**** BEGIN LOGGING AT Sun Mar 12 10:59:56 2006 Mar 12 11:18:27 hpmf... why do powerbuttons always have a light on them Mar 12 11:18:52 irresistable to small kids ... my son just pressed the power on the slug ;) Mar 12 11:27:29 heh Mar 12 11:35:45 why on earth did i just install php + mysql on the slug? Mar 12 11:36:22 not performing too well ? Mar 12 12:00:17 pepijn: and? Mar 12 12:00:35 morning all Mar 12 12:01:31 morning Mar 12 12:06:50 had problems with file-native the other day, I now started from scratch and everything hums along, no idea what was wrong Mar 12 12:13:12 this was mentioned here and on the mailinglist Mar 12 12:13:53 yup Mar 12 12:13:59 as well as dropbear Mar 12 12:14:14 tbm: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=356480 Mar 12 12:14:24 yes, dropbear 0.48 had the same problem Mar 12 12:14:50 pepijn: yeah, just saw it, thanks Mar 12 12:15:57 any clue about the TERM problem? Mar 12 12:16:22 i mean, your screendumps look ok, so what is different here? Mar 12 12:18:15 just woke up, haven't read the report yet. hold on a bit Mar 12 12:18:57 * pepijn is looking for a small memory footprint bind9 replacement Mar 12 12:19:00 any suggestions? Mar 12 12:19:13 local domain, jsut a small number of hosts Mar 12 12:19:21 not meant to be served publically Mar 12 12:21:36 I've heard good things about pdns but I've no idea how small it is Mar 12 12:21:46 PowerDNS tries to emphasize speed and security. Mar 12 12:22:02 yes, looking at that one now, i am running ldap, so that might make a nice combo Mar 12 12:22:04 not sure if that includes size, but apparently there are different backends (I suppose one might be light-wight) Mar 12 12:23:31 maradns looks like another alternative Mar 12 12:23:43 problem: i'd like to have dhcp-ddns support Mar 12 12:23:45 ;) Mar 12 12:25:09 pepijn: i know the file-native was mentioned; the suggested fix of removing zlib did not work for me, as I needed to move to this new system I started from the very scratch and it worked, wanted to mention that Mar 12 12:25:42 k, i started from scratch as well, so no idea wether the zlib fix will work ;) Mar 12 12:29:09 pepijn: as I had some issues upgrading yesterday, I've installed dapper rel 0.5 or whatever from scratch Mar 12 12:29:41 pepijn: did you install via the serial console, or when did you see those TERM problems? Mar 12 12:29:47 actually smp seems to be detected automagically by now: Mar 12 12:29:48 frans@LinuxPC:/etc/X11$ cat /proc/version Mar 12 12:29:48 Linux version 2.6.15-18-686 (buildd@rothera) (gcc version 4.0.3 20060304 (prerelease) (Ubuntu 4.0.2-10ubuntu1)) #1 SMP PREEMPT Thu Mar 9 15:29:22 UTC 2006 Mar 12 12:30:40 eFfeM: yes, i figured that, given the removal of the seperate SMP linux-image Mar 12 12:30:53 Linux version 2.6.15-18-386 (buildd@rothera) (gcc version 4.0.3 20060304 (prerelease) (Ubuntu 4.0.2-10ubuntu1)) #1 PREEMPT Thu Mar 9 14:41:49 UTC 2006 Mar 12 12:30:54 tbm: no serial console available, so it was ssh Mar 12 12:31:03 * NAiL doesn't run SMP Mar 12 12:31:13 pepijn: odd. Because I know serial console is garbled, but SH should be fine Mar 12 12:31:16 SSH Mar 12 12:31:25 ssh in iTerm from MacOSX, but also ssh from console on Ubuntu/breezy Mar 12 12:31:40 basically same issues Mar 12 12:32:14 ssh from iTerm on MacOSX works ok when accessing Debian Mar 12 12:32:27 just not within the d-i environment Mar 12 12:33:33 anyone know the command to merge my mt key in the database by heart ? Mar 12 12:33:44 it isn't done that way any longer Mar 12 12:34:12 place the key in ~/.monotone/keys/somekey Mar 12 12:34:22 ok undersood Mar 12 12:34:31 s/ood/tood. Mar 12 12:35:35 done, now let it hum along Mar 12 12:51:17 tbm: fwiw, iTerm/MacOSX has TERM=xterm, while the Ubuntu console has TERM=linux Mar 12 13:01:00 pepijn: can you boot into d-i again (no need to do a full install, jut boot into it) and take a screenshot? Mar 12 13:04:24 yep Mar 12 13:20:49 hmm, i've been having some issues with the console on debian too. Mar 12 13:21:38 logging in through ssh, when i quit aptitude or nano, the cursor ends up half way up the screen, but the screen isn't cleared. Mar 12 13:22:00 so its a little hard to read the install log while apt is installing stuff Mar 12 13:23:26 * pepijn running poweroff on slug Mar 12 13:24:51 * pepijn has establised telnet access to redboot Mar 12 13:32:20 loaded ramdisk Mar 12 13:32:23 loaded kernel Mar 12 13:32:42 executing... Mar 12 13:34:45 guess it takes so long because of key generation ? Mar 12 13:35:02 that beep would be nice ;) Mar 12 13:36:10 yeah, key generation takes a while Mar 12 13:49:33 does bugs.debian.org accept attachments ? Mar 12 13:49:52 tbm: where do you want the screendump ? Mar 12 13:50:10 pepijn: how big is the file? Mar 12 13:52:29 72Kb (png) Mar 12 13:53:00 imho it would be best to have it stored along with the report Mar 12 13:53:22 running d-i a bit further to find out about the partitioning Mar 12 13:53:37 pepijn: yeah, please send it to 356480@bugs.debian.org Mar 12 13:55:25 you want to be cc-ed directly ? Mar 12 14:06:56 When choosing 'Erase entire disk: SCSI1 (0,0,0) (sda) - 257.9 MB USB 2.0 Mobile Disk' under Guided partitioning, I am presented a dialog with the following information: Mar 12 14:06:57 Failed to partition the selected disk Mar 12 14:06:57 This probably happened because the selected disk or free space is too small to be automatically partitioned. Mar 12 14:07:12 any logs to look at and include in the report ? Mar 12 14:08:04 uh.. you're trying to install debian on a 256mb usbstick? Mar 12 14:08:09 hehe Mar 12 14:08:34 tbm wanted me to run d-i, but I am not gonna overwrite my 250Gb USB disk ;) Mar 12 14:08:52 just test running this Mar 12 14:08:57 well, unless I'm mistaken, it won't work. Mar 12 14:09:10 It's simply, as it says, "too small" Mar 12 14:09:23 I am happy with the manual debootstrap I performed Mar 12 14:09:52 yes, i would think so, and i did include that info in the report Mar 12 14:15:12 k, reports sent Mar 12 14:15:18 gonna boot back Mar 12 14:16:13 ah, okay. that error makes sense I guess Mar 12 14:18:03 maybe it should be more extensive, as in what the minimum then actually is Mar 12 14:26:47 tbm: then you should have the screendump Mar 12 14:26:53 any good? ;) Mar 12 14:28:00 pepijn: interesting. thanks a lot Mar 12 14:31:27 np Mar 12 15:00:42 tbm: what environment are you ssh-ing from ? Mar 12 15:01:00 pepijn: standard Debian Mar 12 15:01:20 4405:tbm@deprecation: ~] echo $TERM Mar 12 15:01:20 xterm Mar 12 15:02:15 which ssh client? Mar 12 15:02:17 hi can anyone tell me how i can change the language on the system an change it to something else and utf-8 to iso... ?? Mar 12 15:09:46 kami22: don't know much about this, could be that iconv will help Mar 12 15:10:11 rebuilding from a fresh archive, but dropbear dependency cannot be resolved: Mar 12 15:10:24 ERROR: Nothing provides runtime dependency dropbear Mar 12 15:10:25 NOTE: no buildable providers for slugos-image Mar 12 15:10:25 ERROR: dependency slugos-image (for openslug-image) not satisfied Mar 12 15:10:33 but there is a packages/dropbear ?? Mar 12 15:10:42 anyone an idea? Mar 12 15:10:53 yeah me Mar 12 15:11:01 read the logs Mar 12 15:11:09 ok Mar 12 15:11:19 my guess is that it is the 'required' inb the dropbear .bb files Mar 12 15:11:23 change it into include Mar 12 15:12:14 i discussed this last friday or so as well in #openslug Mar 12 15:12:41 at that point I did get a different message, so I did not think this was the same problem Mar 12 15:13:00 according to koen bitbake is broken as it should deal with require Mar 12 15:13:35 hehe, should... but for me it doesn't Mar 12 15:13:50 yup, actually I see the same errors again Mar 12 15:15:54 it does contine with this change, also asked the Q in #oe Mar 12 16:51:47 discovered an interesting challenge, did an ipkg upgrade and it tried to upgrade dropbear, causing me to loose the connection which in turn stopped ipkg, so no way out except reboot. retrying now by running ipkg upgrade in a background process (nohupped) Mar 12 16:52:22 is there a better way around this? Mar 12 16:53:15 If you were on unslung, I'd tell you to just telnet in... but I bet you don't have telnetd on your system? Mar 12 16:53:36 nope, actually I could have enabled it before but I didn't Mar 12 16:53:58 the background solution seems to work though, but might need a reboot later on Mar 12 16:54:17 i'm getting a connection refused now but still a lot of activity Mar 12 16:56:07 upgrading dropbear shouldn't actually drop the connection though... Mar 12 16:56:35 It should technically just kill the listening process Mar 12 16:56:40 not the login processes Mar 12 17:02:20 NAiL: it did not work for he, it closed the connection, might be that I didn't wait long enough to upgrade things Mar 12 17:02:35 i'll do another reboot to see what happens Mar 12 17:03:23 NAiL: i am sure that that is the case for OpenSSH Mar 12 17:03:40 yeah, openssh is safe to upgrade Mar 12 17:04:02 (but openssh also takes like 6-7% ram for each of my login processes) Mar 12 17:04:09 i routinely upgrade ssh related things and *first* setup a secondary connection before closing the first (non-upgraded) one Mar 12 17:04:31 yes Mar 12 17:04:40 that way you have some way of recovering Mar 12 17:04:45 must be a killall dropbear vs. kill $(cat /var/run/dropbear.pid) thing Mar 12 17:05:22 btw. i have telnetd with root on pts/0 starting right after networking so i have a backdoor Mar 12 17:05:34 hehe Mar 12 17:05:38 maybe should put a hosts.deny in there too Mar 12 17:05:42 I have serial console (albeit with some work) Mar 12 17:05:47 hehe Mar 12 17:05:49 could do that as well Mar 12 17:05:53 that would be the best, yes Mar 12 17:06:00 that is a *good* solution :) Mar 12 17:06:31 my serial for some reason does not take input maybe something wrong with the cable Mar 12 17:06:42 need to check that with someone else Mar 12 17:06:44 yeah Mar 12 17:06:45 or faulty soldering Mar 12 17:07:03 I thought I had blown the input on my slug, until I tried a diff cable Mar 12 17:07:08 anyway the backdoor I thought of is using the usb<->ethernet cable I have Mar 12 17:07:49 but when I wanted to set that up, things were not working when the ixp drivers failed (probably it then becomes eth0 iso eth1 Mar 12 17:14:57 problem with dropbear is that it does not support kerberos login (which i use) Mar 12 17:17:11 although i do use dropbear on my dreambox and wrt54gl's Mar 12 17:49:05 03bzhou * 10unslung/make/rsync.mk: rsync upstream upgrade to 2.6.7 Mar 12 19:34:18 03bzhou * 10unslung/make/ (py-turbogears.mk py-sqlobject.mk py-paste.mk): changed py-paste py-sqlobject versioning scheme to not use _ Mar 12 20:32:23 woohoo - 2d3h19m uptime - better than my last server :) Mar 12 20:32:31 hehe Mar 12 20:32:39 so a newer kernel fixed your problems? Mar 12 20:32:42 yep Mar 12 20:32:43 dwery: ping? Mar 12 20:33:00 NAiL: host busy, answer might be delayed. Mar 12 20:33:00 :-D Mar 12 20:33:12 half the problem with the last server was the noise level... the nslu is nice and quiet, even with a 7200RPM hard drive hooked up to it. Mar 12 20:33:22 dwery: Have you tried compiling -rc6? Stuff doesn't apply cleanly, so I haven't gotten around to that yet. Mar 12 20:33:31 not yet Mar 12 20:33:46 philpem: definitely. I've got a 4200rpm laptop drive connected though. It's *very* silent :-D Mar 12 20:34:05 could do with getting an 80gig drive for it at some point. 15 isn't really big enough for a server. Mar 12 20:34:27 half the reason i bought a 3.5" casing was because the drives are so much cheaper :) Mar 12 20:34:44 the drive I'm using is "leftover" from my old laptop Mar 12 20:35:04 It's not exactly large (6GB IIRC), but that's plenty to run the debian stuff I need ;) Mar 12 20:35:17 I've got another slug that has a 120GB drive that I use for fileserving Mar 12 20:35:49 i've just installed php, mysql, cherokee (compiled on the slug) and torrentflux. Mar 12 20:35:54 nice Mar 12 20:36:18 so now i can leave the thing downloading.. um.. "the latest version of fedora core".. while i'm asleep. which is nice. Mar 12 20:36:29 hahaha Mar 12 20:36:31 yeah Mar 12 20:36:47 Naruto? ;) Mar 12 20:37:02 Naruto, Bleach and whatever else I see on BoxTorrents that looks interesting :) Mar 12 20:37:13 my isp likes to whine about people running their connections flat out from 6pm-midnight. fine, how's midnight to 6pm then? :D Mar 12 20:38:04 actually, i need to reinstall those hotplug scripts and test them. thing is, i've had this connection up for almost as long as the server and i don't feel like disconnecting. Mar 12 20:38:26 Got more or less complete Bleach & Naruto here Mar 12 20:38:35 Same here Mar 12 20:39:23 as well as Yu Yu Hakusho and OnePiece.. I'm at a friend's place, and he's kinda anime-fan Mar 12 20:39:39 Complete Bleach from 1 through 71, mix of LunarAnime and Dattebayo subs, Naruto 1 through 175 mix of ANBU, ANBU|AonE, AonE and Dattebayo subs. Mar 12 20:40:39 also got Blue Gender, YuYuHakusho s1/s2, InuYasha complete series, InuYasha manga scanslations, Mononoke Hime, and some other stuff too.. Mar 12 20:41:07 most of the stuff i've got is on my anidb userpage... i keep forgetting to update it though Mar 12 20:41:46 interestingly enough, no pirate software on any of my machines. Mar 12 20:42:10 hmm Mar 12 20:42:30 does winzip count as pirate software after 30 days? :-P Mar 12 20:42:55 technically... Mar 12 20:43:04 hrrm Mar 12 20:43:04 i've been using powerarchiver - www.powerarchiver.com Mar 12 20:43:30 gave up on winzip a long time ago. PA is a little quirky at times, but the price is right :) Mar 12 20:43:40 heh Mar 12 20:43:50 not free, but very cheap Mar 12 20:43:56 yeah Mar 12 20:45:02 also got Acronis TrueImage for a pretty good price - it was on a magazine CD and the serial number i got was valid for an upgrade to the latest version. after seeing the price (and lameness) of ghost, i bought it. Mar 12 20:45:10 s/bought it/bought trueimage/ Mar 12 20:45:10 philpem meant: also got Acronis TrueImage for a pretty good price - it was on a magazine CD and the serial number i got was valid for an upgrade to the latest version. after seeing the price (and lameness) of ghost, i bought trueimage. Mar 12 20:45:34 TrueImage? It's like symantec ghost? Mar 12 20:45:40 * NAiL hasn't really used ghost much Mar 12 20:45:57 worth every penny, especially when you know you've got an hour to get every damn thing off your laptop hard drive before it goes titsup.com. Mar 12 20:46:16 yeah Mar 12 20:46:26 I tried getting everything off a 3.5" drive yesterday Mar 12 20:46:31 "SMART FAILURE ALERT: Drive PRI:MASTER reports IMMINENT FAILURE. F1 to continue." Mar 12 20:46:42 unfortunately, it went belly-up in the process Mar 12 20:47:06 i copy my documents off onto the lan with knoppix, then image it if the drive still works. Mar 12 20:47:48 got a lovely error message on a laptop the other day. "Graphics card failure. F1 to continue." -- the clincher was that this error message was displayed on the lcd... Mar 12 20:48:01 rebooted and it was fine. weird little machine... Mar 12 20:48:13 heh Mar 12 20:48:29 * mwester loves ghost -- just wishes it could handle UNIX filesystems better... Mar 12 20:49:04 trueimage is neat, it mounts a backup image as a virtual drive. Mar 12 20:49:21 But doesn't that require drivers? Mar 12 20:49:26 which means one of the many windows ext2 readers will work fine to read the contents of an image Mar 12 20:49:47 mwester: yes, they get installed when you install TW Mar 12 20:49:50 s/TW/TI/ Mar 12 20:49:52 philpem meant: mwester: yes, they get installed when you install TI Mar 12 20:50:03 doesn't help much when you use XFS/JFS/etc though Mar 12 20:50:16 That's unfortunate. I need to be able to bacup and recover hard drives on systems that cannot boot into any OS Mar 12 20:50:39 TI has two operation modes though - the normal windows-hosted mode and the "rescue cd" mode. Mar 12 20:51:07 I currently boot using PXE (the boot server is a slug, of course) and the boot image contains ghost on a ramdisk floppy image. Mar 12 20:51:16 boot off the rescue cd and you get a full network-enabled version of trueimage that can read and write images on samba shares. Mar 12 20:51:22 don't think you can netboot it though Mar 12 20:52:15 Still, it might be worth a look - that's not a free package, is it? Mar 12 20:52:16 can't say i tried though - the first thing i disabled on the A8V was the LAN bootrom (what's the point when the machine has 100GB worth of hard drives, plenty of space for an OS) and then the SATA and RAID controllers. Mar 12 20:52:32 mwester: trueimage? unfortunately not. there's a demo version available though iirc. Mar 12 20:53:39 I'll check it out... might be a good addition for some of the machines that have CDROM drives if it does a better job with Linux and BSD fs. - thanks. Mar 12 20:54:31 mwester: www.acronis.com Mar 12 20:54:59 best thing is it doesn't throw out those "unknown fatal error" messages ghost is so fond of :) Mar 12 22:26:18 03jp30 * 10unslung/Makefile: optware: promote py-urwid, w3m is now cross Mar 13 01:07:07 NAiLzZz: rc6 applied fine here. Mar 13 07:03:19 03bzhou * 10unslung/make/ (py-paste.mk py-sqlobject.mk py-turbogears.mk): Mar 13 07:03:19 complex VERSION= line not picked up by scripts/optware-autoclean.pl, Mar 13 07:03:19 which is the real reason causing py-paste py-sqlobject to build everytime, Mar 13 07:03:19 simplify VERSION line to work around that Mar 13 07:34:34 03oleon * 10unslung/ (make/libgc.mk sources/libgc/backtrace.patch): libgc: fix missing backtrace() on wl500g Mar 13 09:39:47 mornin' .. could anybody tell me more about usb disk-powerdown? how to (and possible) to controle? should supporting hardware do this automatically? Mar 13 09:45:22 http://www.nslu2-linux.org/wiki/FAQ/SpinDownUSBHarddisks Mar 13 10:05:51 03oleon * 10unslung/Makefile: optware: demote w3m on wl500g - until uClibc is upgraded libgc will be broken as it needs __data_start pointer and test_and_set() Mar 13 10:11:35 Thx Evil ... seems to be a topic for discussions ;) Mar 13 10:13:29 np. not really.. either it works or it doesn't ;) Mar 13 10:22:29 ;) hehe evil: i've a debianslugle with external WesternDigital 320 Essential Drive ... Mar 13 10:24:41 which method do you suggest, evil? **** ENDING LOGGING AT Mon Mar 13 10:59:57 2006