[DOC] Fix for the missing pg_stat_progress_cluster view phase column value
Hi,
I found a missing column value in the pg_stat_progress_cluster view document.
I read the src/backend/catalog/system_views.sql file, there seems to be a possibility that 'writing new heap' is output in the 'phase' column.
The attached patch adds a description of the 'writing new heap' value output in the 'phase' column.
Regards,
Noriyoshi Shinoda
Attachments:
monitoring_sgml_v1.diffapplication/octet-stream; name=monitoring_sgml_v1.diffDownload
diff --git a/doc/src/sgml/monitoring.sgml b/doc/src/sgml/monitoring.sgml
index 828e908..e6c49ee 100644
--- a/doc/src/sgml/monitoring.sgml
+++ b/doc/src/sgml/monitoring.sgml
@@ -4079,6 +4079,12 @@ SELECT pg_stat_get_backend_pid(s.backendid) AS pid,
</entry>
</row>
<row>
+ <entry><literal>writing new heap</literal></entry>
+ <entry>
+ <command>CLUSTER</command> is currently writing the new heap.
+ </entry>
+ </row>
+ <row>
<entry><literal>swapping relation files</literal></entry>
<entry>
The command is currently swapping newly-built files into place.
On Sat, Oct 26, 2019 at 05:13:49AM +0000, Shinoda, Noriyoshi (PN Japan A&PS Delivery) wrote:
The attached patch adds a description of the 'writing new heap'
value output in the 'phase' column.
Indeed, fixed. Thanks for the patch.
--
Michael
At Sat, 26 Oct 2019 05:13:49 +0000, "Shinoda, Noriyoshi (PN Japan A&PS Delivery)" <noriyoshi.shinoda@hpe.com> wrote in
I found a missing column value in the pg_stat_progress_cluster view document.
I read the src/backend/catalog/system_views.sql file, there seems to be a possibility that 'writing new heap' is output in the 'phase' column.
The attached patch adds a description of the 'writing new heap' value output in the 'phase' column.
Good catch!
By the way the table mentions the phases common to CLUSTER and VACUUM FULL. I wonder why some of them are described as "CLUSTER is" and others are "The command is"..
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
On Mon, Oct 28, 2019 at 02:26:39PM +0900, Kyotaro Horiguchi wrote:
By the way the table mentions the phases common to CLUSTER and
VACUUM FULL. I wonder why some of them are described as "CLUSTER is"
and others are "The command is"..
Because VACUUM FULL does not use the sort-and-scan mode, no?
--
Michael
Thank you for your response.
By the way the table mentions the phases common to CLUSTER and VACUUM FULL. I wonder why some of them are described as "CLUSTER is" and others are "The command is"..
The 'writing new heap' phase seems to appear only when the CLUSTER statement is executed. When I read the table_relation_copy_for_cluster function, it seems to be a phase that is executed only during sorting.
Regards,
Noriyoshi Shinoda
-----Original Message-----
From: Kyotaro Horiguchi [mailto:horikyota.ntt@gmail.com]
Sent: Monday, October 28, 2019 2:27 PM
To: Shinoda, Noriyoshi (PN Japan A&PS Delivery) <noriyoshi.shinoda@hpe.com>
Cc: pgsql-hackers@lists.postgresql.org
Subject: Re: [DOC] Fix for the missing pg_stat_progress_cluster view phase column value
At Sat, 26 Oct 2019 05:13:49 +0000, "Shinoda, Noriyoshi (PN Japan A&PS Delivery)" <noriyoshi.shinoda@hpe.com> wrote in
I found a missing column value in the pg_stat_progress_cluster view document.
I read the src/backend/catalog/system_views.sql file, there seems to be a possibility that 'writing new heap' is output in the 'phase' column.
The attached patch adds a description of the 'writing new heap' value output in the 'phase' column.
Good catch!
By the way the table mentions the phases common to CLUSTER and VACUUM FULL. I wonder why some of them are described as "CLUSTER is" and others are "The command is"..
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center