Thread: Correction for replication slot creation error message in 9.6
Hi
Currently pg_create_physical_replication_slot() may refer to
the deprecated wal_level setting "archive":
postgres=# SHOW wal_level ;
wal_level
-----------
minimal
(1 row)
postgres=# SELECT pg_create_physical_replication_slot('some_slot');
ERROR: replication slots can only be used if wal_level is >= archive
Patch changes the error message to:
ERROR: replication slots can only be used if wal_level is "replica" or "logical"
Explicitly naming the valid WAL levels matches the wording in the wal_level
error hint used in a couple of places, i.e.
"wal_level must be set to "replica" or "logical" at server start."
Regards
Ian Barwick
--
Ian Barwick http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
Attachment
On Thu, Mar 31, 2016 at 10:15 AM, Ian Barwick <ian@2ndquadrant.com> wrote: > Currently pg_create_physical_replication_slot() may refer to > the deprecated wal_level setting "archive": (Adding Peter in CC who committed this patch). > Patch changes the error message to: > > ERROR: replication slots can only be used if wal_level is "replica" or "logical" Sounds right to me. > Explicitly naming the valid WAL levels matches the wording in the wal_level > error hint used in a couple of places, i.e. > > "wal_level must be set to "replica" or "logical" at server start." It is worth telling that Peter and I both had this code in front of our eyes during the review :) Still we missed that. -- Michael
On 2016-03-31 10:15:21 +0900, Ian Barwick wrote:
> Hi
>
> Currently pg_create_physical_replication_slot() may refer to
> the deprecated wal_level setting "archive":
>
> postgres=# SHOW wal_level ;
> wal_level
> -----------
> minimal
> (1 row)
>
> postgres=# SELECT pg_create_physical_replication_slot('some_slot');
> ERROR: replication slots can only be used if wal_level is >= archive
>
> Patch changes the error message to:
>
> ERROR: replication slots can only be used if wal_level is "replica" or "logical"
>
> Explicitly naming the valid WAL levels matches the wording in the wal_level
> error hint used in a couple of places, i.e.
The explicit naming makes it much more verbose to change anything around
wal level though, so consider me not a fan of spelling out all levels.
Greetings,
Andres Freund
Andres Freund wrote: > On 2016-03-31 10:15:21 +0900, Ian Barwick wrote: > > Patch changes the error message to: > > > > ERROR: replication slots can only be used if wal_level is "replica" or "logical" > > > > Explicitly naming the valid WAL levels matches the wording in the wal_level > > error hint used in a couple of places, i.e. > > The explicit naming makes it much more verbose to change anything around > wal level though, so consider me not a fan of spelling out all levels. I thought we had agreed that we weren't going to consider the wal_level values as a linear scale -- in other words, wordings such as "greater than FOO" are discouraged. That's always seemed a bit odd to me. -- Álvaro Herrera http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
On Thu, Mar 31, 2016 at 11:18 PM, Alvaro Herrera <alvherre@2ndquadrant.com> wrote: > Andres Freund wrote: >> On 2016-03-31 10:15:21 +0900, Ian Barwick wrote: > >> > Patch changes the error message to: >> > >> > ERROR: replication slots can only be used if wal_level is "replica" or "logical" >> > >> > Explicitly naming the valid WAL levels matches the wording in the wal_level >> > error hint used in a couple of places, i.e. >> >> The explicit naming makes it much more verbose to change anything around >> wal level though, so consider me not a fan of spelling out all levels. > > I thought we had agreed that we weren't going to consider the wal_level > values as a linear scale -- in other words, wordings such as "greater > than FOO" are discouraged. That's always seemed a bit odd to me. Yes, that's what I thought as well. -- Michael
On 16/04/01 8:15, Michael Paquier wrote: > On Thu, Mar 31, 2016 at 11:18 PM, Alvaro Herrera > <alvherre@2ndquadrant.com> wrote: >> Andres Freund wrote: >>> On 2016-03-31 10:15:21 +0900, Ian Barwick wrote: >> >>>> Patch changes the error message to: >>>> >>>> ERROR: replication slots can only be used if wal_level is "replica" or "logical" >>>> >>>> Explicitly naming the valid WAL levels matches the wording in the wal_level >>>> error hint used in a couple of places, i.e. >>> >>> The explicit naming makes it much more verbose to change anything around >>> wal level though, so consider me not a fan of spelling out all levels. >> >> I thought we had agreed that we weren't going to consider the wal_level >> values as a linear scale -- in other words, wordings such as "greater >> than FOO" are discouraged. That's always seemed a bit odd to me. > > Yes, that's what I thought as well. I don't remember if I saw that particular discussion, but same here. I suppose the alternative would be something like this: ERROR: replication slots cannot be used if wal_level is "minimal" (providing it remains the only "sub-replica" WAL level ;) ). Regards Ian Barwick -- Ian Barwick http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, RemoteDBA, Training & Services
On 03/30/2016 09:15 PM, Ian Barwick wrote: > Currently pg_create_physical_replication_slot() may refer to > the deprecated wal_level setting "archive": I have fixed this in the most direct way, since there was some disagreement about rewording.
On 03/31/2016 10:18 AM, Alvaro Herrera wrote: > I thought we had agreed that we weren't going to consider the wal_level > values as a linear scale -- in other words, wordings such as "greater > than FOO" are discouraged. That's always seemed a bit odd to me. I don't think there was any agreement about that.
On 05/04/16 10:24, Peter Eisentraut wrote: > On 03/30/2016 09:15 PM, Ian Barwick wrote: >> Currently pg_create_physical_replication_slot() may refer to >> the deprecated wal_level setting "archive": > > I have fixed this in the most direct way, since there was some disagreement about rewording. Thanks! Regards Ian Barwick -- Ian Barwick http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training & Services