Hi,
> I've attached a draft patch for this idea.
I think it needs a rebase after edca3424342da323499a1998d18a888283e52ac7.
Also, as discussed in [1], I think one improvement we had was to
keep IsIndexUsableForReplicaIdentityFull() in a way that it is easier to
read & better documented in the code. So, it would be nice to stick to that.
Overall, the proposed changes make sense to me as well. Once the patch is
ready, I'm happy to review & test in detail.
Thanks,
Onder
[1]
https://www.postgresql.org/message-id/CAA4eK1Jcyrxt_84wt2%3DQnOcwwJEC2et%2BtCLjAuTXzE6N3FXqQw%40mail.gmail.com