Logging in via gui and running:

select * from pg_tables where schemaname='public';

does not show the table.

Running "vagrant halt", then "vagrant up" gets

==> debian: Machine already provisioned. Run vagrant provision or use the --provision
==> debian: flag to force provisioning. Provisioners marked to run always will still run.

So running "vagrant provision".

from sovereign.

The Roundcube error does not appear when running the provisioning again. But psql command just quits silently, so it's hard to verify if anything PostgreSQL is working.

from sovereign.

Nope, still happening.

NOTIFIED: [webmail | set roundcube password] **********************************
failed: [debian] => {"changed": true, "cmd": ["sudo", "-u", "postgres", "psql", "-d", "roundcube", "-c", "ALTER USER roundcube with password 'testPassword';"], "delta": "0:00:00.086619", "end": "2015-12-17 09:59:23.601677", "rc": 2, "start": "2015-12-17 09:59:23.515058", "warnings": []}
stderr: psql: FATAL: database "roundcube" does not exist

FATAL: all hosts have already failed -- aborting

TASK: [vpn | Copy dnsmasq configuration file into place] **********************
FATAL: no hosts matched or all hosts have already failed -- aborting

after which, can't ssh into the system, either.

from sovereign.

After

vagrant halt
vagrant up
vagrant provision debian

seemed to get through the remainder of provisioning.

from sovereign.

relates to an issue upstream: sovereign#417

from sovereign.

Saw this again. Sometimes it happens, sometimes it doesn't.

from sovereign.

NOTIFIED: [webmail | set roundcube password] **********************************
failed: [debian] => {"changed": true, "cmd": ["sudo", "-u", "postgres", "psql", "-d", "roundcube", "-c", "ALTER USER roundcube with password 'testPassword';"], "delta": "0:00:00.084767", "end": "2016-01-02 21:50:26.886369", "rc": 2, "start": "2016-01-02 21:50:26.801602", "warnings": []}
stderr: psql: FATAL: database "roundcube" does not exist

FATAL: all hosts have already failed -- aborting

Still happening.

from sovereign.

Still happening. Saw it on both master and jessie branches.

NOTIFIED: [webmail | set roundcube password] **********************************
failed: [debian] => {"changed": true, "cmd": ["sudo", "-u", "postgres", "psql", "-d", "roundcube", "-c", "ALTER USER roundcube with password 'testPassword';"], "delta": "0:00:00.279713", "end": "2016-02-21 20:35:03.072869", "rc": 2, "start": "2016-02-21 20:35:02.793156", "warnings": []}
stderr: psql: FATAL: database "roundcube" does not exist

FATAL: all hosts have already failed -- aborting

from sovereign.