Subject: Re: Proposal: More structured logging
To: Ronan Dunklau <ronan.dunklau@aiven.io>, pgsql-hackers@lists.postgresql.org
References: <3207907.AWbSqkKDnR@aivenronan>
 <e9b5a975-503d-61c7-c60e-d0ad89f003ad@enterprisedb.com>
 <163276811.1OrvvjSa7g@aivenronan>
From: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
Message-ID: <03fa81f6-77b0-e7eb-b390-dfe04f180495@enterprisedb.com>
Date: Wed, 8 Sep 2021 11:51:31 +0200
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:78.0)
 Gecko/20100101 Thunderbird/78.13.0
MIME-Version: 1.0
In-Reply-To: <163276811.1OrvvjSa7g@aivenronan>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Archived-At: 
 <https://www.postgresql.org/message-id/03fa81f6-77b0-e7eb-b390-dfe04f180495%40enterprisedb.com>
Precedence: bulk

On 01.09.21 10:00, Ronan Dunklau wrote:
> In-core it would open up the possibility to split log messages into different
> fields, for example the different statistics reported in the logs by VACUUM /
> ANALYZE VERBOSE and make it easier to consume the output without having to
> parse the message. Parsing the message also means that any tool parsing it
> needs to either be aware of the locale, or to force the user to use a specific
> one.

I think those messages would themselves have substructure.  For example, 
the current message

     "I/O timings: read: %.3f ms, write: %.3f ms\n"

might be

     {"I/O timings": {"read": ..., "write": ...}}

and that in turn is already part of a larger message.

So just having a single level of tags wouldn't really work for this.