Re: Cost limited statements RFC

On Fri, Jun 7, 2013 at 12:55 PM, Greg Smith <greg(at)2ndquadrant(dot)com> wrote:
> On 6/7/13 12:42 PM, Robert Haas wrote:
>> GUCs in terms of units that are meaningful to the user. One could
>> have something like io_rate_limit (measured in MB/s),
>> io_read_multiplier = 1.0, io_dirty_multiplier = 1.0, and I think that
>> would be reasonably clear.
>
> There's one other way to frame this:
>
> io_read_limit = 7.8MB/s # Maximum read rate
> io_dirty_multiplier = 2.0 # How expensive writes are considered relative to
> reads
>
> That still gives all of the behavior I'd like to preserve, as well as not
> changing the default I/O pattern. I don't think it's too complicated to ask
> people to grapple with that pair.

That's unsatisfying to me because the io_read_limit is not really an
io_read_limit at all. It is some kind of combined limit, but the name
doesn't indicate that.

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