2nd run on MB mirror-server: How to solve error "wget: memory exhausted"

MusicBrainz Development
1

I accidentally entered the wrong password for my virtual machine MB database inside the Docker MB mirror server.
So I started again and saw that all the dump files had already been downloaded.
Unfortunately, I get the error after the dump file recording.tar.zst
wget: memory exhausted

...

--2026-06-18 07:11:34--  https://data.metabrainz.org/pub/musicbrainz/data/solr-backups/20260617-145734/instrument.tar.zst
Reusing existing connection to data.metabrainz.org:443.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable

    The file is already fully retrieved; nothing to do.

--2026-06-18 07:11:34--  https://data.metabrainz.org/pub/musicbrainz/data/solr-backups/20260617-145734/label.tar.zst
Reusing existing connection to data.metabrainz.org:443.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable

    The file is already fully retrieved; nothing to do.

--2026-06-18 07:11:36--  https://data.metabrainz.org/pub/musicbrainz/data/solr-backups/20260617-145734/place.tar.zst
Reusing existing connection to data.metabrainz.org:443.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable

    The file is already fully retrieved; nothing to do.

--2026-06-18 07:11:37--  https://data.metabrainz.org/pub/musicbrainz/data/solr-backups/20260617-145734/recording.tar.zst
Reusing existing connection to data.metabrainz.org:443.
HTTP request sent, awaiting response... 416 Requested Range Not Satisfiable

    The file is already fully retrieved; nothing to do.

wget: memory exhausted

I have already rebooted my VM, that didn’t help.
My VM has actually 40 GB RAM.

Which process is currently causing the “memory exhausted” error?
What can I do now?

2

After examining the memory, processes and log files, an AI assisted me in finding a workaround to restore my MB mirror-server.

I checked that the dump files that had already been downloaded were really complete
(you need to start the container musicbrainz manually, if it is not running with
sudo docker compose up -d musicbrainz
)

sudo docker compose exec musicbrainz bash
cd /media/dbdump
md5sum -c MD5SUMS

As all the files were reported as OK **, I proceeded with:

sudo docker compose exec search load-backup-archives

and after that (again, without reporting any errors):

sudo docker compose run --rm musicbrainz recreatedb.sh
sudo docker compose up -d

The AI has guessed that the command -r (recursive download) in
wget -nd -nH -c -r -P ...
inside the script fetch-backup-archives in combination with
a large dataset (Solr backups, multi-GB files)
many already-existing files
repeated 416 Requested Range Not Satisfiable

is causing the error. The AI suggests that, in this case, the following script load-backup-archives should not rely on the “wrong” wget state and should instead use the MD5SUM check results.

As I am a bloody Ubuntu noob, I have no idea if this is just another AI hallucination :innocent:

** Note:
The MD5SUM command will show 3 missing files, this is the expected result!
Neither the mbdump-documentation nor the mbdump-editor nor mbdump-edit data can be downloaded as data dumps.

root@5623e2322721:/media/dbdump# md5sum -c MD5SUMS
mbdump-cdstubs.tar.bz2: OK
mbdump-cover-art-archive.tar.bz2: OK
mbdump-derived.tar.bz2: OK
md5sum: mbdump-documentation.tar.bz2: No such file or directory
mbdump-documentation.tar.bz2: FAILED open or read
md5sum: mbdump-editor.tar.bz2: No such file or directory
mbdump-editor.tar.bz2: FAILED open or read
md5sum: mbdump-edit.tar.bz2: No such file or directory
mbdump-edit.tar.bz2: FAILED open or read
mbdump-event-art-archive.tar.bz2: OK
mbdump-stats.tar.bz2: OK
mbdump.tar.bz2: OK
mbdump-wikidocs.tar.bz2: OK
md5sum: WARNING: 3 listed files could not be read