Re: How to store "blobs" efficiently for small and large sizes, with random access

On 10/19/22 04:47, Dominique Devienne wrote:
[snip]
> PS: Another drawback of lo is that because it's a single table, it
> still subject to the 32 TB limit on a relation.
>    The 4TB limit per lo is way more than we need, but the 32 TB limit
> may actually be more of an issue for
>    our larger clients, which have thousands of projects, each with
> upwards of a few millions of those blobs.

Partition the table on the first segment of the Primary Key.  From 
experience, anything else can lead to serious query degradation.

-- 
Angular momentum makes the world go 'round.