Önder Kalacı

I'm slowly ramping up on this, so please let me know if the following makes sense/correct or not: #### 1) > Right now the only way to run pg_auto_ctl is...

> There are messages like this (with host=localhost): For future reference: My installation/configuration might be different, but I couldn't see `localhost` in the output on my Mac: ``` MacBook-Pro-Onders:pg_auto_failover onderkalaci$...

> Add --pghost flag to create monitor so you can override the PGHOST variable. We seem to have ` create monitor --nodename` parameter. Do you have something different in mind...

> I think we should definitely have some sane defaults for pg_autoctl create postgres too. The idea is follows: The monitor is an embedded database to our application, and we're...

> agree that that's the case when the database already exists. But if we create it we should not open it up by default when pg_autoctl does not need it....

> ignore PGHOST and skip implementing --pghost on the monitor, I wonder if this is a good idea as we seem to have a dependency in the regression tests for...

This GUC is dropped on Citus 11, so probably not much relevant any more. The PR: https://github.com/citusdata/citus/pull/5444

> f I understand it correctly, it's an alias for citus_add_node, right? Yes, we are deprecating APIs starting with `master_`, so suggest using `citus_add_node(..., groupid:=0)` or `citus_set_coordinator_host`: https://docs.citusdata.com/en/v11.0/develop/api_udf.html#set-coordinator-host I agree...

Just FYI: An example query/use-case we've been asked on github: https://github.com/citusdata/citus/issues/2321 Note that the original query that we need to rewrite is not a subquery, but a repartition outer join...

Maybe we should rename the function to `create_colocated_function` or such