the list to validate each slot name.

Best regards,
Hayato Kuroda
FUJITSU LIMITED


pgsql-hackers by date:

Previous
From: Shlok Kyal
Date:
Subject: Re: issue with synchronized_standby_slots
Next
From: Chao Li
Date:
Subject: Re: Optimize LISTEN/NOTIFY