On Fri, Sep 26, 2014 at 1:22 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
>> If we want the narrowest possible fix for this, I think it's "complain
>> if a non-zero value would round to zero". That fixes the original
>> complaint and changes absolutely nothing else. But I think that's
>> kind of wussy. Yeah, rounding 29 seconds down to a special magic
>> value of 0 is more surprising than rounding 30 seconds up to a minute,
>> but the latter is still surprising. We're generally not averse to
>> tighter validation, so why here?
>
> So in other words, if I set "shared_buffers = 100KB", you are proposing
> that that be rejected because it's not an exact multiple of 8KB?
Absolutely. And if anyone is inconvenienced by that, then they should
upgrade to a 386. Seriously, who is going to set a value of
shared_buffers that is not measured in MB? And if they do, why
shouldn't we complain if we can't honor the value exactly? If they
really put in a value that small, it's not stupid to think that the
difference between 96kB and 104kB is significant. Honestly, the most
likely explanation for that value is that it's a developer doing
testing.
> I apparently have got to make this point one more time: if the user
> cares about the difference between 30sec and 1min, then we erred in
> designing the GUC in question; it should have had a smaller unit.
> I am completely not impressed by arguments based on such cases.
> The right fix for such a case is to choose a different unit for the GUC.
The guy who wrote the GUC system doesn't seem to agree with you, and
neither do I.
Changing the unit for log_rotation_age from minutes to seconds doesn't
fix the original complaint, which was that a non-zero value like 1s
gets rounded down to zero. Even after you change the unit to seconds,
you still have the same problem with a specification of 1ms. You
could change the log rotation unit to 1ms, but that's silly, and as
soon as we get a unit measured in microseconds or nanoseconds, which
seems like just a matter of time, you have the same problem again.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company