Re: Proposal: pg_confcheck - syntactic & semantic validation of postgresql configuration files

On Fri, Oct 9, 2015 at 4:38 PM, Amir Rohan <amir(dot)rohan(at)zoho(dot)com> wrote:
> It does catch bad syntax, but in most cases all you get is
> "The setting could not be applied". that's not great for enums
> or a float instead of an int. I guess a future version will fix that
> (or not).

I expect we would consider patches to improve the error messages if
you (or someone else) wanted to propose such. But you don't have to
want to do that.

> You need a running server to run a check. You need to monkey
> with said server's configuration in place to run a check. You must be on
> 9.5+. The checking mechanism isn't extensible. Certainly not as easily
> as dropping a new rule file somewhere. It doesn't check (AFAICT) for bad
> combinations of values, for example it will tell you that you can't
> change `wal_archive` without restart (without showing source location
> btw, bug?), but not that you better set `wal_level` *before* you
> restart. It doesn't do any semantic checks. It won't warn you
> about things that are not actually an error, just a bad idea.

So, I'm not saying that a config checker has no value. In fact, I
already said the opposite. You seem to be jumping all over me here
when all I was trying to do is explain what I think Tom was getting
at. I *do* think that pg_file_settings is a helpful feature that is
certainly related to what you are trying to do, but I don't think that
it means that a config checker is useless. Fair?

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