Hi All!
First time I see that the sum of written buffers:
3765929+878326121 = 882092050 = select
pg_size_pretty(882092050::numeric*8192) = 6730 GB
is bigger than buffers_alloc value in pg_stat_bgwriter view:
buffers_alloc | 20426161 = 156 GB
buffers_checkpoint | 878599316
buffers_clean | 0
maxwritten_clean | 0
buffers_backend | 3766409
buffers_backend_fsync | 0
buffers_alloc | 20426161
It's kinda mysterious... Any explanations?
Does it mean that the same buffers were written over and over again?
--
Sent from: http://www.postgresql-archive.org/PostgreSQL-general-f1843780.html
On Sun, Apr 8, 2018 at 11:28 AM, pinker <pinker@onet.eu> wrote:
It's kinda mysterious... Any explanations? Does it mean that the same buffers were written over and over again?
Yeah, checkpoints will write all dirty buffers, but doesn't evict them. Next time the page is needed, it doesn't need to be re-read as it is still there.