Follow-up Comment #2, patch #4917 (project mldonkey):
The patch is not finished yet, I put it here for discussion.
I am planing to send warning e-mails in a user-configurable interval (default
one hour after the first e-mail).
"but that is why the config_files_space.tmp was used"
True, but the diskspace config_files_space.tmp occupies is only
released before saving ini files on core shutdown. During core run
saving ini files happens without removing config_files_space.tmp first.
I think (untested) that removing config_files_space.tmp before
each saving of ini files and to re-create it afterwards will block
the core for some time. Saving of ini files is not threaded...
Having an option whether to shutdown the core if core_dir diskspace
is too low is no problem.
Also I want to code a check in CommonInteractive.file_commit if
diskspace on incoming dir is big enough to commit a file. If not,
the file will stay in the commit queue. The only problem I am
seeing is how to determine if !!temp_directory and incoming_dir
are on the same physical partition, in this case commit means
move which is unproblematic in regard of free diskspace.
The only idea I have is to compare total and free disk space on
!!temp_directory and incoming directory.
Is it save to assume that when these values match both dirs are
on the same partition?
_______________________________________________________
Reply to this item at:
<http://savannah.nongnu.org/patch/?func=detailitem&item_id=4917>
_______________________________________________
Nachricht geschickt von/durch Savannah
http://savannah.nongnu.org/
_______________________________________________
Mldonkey-users mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/mldonkey-users
