Re: Security issues concerning pgsql replication - Mailing list pgsql-general

Perhaps you misunderstand me.

It is not user-database, but master-slave interaction that I am concerning.

The master-slave replication proceeds continually and requires no manual interference.  Both master and slave’s private key are involved, but ssl_passphrase_command is only used to parse passphrase of master’s private key. Pgsql cannot get slave’s private key automatically, so replication failed. I’ve tried and proved it did not work.

I refered to the list of pgsql’s authenticate methods but did not find an appropriate one for replication.

https://www.postgresql.org/docs/12/client-authentication.html

发件人: Magnus Hagander [mailto:magnus@hagander.net] 

发送时间: 2020年10月27日 17:00
收件人: xiebin (F) <xiebin18@huawei.com>
抄送: pgsql-general@postgresql.org; zhubo (C) <zhubo31@huawei.com>; Zhuzheng (IT) <zhuzheng@huawei.com>; houxiaowei <brian.hou@huawei.com>; yangshaobo (A) <yangshaobo6@huawei.com>; mapinghu <mapinghu@huawei.com>; Songyunpeng <songyunpeng@huawei.com>; luoqi (F) <luoqi25@huawei.com>
主题: Re: Security issues concerning pgsql replication

 

 

 

On Tue, Oct 27, 2020 at 9:52 AM xiebin (F) <xiebin18@huawei.com> wrote:

Hi,

 

I was setting up a master/slave pgsql(version 12.4) cluster using stream replication. I found 3 ways to authenticate, but all of them has some security issue.

 

1.  Disable authentication.

cat pg_hba.conf

    host   all           all    0/0               md5

    host   replication   xie    192.168.1.31/32   trust

 

In this case, untrusted users on slave may use pg_basebackup to stole data.

 

2.  Using password.

cat pg_hba.conf

    host   all           all    0/0               md5

    host   replication   xie    192.168.1.31/32   md5

 

cat /var/lib/pgsql/.pgpass (on slave)

    192.168.1.30:5432:xie:mydb:xie

 

In this case, the password is stored unencrypted. File access control may help, but it’s not secure enough.

 

 

Why not? The user who can read that file, can also read the entire database on the standby node already.

 

3.  Using certificate.

cat pg_hba.conf

    host   all           all    0/0               md5

    hostssl    replication   xie    192.168.1.31/32   cert clientcert=1

 

cat postgresql.conf | grep ssl

    ssl = on

ssl_ca_file = 'root.crt'

ssl_cert_file = 'server.crt'

ssl_crl_file = ''

ssl_key_file = 'server.key'

 

cat recovery.conf

    primary_conninfo = 'host=192.168.1.30 port=5432 user=xie application_name=stream_relication sslrootcert=/tmp/root.crt sslcert=/tmp/xie.crt sslkey=/tmp/xie.key'

restore_command = ''

recovery_target_timeline = 'latest'

primary_slot_name = 'rep_slot'

 

The certificates are created by official instructions https://www.postgresql.org/docs/12/ssl-tcp.html#SSL-CERTIFICATE-CREATION. But the private key is not encrypted.

I noticed in psql 11+ version, a new configuration ssl_passphrase_command is added, so that encrypted private key can be used.

But as far as I know, encrypted private key is not supported in stream replication.

 

I wonder if there is another way to authenticate in replication? Or does pgsql has any plan to support encrypted private key in replication?

 

 

PostgreSQL replication supports all authentication methods that PostgeSQL supports for regular connections, in general. While I haven't tried it, ssl_passphrase_command should work for this as well as long as it doesn't require manual user interaction. But it could for example read the passphrase from a pipe where it's provided off,or from a hardware device. Do keep in mind that replication might need multiple authentications (for example if the network disconnects, it has to reconnect).

 

You can also use for example GSSAPI and Kerberos to do the login. You will then of course have to figure out how to securely authenticate the postgres OS user on the standby node to the Kerberos system, but that's doable. (Though I believe most Kerberos implementations also rely on filesystem security to protect the tickets, so if you don't trust your filesystem, you may have a problem with that -- as well as indeed most other authentication systems -- so you'd have to investigate that within the kerberos system).