[vdr] Near-GAU of AssertFreeDiskSpace
udo_richter at gmx.de
Sun Feb 19 16:28:18 CET 2006
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
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.
More information about the vdr