Re: request for new parameter for disable promote (slave only mode)

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Ioseph Kim <pgsql-kr(at)postgresql(dot)kr>
Cc: "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: request for new parameter for disable promote (slave only mode)
Date: 2018-07-30 17:10:58
Message-ID: CA+TgmobfQSGCYhqGUQu20QtmXAJES=611qG8JU0onZtSBvfmeA@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Fri, Jul 27, 2018 at 12:05 PM, Ioseph Kim <pgsql-kr(at)postgresql(dot)kr> wrote:
> I want to build one master & multi slave environments to use physical replication.
> Slave nodes have low hardware spec, so I changed max_connection server parameters, and try start slave node.
> But I could not start slave nodes,
> because CheckRequiredParameterValues function (in src/backend/access/transam/xlog.c) reject to set less values then master’s values.
>
> It is maybe,
> This concept is base on some slave node can be promte master.

No, it's because the machinery that computes MVCC snapshots on slaves
relies on max_connections being set high enough.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Alexander Kuzmenkov 2018-07-30 17:13:21 Re: [HACKERS] PoC: full merge join on comparison clause
Previous Message Robert Haas 2018-07-30 17:05:43 Re: [HACKERS] Parallel Append implementation