[01:00:34] --- Russ has left: Disconnected [01:43:18] --- reuteras has become available [01:43:33] --- reuteras has left [02:09:02] --- sxw has become available [02:12:38] --- sxw has left [02:12:38] --- sxw has become available [02:45:11] --- sxw has left [03:15:38] --- sxw has become available [03:19:13] --- sxw has left [03:19:13] --- sxw has become available [05:26:33] --- SecureEndpoints has left: Replaced by new connection [05:37:26] --- SecureEndpoints has become available [06:40:34] --- reuteras has become available [07:03:01] --- reuteras has left [07:06:57] --- SecureEndpoints has left: Disconnected [07:26:57] --- matt has become available [09:14:05] --- dev-zero@jabber.org has become available [09:14:11] --- dev-zero@jabber.org has left: offline [09:19:07] --- SecureEndpoints has become available [10:55:02] --- Russ has become available [12:24:51] --- mmeffie has become available [13:03:30] secureendpoints (or others) - are you familiar with the .msf filename extension? [13:03:46] is that an older Microsoft development environment suffix? [13:05:33] not an email message or something? [13:09:43] no [13:09:58] it's not the mozilla format. it is an alternate suffix instead of '.c'. [13:10:12] e.g., maybe 'microsoft framework'? [13:10:12] sorry [13:10:29] (btw, steven, did you ever split the piov patch)? [13:12:23] piov? [13:13:12] you mean the parallel I/O vec stuff? (readv, writev..) [13:17:38] yah [13:18:21] well, FDH_PREAD and friends [13:23:36] it started out as a separate patch, then got rolled into a jumbo. I don't know the latest status (haven't looked at it in close to a year) [13:23:46] iirc, the testing showed no improvement, though. [13:24:34] ok [13:25:25] iirc, adeason 'touched it last'. [13:25:34] so he would be the one w/the latest info. [13:25:41] tkeiser might also know, though. [13:44:06] pread/pwrite would probably bring more improvement in conjunction with other threading improvements i'd venture should yet be done [13:47:47] shadow - point noted. [13:48:31] also, let's just talk about NetInfo/setcbaddr here [13:54:06] sure. [13:54:34] afaict, setcbaddr overrides NetInfo, but only as long as the addr is valid. [13:56:08] i wrote it, but i don't promise i remember: i though setcbaddr simply allowed you to specify a valid address. if NetInfo is in play, then NetInfo must allow (or NetRestrict must not restrict, if that's in play) [14:07:18] thanks. I may take the low road and not specify it for now. [14:07:30] (I'm down to the last 6 man pages..) [14:08:47] it would be helpful if someone w/OAfW installed would do fs monitor and fs trace. Otherwise, I'll need to wrestle w/that. [14:09:03] take the low road. i will try to come up with text by when you submit [14:09:07] tx. [14:09:37] I should have all but those two done by tomorrow, but today is the first sunshine in a while, so I need to do some mowing tonight. [14:09:51] and there are these people called 'clients' and 'bosses' that keep asking me to do things... :) [14:12:06] i don't even have a wife asking me to do things. i do it because of inner drive. i am clearly nuts. [14:46:43] --- dev-zero@jabber.org has become available [14:46:47] --- dev-zero@jabber.org has left: offline [14:50:09] fs trace [-on] [-off] [-reset] [-dump] [-help] [14:50:09] "fs monitor" does not exist on Windows [14:50:25] see release notes for info on fs trace [14:50:26] monitor is mariner, right? [14:53:07] Yes, monitor is mariner. [14:53:39] Actually, mariner is a Dell Dimension 9150/XPS 400, like tuzanor [14:55:32] yes, i was going to make that joke myself, later [15:44:35] --- matt has left [16:06:37] I wonder if we should return EIO when we can't open the file in the dcache. Instead of, you know, deadlocking the user's machine. [16:12:20] --- matt has become available [16:12:33] ouch [16:13:44] --- matt has left [17:47:00] and the harris grill network asploded. sigh. [18:14:39] returning EIO might be smarter, yes. perhaps also with more debugging info [19:02:41] secureendpoints - tx. am also wanting to include a few examples. [19:04:08] --- Russ has left: Disconnected [19:23:10] --- Russ has become available [19:31:53] for fs monitor/mariner - any reason not to change it from CMD_HIDDEN? that would at least let people see that the command is there.. [19:32:15] probably no reason [19:33:24] I'll change that in my doc patch unless there are objections. [19:34:44] having one patch to change that and document it is probably smarter than rolling it in with something else [19:34:54] makes sense. [19:36:20] I'm not clear what monitor/mariner is doing, though. I had thought it's managing afsmonitor stat collection -- that's clearly not the case. So what's the other end? [19:36:35] s/thought/guessed/ [19:37:06] it says "send messages from this client to X", iirc [19:37:21] right. But how to you manage the collection on host X. [19:37:39] it's running a cache manager. it collects it [19:37:42] ie, do you need to tell host X to listen for it? (I'm looking at afs_mariner.c atm) [19:38:13] EBUSY? [19:41:07] so how do you extract the data from the monitoring station? afsmonitor? [19:41:23] or does it automatically go into that hosts log? [19:41:41] it gets logged just like local data, iirc [19:41:48] ok, tx. [20:03:54] --- asedeno has left [20:03:54] --- asedeno has become available [20:03:57] --- asedeno has left [20:04:17] --- asedeno has become available [21:06:45] --- mmeffie has left [22:33:04] --- Simon Wilkinson has left [23:05:45] --- dev-zero@jabber.org has become available [23:05:48] --- dev-zero@jabber.org has left: offline [23:30:48] --- tkeiser@sinenomine.net/owl has left [23:58:51] --- sxw has left