[00:30:36] --- Russ has left: Disconnected [00:56:31] --- dev-zero@jabber.org has left [03:13:43] --- kaj has become available [03:36:00] jaltman - yes... appears to work properly... I get tokens on login [03:47:21] --- kaj has left [03:47:21] --- abo has left [03:47:52] --- Jeffrey Altman has become available [03:47:57] --- jaltman has left: Replaced by new connection [03:47:57] --- jaltman has become available [04:26:41] --- haba has left [06:10:11] there's LRU code in DNLC, uncompiled. i wonder if there's any benefit in using it. it will require the xdnlc write lock [06:10:37] Is the dnlc ever a bottle neck? [06:10:43] --- kaj has become available [06:12:34] possibly. i dunno [06:13:53] --- Simon Wilkinson has left [06:13:55] --- Simon Wilkinson has become available [06:29:58] --- kaj has left [06:45:24] --- Russ has become available [07:29:54] after implementing the B+tree lookups in Windows, I found that the DNLC functionality was slower than the B+tree lookups [07:32:54] It should really get turned off on Linux, because the dcache provides the same functionality, only faster. [08:19:20] jaltman - if you could make a 64-bit version of the afslogon.dll, I'll give that a shot, too [08:42:49] pretty sure the no-dnlc case does work, and istr there was a command line afsd option for it at some point [10:20:51] --- jaltman has left: Disconnected [11:56:16] --- Russ has left: Disconnected [12:11:53] --- Russ has become available [14:15:41] --- deason has become available [15:02:12] --- haba has become available [15:37:44] OpenAFS 1.4.12+dfsg-3 is now available for Ubuntu at https://launchpad.net/~openafs/+archive/stable . [16:00:14] dfsg ? [16:01:02] 1.4.12 has to be repacked for Debian to remove APSL-covered code. This won't be a problem for 1.5, although I will probably still repack it to remove the WINNT directory, which comes close to halving the size of the source package. [16:01:49] Or I could try to talk Derrick and Jeff into separating the WINNT directory into a separate source tarball like we do with docs. I haven't decided yet if I think that's a good idea. [16:03:33] BTW, are TechNotes-WIN9X and the NTMake9x and lwp/NTMake9x files obsolete? [16:04:10] Likewise README-WIN9X and probably Makefile.djgpp? [16:04:17] I haven't heard anything about anyone using that stuff in quite a while. [16:05:15] Oh, am I on the wrong branch? [16:05:32] Oh, I am. [16:05:33] Never mind. [16:05:54] (Answer: yes, they are obsolete, which is why they have all been removed already.) [16:06:48] I had a moment like, that, when I was boggling at some code that was #ifdef AFS_DARWIN_ENV in src/afs/LINUX. Went to remove it, discovered I had been looking at an old branch. [16:51:47] --- haba has left [17:16:24] --- jaltman has become available [18:44:11] anyone familiar with ubik (both in ptserver and vlserver) complaining about error 6 when trying to sync with the sync site? [18:44:46] one of our db servers failed, this is what the replacement hardware is complaining about. [18:54:24] osi_Alloc does not sleep? [18:57:55] It sure looks like it does. [18:59:47] Right. Apparently I am blind. Need more caffeine ... [19:17:59] --- jaltman has left: Replaced by new connection [19:18:00] --- jaltman has become available [19:31:44] Is there any reason to keep using (afs_proc_t)->p_rcred instead of p_ucred (for the AFS_XBSD_ENV case, at least)? [20:34:26] kula: what version of openafs? [21:32:15] --- jaltman has left: Disconnected [21:35:55] --- jaltman has become available [22:09:55] --- Russ has left: Disconnected [22:25:56] --- Born Fool has become available [22:36:56] --- jaltman has left: Disconnected [23:01:49] shadow_gmailcom: on the db servers? 1.4.8. i even found the rt ticket that explains the problem. it's the same thing i ran into in one of our test cells. [23:02:11] RT#123723 [23:02:43] --- Born Fool has left [23:02:45] we're going to go to 1.4.12 soon.