Message-ID: <1520845041.2810.3.camel@cybertec.at>
Subject: Re: streaming replication and wal 's usages
From: Laurenz Albe <laurenz.albe@cybertec.at>
To: ghiureai <isabella.ghiurea@nrc-cnrc.gc.ca>, "pgsql-admin@postgresql.org"
	 <pgsql-admin@postgresql.org>
Date: Mon, 12 Mar 2018 09:57:21 +0100
In-Reply-To: <63b50fbd-4822-8a36-4fea-2dc473cc1242@nrc-cnrc.gc.ca>
References: <63b50fbd-4822-8a36-4fea-2dc473cc1242@nrc-cnrc.gc.ca>
Organization: Cybertec
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

ghiureai wrote:
> I am running PG 10.2 with streaming replication, no archiving option 
> with 10 slots replication option , I had the slave down for 6 hours  
> which generated 56GB of wal files in pg_wal dir,  after slave was brougt 
> online and catchup with master , the wal dir was still growing on slave 
> with more than 25GB where in archive_status where all the files where in 
> *.done files , eventually the /wal directory reached 100% but still no 
> wal files removed from archive_status
> 
> I am trying to understand what I am missing in this confg ( the master 
> and slave are cfg for pg_dump backups the switch-over is done with 
> pg_basebackup using repmgr4.2.last release )
> 
> the present db size 140GB OLTP type transactions , with an avg 10-12GB 
> new data  growth per day.
> 
> Here is my cfg
> 
> wal_keep_segments = 300
> max_replication_slots=10
> wal_log_hints = on
> checkpoint_timeout = 5min
> max_wal_size = 1GB
> min_wal_size = 80MB
> archive_mode=off
> 
> hot_standby =on

You are not using replication slots, right?

Check pg_replication_slots to see if you have abandoned replication slots;
that could cause old WAL to be retained.

Yours,
Laurenz Albe
-- 
Cybertec | https://www.cybertec-postgresql.com