How to handle later configuration changes of installed software ?

11 views
Skip to first unread message

Norbert Klamann

unread,
Jan 4, 2023, 3:27:16 AM1/4/23
to Ansible Project
Hello all,
I use ansible to install software and do the first configuration and i like this very much.

But I want to achieve that I can recreate the running systems as detailled as posible from the outside with scripts, with ansible for example.

Maybe an example helps to clarify the problem:

Consider necxtcloud. The software consists of a webserver, a bunch of php-software and a database.

The **configuration** consists of
* webserver configuration
* php configuration
* config.php (nextcloud)
* tables in the database

The **data** consists of
* files in the filesystem
* tables in the databas

The webserver configuration and the php configuration can be defined in ansible and
distributed, no problem there.

The  configuration tables in the database are changed via the web interface, i don't see a realistic chance to to handle them with ansible at all.

The *config.php* can and will be changed via the web interface and If i overwrite the file with ansible I lose all these changes.

How do peaople handle this kind of situation ? I try to be disciplined and change the ansible files synchronous to configuration changes on the running system, but I would like to make this process bullet-proof.


Any ideas ?

Thanks a lot

Norbert

Rowe, Walter P. (Fed)

unread,
Jan 4, 2023, 7:17:06 AM1/4/23
to ansible...@googlegroups.com
Database changes can be scripted in SQL by DBAs. These SQL scripts can be executed from a command line. Good app config mgmt should require this as part of DevSecOps automated deployments and upgrades. It also serves to document changes to tables, indexes, etc. Database server configuration changes also should be scripted and configuration files placed under version control. We use GitLab CI/CD workflows for app deployments. We use ansible for server and infrastructure management.

Walter
--
Walter Rowe, Division Chief
Infrastructure Services, OISM
Mobile: 202.355.4123

--
You received this message because you are subscribed to the Google Groups "Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ansible-proje...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ansible-project/cc3b974b-3cdc-4d20-9198-9e00b9edfd98n%40googlegroups.com.

Dick Visser

unread,
Jan 4, 2023, 7:20:02 AM1/4/23
to ansible...@googlegroups.com
You can't really have two sources of truth. 
I would just keep the ansible part and think of it as initial configuration (as you already do). 
From there on lots of entropy is added, so I would use something to keep that, like a backup strategy for the files and tables etc.
You could still use ansible, but then to automate the recovery process. 

--
You received this message because you are subscribed to the Google Groups "Ansible Project" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ansible-proje...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/ansible-project/cc3b974b-3cdc-4d20-9198-9e00b9edfd98n%40googlegroups.com.
--
Sent from Gmail Mobile
Reply all
Reply to author
Forward
0 new messages