Last night we did a vacuum full operation on partitioned tables. A few hours after the vacuum full, I found cpu rising with an unknown session called “startup” when queried by top command. We cannot conclude that this is related…
I ran a simple delete query on a telemetries table to clear up some data but cancelled as it was taking really long. I noticed that after I cancelled the query some of the tables (all partitions of the telemetries…
I don't understand why simple stored procedure or anonymous block that doesn't query any tables would prevent vacuum to remove dead rows. I connect to postgres and run the following: DO $BODY$ DECLARE v_id INT; BEGIN for v_id in 1..10…
I have a table store_record with 45 million records. I want to run a simple count query on the largest database_id. Note I have an index on database_id. SELECT COUNT(*) FROM store_record WHERE database_id='123'; -- returns ~17.2 million The query…
Is it possible to run PostgreSQL 11's VACUUM FULL for a short while and then get some benefit? Or does cancelling it midway cause all of its progress to be lost? I've read about pg_repack (https://aws.amazon.com/blogs/database/remove-bloat-from-amazon-aurora-and-rds-for-postgresql-with-pg_repack/) but the way it…