100/s is 3 billion per year. Postgres has a hard limit of 4 billion blobs in a column.
that ignores the reality that RDS storage is comparatively expensive.
If you think a table might get anywhere near this size (blobs or not), I highly recommend table partitioning.
My problem is customer ops, not mine so much. They run our software.
That's not true. Our ops struggle too. 32bit oids really are a barrier.
Yes, but it does blow up TOAST and has a lot of impact on the deletion behavior on busy tables. We removed all larger json blobs from PG. typical settings or config stuff in json in PG are fine. We use that all the time. But larger json blobs of several kilobytes are still an issue for semi timeseries data.
Could you elaborate on the TOAST issues you're having? We're pretty liberal with our use of large JSONB objects and might hit a billion objects in a year or so.
Have you tried using PostgreSQL Large Objects (LOs)?