Postgresql – libreadline.so.6

The error is harmless but to get rid of it, I think you need to break this restore into two commands, as in:

dropdb -U postgres mydb && \
 pg_restore --create --dbname=postgres --username=postgres pg_backup.dump

The --clean option in pg_restore doesn't look like much but actually raises non-trivial problems.

For versions up to 9.1

The combination of --create and --clean in pg_restore options used to be an error in older PG versions (up to 9.1). There is indeed some contradiction between (quoting the 9.1 manpage):

--clean Clean (drop) database objects before recreating them

and

--create Create the database before restoring into it.

Because what's the point of cleaning inside a brand-new database?

Starting from version 9.2

The combination is now accepted and the doc says this (quoting the 9.3 manpage):

--clean Clean (drop) database objects before recreating them. (This might generate some harmless error messages, if any objects were not present in the destination database.)

--create Create the database before restoring into it. If --clean is also specified, drop and recreate the target database before connecting to it.

Now having both together leads to this kind of sequence during your restore:

DROP DATABASE mydb;
...
CREATE DATABASE mydb WITH TEMPLATE = template0... [other options]
...
CREATE SCHEMA public;
...
CREATE TABLE...

There is no DROP for each individual object, only a DROP DATABASE at the beginning. If not using --create this would be the opposite.

Anyway this sequence raises the error of public schema already existing because creating mydb from template0 has imported it already (which is normal, it's the point of a template database).

I'm not sure why this case is not handled automatically by pg_restore. Maybe this would cause undesirable side-effects when an admin decides to customize template0 and/or change the purpose of public, even if we're not supposed to do that.

After asking on pgsql-performance list, Jeff Janes figured out that the cause was associated to the default collation used by Postgres (see this link for more informations). MacMini was using the much performing collation while Dell T420 was using the en/US collation.

T420 (Postgres 9.4.1)

                                  List of databases
   Name    |  Owner   | Encoding |   Collate   |    Ctype    |   Access privileges   
-----------+----------+----------+-------------+-------------+-----------------------
 pen       | pen      | UTF8     | en_US.UTF-8 | en_US.UTF-8 | 
 postgres  | postgres | UTF8     | en_US.UTF-8 | en_US.UTF-8 | 
 template0 | postgres | UTF8     | en_US.UTF-8 | en_US.UTF-8 | =c/postgres          +
           |          |          |             |             | postgres=CTc/postgres
 template1 | postgres | UTF8     | en_US.UTF-8 | en_US.UTF-8 | =c/postgres          +
           |          |          |             |             | postgres=CTc/postgres
(4 rows)

MacMini (Postgres 9.0.13)

                                    List of databases
       Name        |   Owner    | Encoding | Collation | Ctype |    Access privileges    
-------------------+------------+----------+-----------+-------+-------------------------
 caldav            | caldav     | UTF8     | C         | C     | 
 collab            | collab     | UTF8     | C         | C     | 
 device_management | _devicemgr | UTF8     | C         | C     | 
 pen               | pen        | UTF8     | C         | C     | 
 postgres          | _postgres  | UTF8     | C         | C     | 
 roundcubemail     | roundcube  | UTF8     | C         | C     | 
 template0         | _postgres  | UTF8     | C         | C     | =c/_postgres           +
                   |            |          |           |       | _postgres=CTc/_postgres
 template1         | _postgres  | UTF8     | C         | C     | =c/_postgres           +
                   |            |          |           |       | _postgres=CTc/_postgres
(8 rows)

After setting collation on T420 "C" the A transaction went from 195 seconds to 33 seconds against 40 seconds on Mac Mini; B type transaction went from 141 seconds to 78 seconds against 101 seconds on Mac Mini. This is the best performance improvement after modifing BIOS settings. Many kernel adjustments didn't provide significant improvements.

So, running the following command will initialize a new database with collation C and encoding UTF8:

/usr/local/pgsql/bin/initdb -D /path/to/your/data --no-locale --encoding=UTF8

I hope this post will help other people in the future. After reading about a lot of digressions about kernel, virtual memory, RAID controllers, disk cache, WAL and other tech stuff I never found someone talking about collations.