Re: Internal key management system - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: Internal key management system
Date
Msg-id 20201028014314.GA16897@momjian.us
Whole thread Raw
In response to Re: Internal key management system  (Bruce Momjian <bruce@momjian.us>)
Responses Re: Internal key management system
List pgsql-hackers
On Tue, Oct 27, 2020 at 10:20:35AM -0400, Bruce Momjian wrote:
> I don't know much about how to hook into that stuff so if you have an
> idea, I am all ears.  I have used OpenSSL with Yubikey via pksc11.  You
> can see the use of it on slide 57 and following:
> 
>     https://momjian.us/main/writings/crypto_hw_config.pdf#page=57
> 
> Interestingly, that still needed the user to type in a key to unlock the
> Yubikey, so we might need PKCS11 and a password for the same server
> start.
> 
> I would like to get this moving forward so I will work on the idea of
> passing an open /dev/tty file descriptor from pg_ctl to the postmaster
> on start.

Here is an updated patch that uses an argument to pass an open /dev/tty
file descriptor to the postmaster.  It uses -R for initdb/pg_ctl, -R ###
for postmaster/postgres, and %R for cluster_passphrase_command.  Here is
a sample session:

-->    $ initdb -R --cluster-passphrase-command '/tmp/pass_fd.sh "%p" %R'
    The files belonging to this database system will be owned by user "postgres".
    This user must also own the server process.
    
    The database cluster will be initialized with locale "en_US.UTF-8".
    The default database encoding has accordingly been set to "UTF8".
    The default text search configuration will be set to "english".
    
    Data page checksums are disabled.
    Key management system is enabled.
    
    fixing permissions on existing directory /u/pgsql/data ... ok
    creating subdirectories ... ok
    selecting dynamic shared memory implementation ... posix
    selecting default max_connections ... 100
    selecting default shared_buffers ... 128MB
    selecting default time zone ... America/New_York
    creating configuration files ... ok
    running bootstrap script ...
-->    Enter database encryption pass phrase: B1D7B405EDCD97B7351DD3B7AE0637775FFBC6A2C2EEADAEC009A75A58A79F50
    ok
    performing post-bootstrap initialization ...
-->    Enter database encryption pass phrase: B1D7B405EDCD97B7351DD3B7AE0637775FFBC6A2C2EEADAEC009A75A58A79F50
    ok
    syncing data to disk ... ok
    
    initdb: warning: enabling "trust" authentication for local connections
    You can change this by editing pg_hba.conf or using the option -A, or
    --auth-local and --auth-host, the next time you run initdb.
    
    Success. You can now start the database server using:
    
        pg_ctl -D /u/pgsql/data -l logfile start
    
    $ pg_ctl stop
    pg_ctl: PID file "/u/pgsql/data/postmaster.pid" does not exist
    Is server running?
-->    $ pg_ctl -l /u/pg/server.log -R start
    waiting for server to start...
-->    Enter database encryption pass phrase: B1D7B405EDCD97B7351DD3B7AE0637775FFBC6A2C2EEADAEC009A75A58A79F50
     done
    server started

Attached is my updated patch, based on Masahiko Sawada's patch, and my
pass_fd.sh script.  

-- 
  Bruce Momjian  <bruce@momjian.us>        https://momjian.us
  EnterpriseDB                             https://enterprisedb.com

  The usefulness of a cup is in its emptiness, Bruce Lee


Attachment

pgsql-hackers by date:

Previous
From: Tomas Vondra
Date:
Subject: Re: MultiXact\SLRU buffers configuration
Next
From: Tomas Vondra
Date:
Subject: Re: Patch to fix FK-related selectivity estimates with constants