On 30-12-2011 17:40, Tom Lane wrote:
> Miguel Silva<miguel.silva@tactis.pt> writes:
>> But with postgresql version 9.0.5 (in version 9.0.4 all was fine), we
>> noticed the program was taking longer to start. In fact, in some clients
>> that had older hardware, it could take around 20 minutes when it usually
>> takes only a few seconds. To make a long story short, the problem was
>> traced and narrowed down to a single auto generated query. Here it is:
>> "SELECT [ snip ]"
>> ... Here are the explain analyzes.
>> enable_material ON: http://explain.depesz.com/s/wen
>> enable_material OFF: http://explain.depesz.com/s/Zaa
> It doesn't really accomplish anything to post anonymized explain output
> when you've already shown us the actual query, does it? Especially when
> said query involves only the system catalogs and by no stretch of the
> imagination could be thought to contain anything proprietary?
>
> regards, tom lane
>
Indeed you are right. Those are explains I created some time ago, when I
didn't really know what that webpage did. I just kept them, and used
them now, didn't even think about that. But the explains are still
there, still useful. Anyway, if it is really necessary, I can post new ones.
Best regards,
Miguel Silva