freezing tuples ( was: Why is vacuum_freeze_min_age 100m? )

[ moving to -hackers ]

If this topic has been discussed previously, please point me to the
earlier threads.

Why aren't we more opportunistic about freezing tuples? For instance, if
we already have a dirty buffer in cache, we should be more aggressive
about freezing those tuples than freezing tuples on disk.

I looked at the code, and it looks like if we freeze one tuple on the
page during VACUUM, we mark it dirty. Wouldn't that be a good
opportunity to freeze all the other tuples on the page that we can?

Or, perhaps when the bgwriter is flushing dirty buffers, it can look for
opportunities to set hint bits or freeze tuples.

Regards,
Jeff Davis