Re: [PATCH v1] Add and report the new "in_hot_standby" GUC pseudo-variable.

On Wed, Mar 22, 2017 at 9:00 PM, Peter Eisentraut <
peter(dot)eisentraut(at)2ndquadrant(dot)com> wrote:

> On 3/22/17 14:09, Robert Haas wrote:
> >> The opposite means primary. I can flip the GUC name to "is_primary", if
> >> that's clearer.
> > Hmm, I don't find that clearer. "hot standby" has a very specific
> > meaning; "primary" isn't vague, but I would say it's less specific.
>
> The problem I have is that there is already a GUC named "hot_standby",
> which determines whether an instance is in hot (as opposed to warm?)
> mode if it is a standby. This is proposing to add a setting named
> "in_hot_standby" which says nothing about the hotness, but something
> about the standbyness. Note that these are all in the same namespace.
>
> I think we could use "in_recovery", which would be consistent with
> existing naming.
>

One thing we might want to consider around this -- in 10 we have
target_session_attrs=read-write (since
721f7bd3cbccaf8c07cad2707826b83f84694832), which will issue a SHOW
transaction_read_only on the connection.

We should probably consider if there is some way we can implement these two
things the same way. If we're inventing a new variable that gets pushed on
each connection, perhaps we can use that one and avoid the SHOW command? Is
the read-write thing really interesting in the aspect of the general case,
or is it more about detectinv readonly standbys as well? Or to flip that,
would sending the transaction_read_only parameter be enough for the usecase
in this thread, without having to invent a new variable?

(I haven't thought it through all the way, but figured I should mention the
thought as I'm working through my email backlog.)

--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/