On Wed, Dec 30, 2009 at 10:24 AM, Alvaro Herrera
<alvherre@commandprompt.com> wrote:
> Tom Lane escribió:
>> Robert Haas <robertmhaas@gmail.com> writes:
>> > If you'd be willing to look over the latest version of my
>> > per-tablespace random_page_cost/seq_page_cost patch, which I posted to
>> > -hackers some time in the last few days, I can get that committed and
>> > then start working on this, if you'd like.
>>
>> I think Alvaro would actually be the right person to review that,
>> since the reloptions code is almost entirely his work.
>
> I can't promise anything right now though, as my wife could get with
> labour very soon ...
In terms of reasons for not being able to guarantee anything, I'd have
to say that's one of the best I've heard.
In all honesty, I'm not very worried about the reloptions stuff
proper. I have copied the existing coding pattern so closely that
it's a little hard to imagine that I've broken anything too badly. My
main concerns are:
1. Am I leaking memory anywhere?
and
2. Can anything bad happen as a result of invalidation events and/or
concurrent updates to pg_tablespace?
If anyone feels qualified to check my work on those two points, that
would be great. In reality, even if I've done something relatively
stupid, it isn't likely to have much practical impact since
pg_tablespace updates figure to be infrequent and many people won't
use this feature at all. But I'd still rather not do something
stupid.
Thanks,
...Robert