Since it is generating 60GB in few mins, my replication lag is increasing upto 10Gb for around 10 mins, which may seen as possible data loss in case of primary node failure
Regards
Sasi
On Thursday, December 12, 2024, 12:06 PM, Laurenz Albe <laurenz.albe@xxxxxxxxxxx> wrote:
On Wed, 2024-12-11 at 23:25 -0700, David G. Johnston wrote:
> On Wednesday, December 11, 2024, SASIKUMAR Devaraj <sashikumard@xxxxxxxxx> wrote:
> > When we are running vacuumdb for our database of 1.5TB it is generating approximately 60GB
> > of WAL files? Any way we can reduce this WAL file generation?
>
> Set a much more aggressive autovacuum so you don’t accumulate as much dead tuples between runs?
You can also activate "wal_compression" for smaller full-page images
and increase "max_wal_size" to get fewer of them.
But 60GB of WAL shouldn't be a problem. Why are they a problem for you?
Yours,
Laurenz Albe