Some restored files consistently have incorrect Group Id

Hello,

I’ve been noticing for a while now that, sometimes when I restore data to a client, the GIDs of some files are not properly set. For example:

drwx------ 19 1001 1001     4096 Feb 19 22:27 .
drwxrwxrwx  4 1001 1001     4096 Feb 18 00:39 ..
-rw-------  1 1001 1001        3 Jul 30  2022 PG_VERSION
drwx------  7 1001 1001     4096 Sep 24  2022 base
-rw-------  1 1001 root 18857984 Feb  5 23:01 core
drwx------  2 1001 1001     4096 Feb 19 22:27 global
drwx------  2 1001 1001     4096 Jul 30  2022 pg_commit_ts
drwx------  2 1001 1001     4096 Jul 30  2022 pg_dynshmem
-rw-------  1 1001 1001     1636 Jul 30  2022 pg_ident.conf
drwx------  4 1001 1001     4096 Feb 20 07:57 pg_logical
drwx------  4 1001 1001     4096 Jul 30  2022 pg_multixact
drwx------  2 1001 1001     4096 Jul 30  2022 pg_notify
drwx------  2 1001 1001     4096 Jul 30  2022 pg_replslot
drwx------  2 1001 1001     4096 Jul 30  2022 pg_serial
drwx------  2 1001 1001     4096 Jul 30  2022 pg_snapshots
drwx------  2 1001 1001     4096 Feb  5 23:01 pg_stat
drwx------  2 1001 1001     4096 Feb 20 07:58 pg_stat_tmp
drwx------  2 1001 1001     4096 Feb  7 00:03 pg_subtrans
drwx------  2 1001 1001     4096 Jul 30  2022 pg_tblspc
drwx------  2 1001 1001     4096 Jul 30  2022 pg_twophase
drwx------  3 1001 1001     4096 Feb 20 00:02 pg_wal
drwx------  2 1001 1001     4096 Jul 30  2022 pg_xact
-rw-------  1 1001 1001       88 Jul 30  2022 postgresql.auto.conf
-rw-------  1 1001 1001      249 Feb 19 22:27 postmaster.opts
-rw-------  1 1001 root       79 Feb 19 22:27 postmaster.pid

The Group ID for ALL the files should be 1001 but one file and one folder have the incorrect GID of root. This means that, after a restore, I need to go in and manually chown the files to have the correct ID’s. Nothing else is accessing the directory that the data is being restored to.

Urbackup Server Version: 2.5.32
Urbackup Client Container: latest

Backups are stored on a BTRFS filesystem.

The logs of the restore don’t seem to indicate anything unusual going on:

Level Time Message
Info 02/20/24 14:53 Starting restore of path “backup/db-data”…
Info 02/20/24 14:53 Starting restore…
Info 02/20/24 14:53 Loading file list…
Info 02/20/24 14:53 Calculating download size…
Info 02/20/24 14:53 Downloading necessary file data…
Info 02/20/24 14:55 Applying file metadata…
Info 02/20/24 14:55 Restore finished successfully.

I have also attached the client log which does display some errors:
2024-02-20 19:24:19: ERROR: Error getting internal id for user with id 1001
urbackupclient.log (213.0 KB)

But I’d think that this shouldn’t affect the actual ID of the data after a restore.