From: | Sami Imseih <samimseih(at)gmail(dot)com> |
---|---|
To: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | improve DEBUG1 logging of parallel workers for CREATE INDEX? |
Date: | 2025-01-03 18:30:14 |
Message-ID: | CAA5RZ0trTUL6_vpvW79daGgkp7B-ZtWUc5yrPz5Sjm8Ns4KRgQ@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
While reviewing patch [1], I realized that the DEBUG1 message
for CREATE INDEX could do better in providing information
about parallel workers launched. Currently, the message just
shows how many workers are planned, but a user may want
to ensure that they have the appropriate number of workers
launched as well when they enable DEBUG1 logging.
Therefore, I want to suggest that instead of:
postgres=# CREATE INDEX tbl_c1 ON tbl(c1);
DEBUG: building index "tbl_c1" on table "tbl" with request for 2
parallel workers
DEBUG: index "tbl_c1" can safely use deduplication
CREATE INDEX
to enhance the message to:
postgres=# create index tbl_c1 on tbl(c1);
DEBUG: building index "tbl_c1" on table "tbl"
DEBUG: launched 0 parallel vacuum workers for index vacuuming (planned: 2)
DEBUG: index "tbl_c1" can safely use deduplication
CREATE INDEX
postgres=#
I prepared a simple patch, attached, for this. The log message matches
the format
used in VACUUM VERBOSE ( for consistency sake ).
Also, I think the documentation in [2] should be updated to mention that DEBUG1
can be used to view parallel worker usage for the build, but I have
not included it in
the patch yet.
Any thoughts?
[2] https://www.postgresql.org/docs/current/sql-createindex.html
Regards,
Sami Imseih
Amazon Web Services (AWS)
Attachment | Content-Type | Size |
---|---|---|
0001-improve-DEBUG1-logging-of-parallel-workers-for-CREAT.patch | application/octet-stream | 2.5 KB |
From | Date | Subject | |
---|---|---|---|
Next Message | Jim Nasby | 2025-01-03 19:08:02 | Re: Vacuum statistics |
Previous Message | Tom Lane | 2025-01-03 18:12:43 | Re: Fwd: Re: A new look at old NFS readdir() problems? |