Re: Barman disaster recovery solution - Mailing list pgsql-general

From Edson Carlos Ericksson Richter
Subject Re: Barman disaster recovery solution
Date
Msg-id e2fd3516-e544-ce81-512a-a7fc720eb389@simkorp.com.br
Whole thread Raw
In response to Barman disaster recovery solution  (Julie Nishimura <juliezain@hotmail.com>)
Responses Re: Barman disaster recovery solution  (Stephen Frost <sfrost@snowman.net>)
List pgsql-general
Em 21/02/2019 04:17, Julie Nishimura escreveu:
> Does anyone use this solution? any recommenations?
>
> Thanks!


We do use it.

IMHO, those are minimum recommendations:


1) start using it! It's easy and robust.

2) for minimal impact over production servers, setup replicated servers 
and create your backup from slave servers.

3) *_test your backups_*. This is a MUST HAVE - no option here.

4) have your backup server in different cities, or states, or even 
countries. Never, ever create a backup on the server at the side of your 
production server.

5) only communicate with your servers using SSH and private key 
certificates. Establish a PKI infrastructure in a way that production 
and backup servers only communicate using SSH and certificates.

6) your backup servers shall never ever be connected directly to the 
internet. Hackers love low attention backup servers running with minimal 
security.


No backup solution (no matter which one you choose) is 100% guaranteed: 
your disks may fail, your network mail fail, your memory may fail, files 
get corrupted - so, setup a regular "restore" to separate "test backup 
server" on daily basis. Having a virtual server for this purpose has 
minimal budget impact if any at all, and you save your sanity in case of 
a disaster.



Regards,



Edson





pgsql-general by date:

Previous
From: Sergey Burladyan
Date:
Subject: Running from 9.6 backups sometimes fails with fatal error
Next
From: Ron
Date:
Subject: Re: adding more space to the existing 9.6 cluster