On 2015-01-29 12:10:41, Ward Dossche wrote to Wilfred van Velzen: about: "Re: down ?":
Wv>> When that happens your system is put on hold for a couple of hours, Wv>> so there are fewer calls... Disk full, maybe?
WD> Interesting to go look in a log-file after all, and this is what i find for WD> you:
WD> *************************************************************************** WD> ** ? 29 Jan 01:12:03 [3944] no enough free space in c:\db\inbound (520192, WD> req-d 2098011) WD> *************************************************************************** WD> **
WD> There is 42GB available on average, there was no unsusual activity, no WD> mail-bomb ... no clue ...
WD> I find a similar report about a dozen times ... 2 this morning ... maybe 10 WD> yesterday.
WD> No clue.
You might want to update your binkd. I notice you are using a really old version: 0.9.5a.
There was a bug in the free disk space calculation that was fixed 11 years ago:
2004/01/02 17:32:02 1.0a-323 stas inbound.c,2.27,2.28 Fix the minfree token usage and fix getfree() on Win >w95
2004/01/02 17:31:33 1.0a-322 stas getfree.c,2.8,2.9 Fix the minfree token usage and fix getfree() on Win >w95
Or a value of 0 for the minfree configuration options might be a work around:
# # Binkd will skip all files from a node if # size_of_the_next_file_for_us_there + minfree < free_space_in_inbound # The zero value and the value 4294967295 (2**32-1) is equivalented to infinity. # minfree 2048 minfree-nonsecure 2048
