On Tue, Jul 21, 2015 at 3:48 PM, Junio C Hamano <gitster@xxxxxxxxx> wrote: > Junio C Hamano <gitster@xxxxxxxxx> writes: > >> While I still think that it is more important to prevent such a >> situation from occurring in the first place, ignoring .idx that lack >> corresponding .pack should be fairly simple, perhaps like this. >> ... > > Sorry for the noise, but this patch is worthless. We already have > an equivalent test in add_packed_git() that is called from this same > place. And a few extra updates from me: we found that this appears to occur even after update to 1.9.5, and setting core.fscache on 2.4.6 has no appreciable impact on the time it takes to run "git fetch", either. Our thought was antivirus (or something else?) might have the file open when git attempts to unlink the .idx, but perhaps it's something else, too? In one case, we had ~560 orphaned .idx files, but 150 seems sufficient to slow a fetch operation for a few minutes until it actually begins transferring objects. The "git gc" approach to cleaning up the mess is certainly looking more and more attractive... :) -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html