Hi list,
Yesterday I was close to a GAU of the AssertFreeDiskSpace recording cleanup of VDR.
I've been working on a remote VDR client using streamdev and NFS, and did a setup where /video was local, and /video/remote1 and /video/remote2 were NFS mounted recording trees. Since the client was not supposed to record at all, /video was rather small.
After fiddling for some time, the disk free info in the main menu caught my eye, saying 0:45min free. Thats quite close to disaster: At 0:39, VDR would have started deleting recordings, and since the recordings were on different mounts, free disk space would have been unaffected. VDR would soon have cleaned my server's disk of everything that he's allowed to clean...
Well, lucky I am. But how to avoid such a disaster? For now I've patched the MINDISKSPACE to 0, but maybe this trap should be fixed generally? I think it may be a good idea to limit AssertFreeDiskSpace to recordings on the same mount (or mounts, if VDR knows of several /video mounts), as deleting on un-monitored mounts doesn't affect 'free disk space' anyway.
Cheers,
Udo