Re: speed up a logical replica setup

Hi,

> Thanks for the feedback. I'm attaching v26 that addresses most of your comments
> and some issues pointed by Vignesh [1].

I have created a top-up patch v27-0004. It contains additional test
cases for pg_createsubscriber.

Currently, two testcases (in v27-0004 patch) are failing. These
failures are related to running pg_createsubscriber on nodes in
cascade physical replication and are already reported in [1] and [2].
I think these cases should be fixed. Thoughts?

The idea of this patch is to keep track of testcases, so that any
future patch does not break any scenario which has already been worked
on. These testcases can be used to test in our development process,
but which test should actually be committed, can be discussed later.
Thought?

[1]: https://www.postgresql.org/message-id/CAHv8Rj+5mzK9Jt+7ECogJzfm5czvDCCd5jO1_rCx0bTEYpBE5g@mail.gmail.com
[2]:
https://www.postgresql.org/message-id/CAA4eK1Kq8qWiBK1-ky%2BjkuJRedoWLh5%3DVOmd%2BYwh9L8PUxdq%2BQ%40mail.gmail.com



Thanks and regards,
Shlok Kyal