Removal of obsolete media files?

[mathieumd]

How are purged obsolete files? Isn't /etc/cron.daily/sheltermanager3 supposed to clean up old files?

We have 4 photos actually used:

sqlite> select ID,DBFSID,MediaSize,MediaName from media;
31|40|307085|31.jpg
32|41|199055|32.jpg
33|42|168479|33.jpg
34|43|128715|34.jpg

But the file folder contains much more files, because I made a few tests and then removed media from the web UI:

ls -l /srv/www/asm3_data/asm3sqlite/
total 1000
-rw-rw-rw- 1 www-data www-data  32529 juin  24 09:32 34.jpg
-rw-rw-rw- 1 www-data www-data  32529 juin  24 09:34 36.jpg
-rw-rw-rw- 1 www-data www-data  32529 juin  24 09:50 37.jpg
-rw-rw-rw- 1 www-data www-data   7702 juin  24 09:57 38.jpg
-rw-rw-rw- 1 www-data www-data   7702 juin  24 09:58 39.jpg
-rw-rw-rw- 1 www-data www-data 307085 juin  24 10:33 40.jpg
-rw-rw-rw- 1 www-data www-data 199055 juin  24 10:40 41.jpg
-rw-rw-rw- 1 www-data www-data 168479 juin  24 10:43 42.jpg
-rw-rw-rw- 1 www-data www-data 128715 juin  24 10:49 43.jpg
-rw-rw-rw- 1 www-data www-data  32529 juin  24 11:04 44.jpg
-rw-rw-rw- 1 www-data www-data  32529 juin  24 11:05 45.jpg
-rw-rw-rw- 1 www-data www-data  32529 juin  24 11:11 46.jpg

By the way, files referenced in DB column MediaName (for example 31.jpg or 34.jpg) are not the name in the filesystem (there 34.jpg is one of my test which were removed from the web UI since). The actual files are 4[0-3].jpg (as can be confirmed by the file sizes), which is the name shown in the colum DBFSID.

[bobintetley]

We have lots of customers who delete things that they didn't really mean to, and since we use S3 compatible storage which is cheap, we never actually delete anything from the dbfs table and we don't remove the real media files. This means in combination with the deletion table, we can restore anything that has been deleted in the last 6 months.

If you do want to remove your deleted media forever, you can run the cron.py script with the target "maint_db_delete_orphaned_media"

This will delete the DBFS files and database entries that have been removed from the media table to bring them back in sync.

[mathieumd]

Nice, thanks.

To add it in a crontab, it looks like I'll have to redirect it's stderr to /dev/null:

sudo python3 cron.py maint_db_delete_orphaned_media
config: /etc/asm3.conf

sudo python3 cron.py maint_db_delete_orphaned_media >/dev/null
config: /etc/asm3.conf

sudo python3 cron.py maint_db_delete_orphaned_media 2>/dev/null

I would prefer to get an output only when there is a real error, though, which would allow me to be notified by email by Cron. As discussed in #1359.

[bobintetley]

I'll remove this in master now, there's no need for the config file to be output to stderr like this, it's unnecessary debug code and if it should have gone anywhere it should have been to the syslog.