Re: [PATCH]make pg_rewind to not copy useless WAL files

From: chenhj <chjischj(at)163(dot)com>
To: "Alexander Korotkov" <a(dot)korotkov(at)postgrespro(dot)ru>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: [PATCH]make pg_rewind to not copy useless WAL files
Date: 2017-09-29 16:53:31
Message-ID: a6023f8.ab46.15ece8eb1c0.Coremail.chjischj@163.com
Views: Raw Message | Whole Thread | Download mbox
Thread:
Lists: pgsql-hackers

On 2017-09-29 19:29:40,"Alexander Korotkov" <a(dot)korotkov(at)postgrespro(dot)ru> wrote:

On Fri, Sep 29, 2017 at 10:07 AM, chenhj <chjischj(at)163(dot)com> wrote:

OK. That makes sense. Thank you for the explanation.

I still have some minor comments.

/*
+ * Save the WAL filenames of the divergence and the current WAL insert
+ * location of the source server. Later only the WAL files between those
+ * would be copied to the target data directory.

Comment is outdated. We don't save filenames anymore, now we save segment numbers.

+ * Note:The later generated WAL files in the source server before the end
+ * of the copy of the data files must be made available when the target
+ * server is started. This can be done by configuring the target server as
+ * a standby of the source server.
+ */

You miss space after "Note:". Also, it seems reasonable for me to leave empty line before "Note:".

# Setup parameter for WAL reclaim

Parameter*s*, because you're setting up multiple of them.

# The accuracy of imodification from pg_ls_waldir() is seconds, so sleep one seconds

One second without "s".

Also, please check empty lines in 006_wal_copy.pl to be just empty lines without tabs.

Thanks for your comments, i had fix above problems.
And also add several line breaks at long line in 006_wal_copy.pl
Please check this patch again.

------
Best Regards
Chen Huajun

Attachment Content-Type Size
pg_rewind_wal_copy_reduce_v6.patch application/octet-stream 9.5 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2017-09-29 16:53:39 Re: [POC] hash partitioning
Previous Message Robert Haas 2017-09-29 16:46:23 Re: Shaky coding for vacuuming partitioned relations