Re: Bloated toast table with empty associated table

2025-04-05 Thread Ron Johnson
On Thu, Mar 20, 2025 at 7:40 AM Paul Allen wrote: > Hello. > > Preconditions. > > I have some empty table and constantly try to execute `insert ... on > conflict do update ...` on it. My data in row which I try to insert is > invalid by violation of foreing key constraint, so I am getting error >

Bloated toast table with empty associated table

2025-04-04 Thread Paul Allen
Hello. Preconditions. I have some empty table and constantly try to execute `insert ... on conflict do update ...` on it. My data in row which I try to insert is invalid by violation of foreing key constraint, so I am getting error while inserting and table keeps being empty. This table have some

Re: Bloated toast table with empty associated table

2025-03-24 Thread Adrian Klaver
On 3/20/25 04:39, Paul Allen wrote: Hello. Problem. My backend application attempts unsuccessfully repeatedly to insert the same ~100 rows with images, and despite table's row count remains 0, toast table's size is growing up permanently, reaching 100, 200, 300 GB until it takes all availab

Re: Bloated toast table with empty associated table

2025-03-20 Thread Paul Allen
> It's the general principle that the _scale_factor defaults are in my > experience too high. Sorry, didn't mentioned Ok, thanks, I will use the cron task. It's toast behaviour still seems odd to me. Is there a way to prevent this behaviour at all, to not store rows which were failed to insert?

Re: Bloated toast table with empty associated table

2025-03-20 Thread Ron Johnson
On Thu, Mar 20, 2025 at 8:45 AM Paul Allen wrote: > > It's the general principle that the _scale_factor defaults are in my > experience too high. > Sorry, didn't mentioned > > Ok, thanks, I will use the cron task. > > It's toast behaviour still seems odd to me. Is there a way to prevent > this be

Re: Bloated toast table with empty associated table

2025-03-20 Thread Ron Johnson
On Thu, Mar 20, 2025 at 8:22 AM Paul Allen wrote: > > Instantaneously and repeatedly, while ignoring the error? > Yes, that's how it should work and I probably can't do anything about it. > > > Your _scale_factor values are too high. Drop them down to about 5%. > > Okay, but what about altering

Re: Bloated toast table with empty associated table

2025-03-20 Thread Paul Allen
> Instantaneously and repeatedly, while ignoring the error? Yes, that's how it should work and I probably can't do anything about it. > Your _scale_factor values are too high. Drop them down to about 5%. Okay, but what about altering controlzone_passage table, where I set all _scale_factor value