Hello GitLab,

I’m using GitLab for 5 years, and one of the most problems is updating GitLab to the latest version.
Firs I was using Ubuntu and in the last 2 years, CentOS but there are some problems in the installation if my GitLab hosted app refuses to finish the update.

I get error:
[root@git ~]# yum -y update
gitlab_gitlab-ee 368 B/s | 862 B 00:02
gitlab_gitlab-ee-source 384 B/s | 862 B 00:02
Dependencies resolved.

Package Architecture Version Repository Size

Upgrading:
gitlab-ee x86_64 13.3.6-ee.0.el8 gitlab_gitlab-ee 819 M

Transaction Summary

Upgrade 1 Package

Total download size: 819 M
Downloading Packages:
gitlab-ee-13.3.6-ee.0.el8.x86_64.rpm 93 MB/s | 819 MB 00:08

Total 93 MB/s | 819 MB 00:08
Running transaction check
The transaction check succeeded.
Running transaction test
The transaction test succeeded.
Running transaction
Preparing: 1/1
Running scriptlet: gitlab-ee-13.3.6-ee.0.el8.x86_64 1/2
gitlab preinstall: Automatically backing up only the GitLab SQL database (excluding everything else!)
2020-09-16 08:17:14 +0200 – Dumping database …
Dumping PostgreSQL database gitlabhq_production … [DONE]
2020-09-16 08:17:18 +0200 – done
2020-09-16 08:17:18 +0200 – Dumping repositories …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping uploads …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping builds …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping artifacts …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping pages …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping lfs objects …
2020-09-16 08:17:18 +0200 – [SKIPPED]
2020-09-16 08:17:18 +0200 – Dumping container registry images …
2020-09-16 08:17:18 +0200 – [SKIPPED]
rake aborted!
Errno::ENOENT: No such file or directory - tar
/opt/gitlab/embedded/service/gitlab-rails/lib/gitlab/popen.rb:38:in popen_with_detail' /opt/gitlab/embedded/service/gitlab-rails/lib/gitlab/popen.rb:14:in popen’
/opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:172:in tar_version' /opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:259:in backup_information’
/opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:21:in block (2 levels) in write_info' /opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:20:in open’
/opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:20:in block in write_info' /opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:19:in chdir’
/opt/gitlab/embedded/service/gitlab-rails/lib/backup/manager.rb:19:in write_info' /opt/gitlab/embedded/service/gitlab-rails/lib/tasks/gitlab/backup.rake:20:in block (3 levels) in <top (required)>’
/opt/gitlab/embedded/bin/bundle:23:in load' /opt/gitlab/embedded/bin/bundle:23:in ’
Tasks: TOP => gitlab:backup:create
(See full trace by running task with --trace)
gitlab preinstall:
gitlab preinstall: Database backup failed! If you want to skip this backup, run the following command and try again:
gitlab preinstall:
gitlab preinstall: sudo touch /etc/gitlab/skip-auto-backup
gitlab preinstall:
error: %prein(gitlab-ee-13.3.6-ee.0.el8.x86_64) scriptlet failed, exit status 1

Error in PREIN scriptlet in rpm package gitlab-ee
Verifying : gitlab-ee-13.3.6-ee.0.el8.x86_64 1/2
Verifying : gitlab-ee-13.3.5-ee.0.el8.x86_64 2/2

Failed:
gitlab-ee-13.3.5-ee.0.el8.x86_64 gitlab-ee-13.3.6-ee.0.el8.x86_64

Error: Transaction failed

For the first time I installed GitLab on MariaDB and behind the router (as Virtual Machine), now I have a server dedicated to GitLab, with dedicated resources, IP and so + I installed it as the default installation.

What is the problem with the upgrade process…

Thanks for the help and assistance in the advance

In case anyone else runs into this specific error:

Turns out, what it’s actually complaining about is that the ‘tar’ command is missing. I was surprised that even a minimal OS install (CentOS 8, in my case) could succeed without including something as basic as tar. At any rate I installed it and reran the upgrade, which then succeeded. (The upgrade performs a backup automatically by default unless you suppress it, and the backup process uses tar to generate its archive file.)

I suppose the error message would be more intuitive if it said something like “command not found” rather than “no such file or directory”.

well, taris not core OS functionality; OS will be fine without tar