On 10/8/20 12:32 PM, Domen Šetar wrote:
Hi guys,
I have following problem: I use POWA to collect performance data from 20 postgresql servers. Vaccum doesn't clean the largest table after POWA database is growing large. Here are data taken at 19:27:
relname | last_vacuum | last_autovacuum | n_tup | dead_tup | av_threshold | expect_av
--------------------------------------------+-------------+------------------+----------------+----------------+----------------+-----------
powa_all_relations_history_current | | 2020-10-08 18:24 | 162,254,592 | 82,045,219 | 162,280 | *
Can you please advise me how to tune poasgresql vacuum parameters to get this table cleaned? All other tables are vacummed as expected.
Is there an open transaction causing the vacuum to not free thee dead tuples?
--
Angular momentum makes the world go 'round.