- Thursday, January 8, 2015

openafs@conference.openafs.org

Thursday, January 8, 2015< ^ >

Room Configuration

Room Occupants

GMT+0
[03:22:24] mvita leaves the room
[08:22:14] ballbery leaves the room
[08:22:15] ballbery joins the room
[08:54:04] Simon Wilkinson joins the room
[10:01:55] ballbery leaves the room
[10:01:55] ballbery joins the room
[10:15:16] Simon Wilkinson leaves the room
[12:55:31] Simon Wilkinson joins the room
[13:31:43] abo joins the room
[16:43:34] kaduk joins the room
[18:30:44] <kaduk> Thinking about that -info posting ... I have some recollection that going from 1.2 to 1.4 the initial cache scan took a very long time, but subsequent ones were faster. Would that still be the case today (e.g., for 1.6)?
[18:30:44] Marc Dionne leaves the room
[18:30:44] Marc Dionne leaves the room
[18:30:44] Marc Dionne leaves the room
[18:38:49] <jhutz@jis.mit.edu/owl> Transitioning across that boundary will take a long time, because the
cache directory structure gets reorganized.
[18:39:42] <kaduk> To be clear, the 1.2 --> 1.4 boundary?
[18:41:16] <jhutz@jis.mit.edu/owl> I don't remember exactly where the change happened. I thought it was
longer ago than that, but I really don't remember.
[20:09:00] <Simon Wilkinson> The cache scan issue is that we perform a syscall for every potential file within the cache at startup.
[22:13:49] kaduk leaves the room
[23:00:27] Simon Wilkinson leaves the room
[23:01:07] Simon Wilkinson joins the room