On 1/29/20 1:58 PM, Perumal Raj wrote:
Hi Tom /Adrian,
Issue is not specific to a table or particular Query. Also there is no
change in DB parameter after upgrade.
You obviously are seeing some difference, so pick a query and do an
EXPLAIN ANALYZE on it. Without some actual performance data this issue
is not going to be solved.
That the only way i can make it most of the the query to run as like
before upgrade.
Note:
Some web reference says , Engine will take some time to adjust until it
runs autovacuum .
The above assumes you have not done a manual ANALYZE on the database, as
ANALYZE is done as part of autovacuum.
On Wed, Jan 29, 2020 at 10:22 AM Tom Lane <tgl@xxxxxxxxxxxxx
<mailto:tgl@xxxxxxxxxxxxx>> wrote:
Perumal Raj <perucinci@xxxxxxxxx <mailto:perucinci@xxxxxxxxx>> writes:
> We have recently upgraded postgres from 9.2 to 11.6 and started
seeing
> performance issue immediately and able to fix the performance
issue after
> disabling parameter: enable_seqscan.
> Question :
> Should i keep the above parameter always disabled ? If not why
the behavior
> changed in Higher version ?
This is unanswerable with the amount of information you've given.
Yes, turning off enable_seqscan is a bad idea in general, but why
you got a worse plan without that requires details.
https://wiki.postgresql.org/wiki/Slow_Query_Questions
regards, tom lane
--
Adrian Klaver
adrian.klaver@xxxxxxxxxxx
