Generally older FD will work with newer SD and DIR. But do make sure your SD and DIR are the same version. Save a lot of headaches.
That said if you can update the FD I would for fixes.
Brock Palen
1 (989) 277-6075
bro...@mlds-networks.com
www.mlds-networks.com
Websites, Linux, Hosting, Joomla, Consulting
> On Apr 27, 2020, at 10:05 AM, Goncalo Sousa <gtemp...@gmail.com> wrote:
>
> I have read https://docs.bareos.org/IntroductionAndTutorial/UpdatingBareos.html but it doesnt specify the order or how to only the database.
>
> How do I upgrade it?
>
> The server is CentOS Linux release 7.5. This means I have to upgrade the Clients FD's too or are they compatible?
>
> --
> You received this message because you are subscribed to the Google Groups "bareos-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to bareos...@googlegroups.com.
su postgres -c /usr/lib/bareos/scripts/update_bareos_tables su postgres -c /usr/lib/bareos/scripts/grant_bareos_privileges
> To view this discussion on the web visit https://groups.google.com/d/msgid/bareos-users/8767d51e-a61a-429f-a033-7e2be528fc63%40googlegroups.com.
In short, yum history rollback is not a good way to do do anything other than very small package changes. i.e. downgrading from version 1.4.27 to 1.4.26 of some package or completely removing package you just installed for testing along with all dependencies.
Longer explanation:
Yum history rollback just finds out which packages are to be installed (and in which versions) and which are to be removed. After that it simply downloads needed pakage files and runs appropriate install/remove actions. It's in no way similar to any "restore snapshot" operation. In case of files contained in the packages, the situation is easy - the "new" package installs new packages, then the files belonging to the "old" package but not belonging to "new" one are removed. Simple.
It gets more tricky with package scripts. RPM runs like this:
1. Runs pre-install script
2. Installs files
3. Runs post-install script
4. Runs pre-uninstall script
5. Removes files
6. Runs post-uninstall script
Steps 1-3 are run if there's any package installation or update (upgrade/downgrade) and steps 4-6 are run if there's any package removal (uninstall/upgrade/downgrade).
Bear with me, it's getting more magical :-)
The scripts executed during the operations are run almost
completely without any additional environment or context. The only
parameter is "a number representing the number of instances of the
package currently installed on the system, after
the current package has been installed or erased"
(http://ftp.rpm.org/max-rpm/s1-rpm-inside-scripts.html). This
means that install scripts might be run with an argument of 1
(initial install) or 2 (upgrade/downgrade) since after the install
stage the number of package instances on the system will be equal
to those numbers and the uninstall scripts might be run with an
argument of 0 (complete removal) or 1 (upgrade/downgrade) since
that's how many instances of package will be left after the
uninstall phase. There is probably an option for
You can see how it's used by querying a package with rpm -q --scripts. A good example here is an openssh-server package which runs some actions only on complete removal.
Unfortunately, the rpm scripts have no more context about the
package versions (and don't have any reasonable way for querying
the state of the rpm database in the middle of a transaction) so
there is no way of knowing whether the update operation is an
upgrade or downgrade. And we need also to remember than usually
"big upgrades" invoke some "external" changes, like database
schema upgrades, which are not easily rollbackable (as we said
before - rpm rollback doesn't work on system-level snapshot).
Therefore it's usually either assumed that update operation is
indeed an upgrade one, not a downgrade. And it's a good practice
to check prerequisites before executing needed "upgrade scripts"
(like check database schema version beforehand as to not apply an
upgrade script twice). Sometimes packagers just create a package
with new files and tell you to upgrade the dependencies (like
database schema) on your own.
So, long story short - no, there is no simple way to perform a yum history rollback and have a usable system if the upgrade had been "deep".
Therefore if you upgrade Bareos installation upgrade I'd do a
full configuration backup and database dump to have a working
state to which you can roll back.
Hope this wasn't too complicated :-)
--