[CPDEV-101420] Fix failure of second backup after restore

[n549]

Description

When doing the following sequence of action:

• backup
• restore
• backup
  the last backup fails at RHEL hosts.

RC

Out of the box RHEL has fs.protected_regular kernel variable set to 1 which means prohibiting O_CREAT open on regular files that we don’t own in world writable sticky directories (which the /tmp folder is), unless they are owned by the owner of the directory.

After the restore procedure the /tmp/kubemarine-backup.tar.gz file is left in /tmp folder, so the second backup procedure tries to update an existing file and fails.

Actually, Ubuntu 20.04 and 22.04 have fs.protected_regular=2 which means the same as 1, but also applies to group writable sticky directories, so the issue must be reproduced there too. Centos 7 doesn't have this variable and doesn't have such an issue.

Solution

Restore procedure leaves /tmp/kubemarine-backup.tar.gz file after itself. This file also can be left after previously failed backup procedures, etc.
So it's better to delete this file anyway before creating new one in the backup procedure.

How to apply

Test Cases

TestCase 1

Deploy a cluster at RHEL nodes and run maintenance procedures:

• backup
• restore
• backup
  ER: All the procedures finish successfully.

TestCase 2

Deploy a cluster at Ubuntu nodes and run maintenance procedures:

• backup
• restore
• backup
  ER: All the procedures finish successfully.

Checklist

• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• Integration CI passed
• Unit tests. If Yes list of new/changed tests with brief description
• There is no merge conflicts

Unit tests

Indicate new or changed unit tests and what they do, if any.