Re: pgsql: Generate backup manifests for base backups, and validate them.

From: David Steele <david(at)pgmasters(dot)net>
To: Robert Haas <robertmhaas(at)gmail(dot)com>, Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com>
Cc: pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: pgsql: Generate backup manifests for base backups, and validate them.
Date: 2020-04-07 17:13:24
Message-ID: ebd9b786-c5b4-b6e1-f3ee-25b02b7706e9@pgmasters.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers pgsql-hackers

On 4/7/20 12:44 PM, Robert Haas wrote:
> On Tue, Apr 7, 2020 at 5:51 AM Peter Eisentraut
> <peter(dot)eisentraut(at)2ndquadrant(dot)com> wrote:
>> On 2020-04-03 21:07, Robert Haas wrote:
>>> A new tool called pg_validatebackup can validate a backup against the
>>> manifest.
>>
>> In software engineering, "verify" and "validate" have standardized
>> distinct meanings. I'm not going to try to explain them here, but you
>> can easily find them online. I haven't formed an opinion on which one
>> of them this tool is doing, but I notice that both the man page and the
>> messages produced by the tool use the two terms seemingly
>> interchangeably. We should try to pick the correct term and use it
>> consistently.
>
> The tool is trying to make sure that we have the same backup that
> we're supposed to have, and that the associated WAL is present and
> sane. Looking at
> https://en.wikipedia.org/wiki/Verification_and_validation, that sounds
> more like verification than validation, but I confess that this
> distinction is new to me.

When I searched I found a two different definitions for validation and
verification. One for software development (as in the link above and
what I think Peter meant) and another for data (see
https://en.wikipedia.org/wiki/Data_validation,
https://en.wikipedia.org/wiki/Data_verification,
https://www.differencebetween.com/difference-between-data-validation-and-vs-data-verification/)

It seems that validation vs. verify as defined in PMBOK (the former
sense) does not really apply here, though. That leaves only the latter
sense which appears less well-documented but points to "verify" as the
better option.

Regards,
--
-David
david(at)pgmasters(dot)net

In response to

Browse pgsql-committers by date

  From Date Subject
Next Message Tomas Vondra 2020-04-07 17:19:18 Re: pgsql: Collect statistics about SLRU caches
Previous Message Thomas Munro 2020-04-07 16:54:40 Re: pgsql: Collect statistics about SLRU caches

Browse pgsql-hackers by date

  From Date Subject
Next Message Etsuro Fujita 2020-04-07 17:24:28 Re: [HACKERS] advanced partition matching algorithm for partition-wise join
Previous Message Jeevan Chalke 2020-04-07 17:02:55 Re: WIP/PoC for parallel backup