Backups with size of ~ 4 GB not copied to remote SAMBA share (SAMBA backup fails)

[bcutter]

SAMBA backup fails for backups of roughly around or greater than 4 GB. Everything below that mark works just fine. For a weekly backup which now hit the 4 GB size mark, it is not uploaded to the remote SAMBA storage.

• Created locally: ✅
• Copied to remote share: ❌
• Final backup state: ❌ failed

Of course this can be highly dangerous if the local storage dies - and the important backup is not available remotely.

Logs of latest occurence:

[23-11-11 02:39:36] INFO: Backup finished
[23-11-11 03:30:00] DEBUG: Input received: {"command":"trigger","backup_name":"{type} Backup nur InfluxDB {version} {date}","exclude_folders":["addons/local","media","ssl","share","homeassistant"],"exclude_addons":["xxxMASKEDxxx"]}
[23-11-11 03:30:01] INFO: Running backup with customized parameters
[23-11-11 03:30:01] INFO: Backup running ...
[23-11-11 03:30:02] DEBUG: Backups local/remote: 3/45
[23-11-11 03:30:02] DEBUG: Total backups succeeded/failed: 937/166
[23-11-11 03:30:02] DEBUG: Last backup: 2023-11-11 02:39
[23-11-11 03:30:02] DEBUG: Posting sensor data to API at /core/api/states/sensor.samba_backup
[23-11-11 03:30:03] DEBUG: API Status: 200
[23-11-11 03:30:03] DEBUG: API Response: {"entity_id":"sensor.samba_backup","state":"RUNNING","attributes":{"friendly_name":"Samba Backup","backups_local":"3","backups_remote":"45","total_backups_succeeded":"937","total_backups_failed":"166","last_backup":"2023-11-11 02:39"},"last_changed":"2023-11-11T02:30:03.102345+00:00","last_updated":"2023-11-11T02:30:03.102345+00:00","context":{"id":"01HEY3E02YMRQ66JES9HNTG9FF","parent_id":null,"user_id":"xxxMASKEDxxx"}}
[23-11-11 03:30:03] INFO: Creating backup "Partial Backup nur InfluxDB 2023.4.6 2023-11-11 03:30"
parse error: Invalid numeric literal at line 1, column 5
[23-11-11 06:31:42] DEBUG: Posting sensor data to API at /core/api/states/sensor.samba_backup
[23-11-11 06:31:42] DEBUG: API Status: 200
[23-11-11 06:31:42] DEBUG: API Response: {"entity_id":"sensor.samba_backup","state":"FAILED","attributes":{"friendly_name":"Samba Backup","backups_local":"3","backups_remote":"45","total_backups_succeeded":"937","total_backups_failed":"167","last_backup":"2023-11-11 02:39"},"last_changed":"2023-11-11T05:31:42.672672+00:00","last_updated":"2023-11-11T05:31:42.672672+00:00","context":{"id":"01HEYDTM6GPDF78HQQ74KZJET5","parent_id":null,"user_id":"xxxMASKEDxxx"}}
[23-11-11 06:32:26] DEBUG: Posting sensor data to API at /core/api/states/sensor.samba_backup
[23-11-11 06:32:26] DEBUG: API Status: 200
[23-11-11 06:32:26] DEBUG: API Response: {"entity_id":"sensor.samba_backup","state":"IDLE","attributes":{"friendly_name":"Samba Backup","backups_local":"3","backups_remote":"45","total_backups_succeeded":"937","total_backups_failed":"167","last_backup":"2023-11-11 02:39"},"last_changed":"2023-11-11T05:32:26.867523+00:00","last_updated":"2023-11-11T05:32:26.867523+00:00","context":{"id":"01HEYDVZBKGG3SCB18C1M45HNX","parent_id":null,"user_id":"xxxMASKEDxxx"}}
[23-11-11 06:32:27] INFO: Backup finished
[23-11-12 02:00:53] INFO: Backup running ...

Possibly similar to #151

Need assistance on this.

[Mikescotland]

I had a similar problem - backup is creating first a tmp backup and then writes it to Samba. In my case, HA drive became full and backup continued - after a few hours stopped.
Interesting thing is that the drive still had 40 GB free, I noticed a notification in Ubuntu about the drive being full. Not sure what happened though - the media folder is around 6GB, I excluded it now and works. Looks like the addon does something in this case that fills up the drive (backup was running for.... 3hr before failed, unlikely on a core i5 machine to take so long)

[bcutter]

I can't follow you completely. Did you have storage issues on the source (HA) or the target (remote SAMBA share)?
As mentioned the backup is created fine locally (where storage could be an issue in theory, but is not, see [*1], but not transferred to the remote storage (which has more than enough space by the way).

[*1]

[Mikescotland]

Source.
Despite having around 50GB free and media folder only 8GB, Ubuntu threw and error after an hour of backup saying drive is completely full.
Tmp file on source drive was massive, like backup was endlessly adding and adding..

[bcutter]

OK, thanks for your report (which made me check storage, didn't think about that before). Very likely not the case (not a storage issue) here.

[ouaibe]

This seems very similar to #65 and #137 and #150 currently not solved until #79 implementation.

[bcutter]

...which is open for almost 2 years. Wow. Addon slowly dying... no big hopes. So sad.

[bcutter]

Intermediate informatione:

For me some supervisor update seems to have boosted backup creation by a lot already somewhere in December 2023. Problem likely never was/is not the actual backup size, but the time needed to create it. Because the latter one decreased dramatically, this issue did not happen anymore for me - currently. Anyway, this addon seems to run into some limitations (which will likely arise again once the backup size and therefore duration increases again over time).

[ouaibe]

Re-enabled the extension on my end and it seems the time to take backups has significantly improved and thus the timeout doesn't happen anymore, 'fixing' this issue. My backup files are > 4GB.

[bcutter]

Re-enabled the extension on my end and it seems the time to take backups has significantly improved and thus the timeout doesn't happen anymore, 'fixing' this issue. My backup files are > 4GB.

So you confirmed #161 (comment).