Attempt to fix unstable regression tests, take 2
authorDavid Rowley <[email protected]>
Mon, 30 Mar 2020 10:41:11 +0000 (23:41 +1300)
committerDavid Rowley <[email protected]>
Mon, 30 Mar 2020 10:41:11 +0000 (23:41 +1300)
commit24566b359d095c3800c2a326d88a595722813f58
treefff12f35f5764869cb55cb91519181b92c2be421
parent64638ccba3a659d8b8a3a4bc5b47307685a64a8a
Attempt to fix unstable regression tests, take 2

Following up on 2dc16efed, petalura has suffered some additional
failures in stats_ext which again appear to be around the timing of an
autovacuum during the test, causing instability in the row estimates.

Again, let's fix this by explicitly performing a VACUUM on the table
and not leave it to happen by chance of an autovacuum pass.

Discussion: https://postgr.es/m/CAApHDvok5hmXr%2BbUbJe7%2B2sQzWo4B_QzSk7RKFR9fP6BjYXx5g%40mail.gmail.com
src/test/regress/expected/stats_ext.out
src/test/regress/sql/stats_ext.sql