>> client_addr does not seem the right place to store this information,
>> and it is already documented for years that NULL is used when using a
>> Unix socket. But I think that we could change *client_hostname* so as
>> the path name is reported instead of NULL when connecting through a
>> Unix domain socket, and there is no need to switch the field type for
>> that.
>
> That seems like a hack, and I think it could still break apps that
> are expecting particular semantics for that field. Why not add a
> new column instead?
Actually I aready proposed to add new column to pg_stat_get_wal_senders():
> Changing this behavior would affect existing pg_stat_get_activity view
> users and I hesitate to do so. I wonder if we could add receiver's
> UNIX domain socket path to from pg_stat_get_wal_senders() (which is
> called from pg_stat_replication view) so that the poor UNIX domain
> socket users could make their own view or access
> pg_stat_get_wal_senders() to get the UNIX domain socket path.
If we were ok to add a new column to pg_stat_activity view or
pg_stat_replication view as well, that will be great.
Best regards,
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese:http://www.sraoss.co.jp