Debian – How to remove a bunch of ancient packages on debian

debianpackage-managementuninstall

I'm running sid, and in the course of trying to cross-grade my system from i386 to amd64 I came across some ancient packages that I couldn't remove. Some background: I've had this system since potato, or maybe earlier.

There are about a hundred packages like this, so I'd like a generic or scriptable answer. Here's one example:

bminton:/var/cache/apt/archives# dpkg --purge libstdc++2.10-dev
(Reading database ... 1352516 files and directories currently installed.)
Removing libstdc++2.10-dev (1:2.95.4-27) ...
install-info: No dir file specified; try --help for more information.
dpkg: error processing package libstdc++2.10-dev (--purge):
 subprocess installed pre-removal script returned error exit status 1
Errors were encountered while processing:
 libstdc++2.10-dev

The prerm script `/var/lib/dpkg/info/libstdc++2.10-dev.prerm script contains the following:

#! /bin/sh -e



install-info --quiet --remove iostream-2.95

Manually running install-info --quiet --remove iostream-2.95 gives the following error:

install-info: No dir file specified; try --help for more information.

Best Answer

dpkg used to have its own install-info script which was used in place of the GNU one. An email about the change gives a suggestion for packages (formatting added):

These packages should just drop their info files in /usr/share/info, and call the update-info-dir script if present (postinst and prerm). They could suggest/recommend the info package.

So, what I'd suggest you do is edit (yes, edit) /var/lib/dpkg/info/libstdc++2.10-dev.prerm and comment out the install-info ... line. Do the same for other packages with that failure. Once you're done purging the packages, manually run update-info-dir.

BTW: In the future, after doing a upgrade, you should check the list of obsolete/local packages on your system and purge them if not needed. Otherwise, you wind up with very outdated maintainer scripts left lying around.

Related Solutions

Debian – How to Remove Group of No Longer Needed Packages

I would suggest using user tags, which is a characteristic of aptitude (apt binary doesn't have this) would help with that. When installing packages add the --add-user-tag option when installing the packages, for example:

sudo aptitude --add-user-tag build-dep-package build-dep package

and to remove them just use:

sudo aptitude remove '?user-tag(build-dep-package)'

You can also remove the tag along the way with --remove-user-tag build-dep-package.

But this pose a problem, what about if the same package has both tags? Well, you just use the and/not conditionals to prevent that from happening:

sudo aptitude remove '?and(?user-tag(build-dep-package),?not(?user-tag(wanted-packages))'

To manipulate several packages with different tags use or:

sudo aptitude remove '?or(?user-tag(build-dep-package),?user-tag(wanted-packages)'

Through this is very useful in the long run, I haven't found a way to list all the current active tags.

Debian – Failed to start MariaDB database server after upgrade to debian 9

In the tutorial you should install mariadb-server in place of mysql-server. I've got the same problem on DEBIAN 9 install and these command worked for me :

sudo apt purge mysql*
sudo apt purge mariadb*
sudo apt autoremove
sudo apt autoclean

Check if still something in the system :

dpkg -l | grep -e mysql -e mariadb

if the list is empty then install mariadb server :

sudo apt install mariadb-server

Now try to check status on stop and start the database :

Status after install :

sudo systemctl status mysql.service
● mariadb.service - MariaDB database server
   Loaded: loaded (/lib/systemd/system/mariadb.service; enabled; vendor preset: enabled)
   Active: active (running) since Wed 2018-03-21 16:08:49 CET; 2min 1s ago
 Main PID: 15699 (mysqld)
   Status: "Taking your SQL requests now..."
   CGroup: /system.slice/mariadb.service
           └─15699 /usr/sbin/mysqld

.../..

Check the PID :

sudo more /var/run/mysqld/mysqld.pid 
15699

Stop the database :

sudo systemctl stop mysql.service
sudo systemctl status mysql.service
● mariadb.service - MariaDB database server
   Loaded: loaded (/lib/systemd/system/mariadb.service; enabled; vendor preset: enabled)
   Active: inactive (dead) since Wed 2018-03-21 16:14:35 CET; 1s ago
 Main PID: 15699 (code=exited, status=0/SUCCESS)
   Status: "MariaDB server is down"

.../...

You could check the PID file is not present as well

sudo more /var/run/mysqld/mysqld.pid 
more: stat on /var/run/mysqld/mysqld.pid: No file or folder

Start the database :

sudo systemctl start mysql.service
sudo systemctl status mysql.service
● mariadb.service - MariaDB database server
   Loaded: loaded (/lib/systemd/system/mariadb.service; enabled; vendor preset: enabled)
   Active: active (running) since Wed 2018-03-21 16:14:54 CET; 3s ago
  Process: 15911 ExecStartPost=/bin/sh -c systemctl unset-environment _WSREP_START_POSITION (code=exited, status=0/SUCCESS)
  Process: 15908 ExecStartPost=/etc/mysql/debian-start (code=exited, status=0/SUCCESS)
  Process: 15791 ExecStartPre=/bin/sh -c [ ! -e /usr/bin/galera_recovery ] && VAR= ||   VAR=`/usr/bin/galera_recovery`; [ $? -eq 0 ]   && systemctl set-environment _WSREP_S
  Process: 15787 ExecStartPre=/bin/sh -c systemctl unset-environment _WSREP_START_POSITION (code=exited, status=0/SUCCESS)
  Process: 15784 ExecStartPre=/usr/bin/install -m 755 -o mysql -g root -d /var/run/mysqld (code=exited, status=0/SUCCESS)
 Main PID: 15881 (mysqld)
   Status: "Taking your SQL requests now..."
    Tasks: 26 (limit: 4915)
   CGroup: /system.slice/mariadb.service
           └─15881 /usr/sbin/mysqld

You could check the PID file it should be present. May help on a empty database. If there is something in the database you should do a backup first.

Best Answer

Related Solutions

Debian – How to Remove Group of No Longer Needed Packages

Debian – Failed to start MariaDB database server after upgrade to debian 9

Related Question