On 8/18/22 02:50, gzh wrote:
Dear Tom,
Thanks for your reply.
Please refer to the information below：
I run following sql in PostgreSQL 8.2 and PostgreSQL 12.5, it returns
different execution plan.
There are ~13 years of improvements to the planner and the database as a
whole, I would mo
But you should do that according to the documentation:
...
After restoring a backup, it is wise to run ANALYZE on each database so the
query optimizer has useful statistics.
...
On 8/17/22 20:01, gzh wrote:
Hi,
I have had a Perl Website working for 7 years and have had no problems
until a few weeks ago I replaced my database server with a newer one.
Did you run ANALYZE on the 12.5 server after restoring the data to it?
gzh
--
Adrian Klaver
adrian.kla...@akl
On Thu, 18 Aug 2022 at 15:32, Tom Lane wrote:
> The 12.5 plan looks like it thinks that the join condition is not
> hashable --- and probably not mergeable as well, else it would have
> done a mergejoin. This is odd if we assume that the lower()
> outputs are just text. But you haven't said anyt
gzh writes:
> I run following sql in PostgreSQL 8.2 and PostgreSQL 12.5, it returns
> different execution plan.
8.2 is ... well, not stone age maybe, but pretty durn ancient.
You really ought to update a bit more often than that. (And
maybe pay more attention to staying up to date with minor re
