Back to Home / #linode / 2006 / 05 / Prev Day | Next Day
#linode IRC Logs for 2006-05-16

---Logopened Tue May 16 00:00:17 2006
00:26<fo0bar>caker: woo, finnix now boots perfectly as a dom0
00:46<@caker>fo0bar: know what eskimos get from sitting on the ice too long?
00:47<@caker>polaroids
00:47<taupehat>haha
00:49<@caker>fo0bar: anyhow, congrats
00:49<@caker>need to announce that on xen-user, er something
00:49<@caker>you're probably more up to date than their demo iso
00:49<Eman>linode needs a canadian counterpart... for... things that arent legal in the us
00:51<@caker>like?
00:51<taupehat>dvdcss?
00:51<Eman>torrent trackers :D
00:52<taupehat>heh
00:53<fo0bar>caker: yes, I've heard a major complaint on the lists is the demo cd isn't updated with each release
00:53<TheFirst>hmmm not legal in the US ... wouldn't it be quicker to list the legal stuff?
00:53<TheFirst>Legal: <empty set>
00:54<fo0bar>caker: besides, I'm not going to replace the existing finnix kernel with the dom0, for a couple reasons
00:55<fo0bar>#1, I'm currently at the mercy of xen development when choosing what kernel to release. I'm already bitter about getting burned by stuff like that (unionfs), and thankfully SKAS patches are released about 24 hours after a kernel
00:55|-|Marcel [~marcel@jharkema.xs4all.nl] has joined #linode
00:56<fo0bar>#2, AFAICT, you have to specify the hardcoded memory to be used in the dom0 on the kernel commandline (I may be wrong about that?)
00:57<fo0bar>#3, xen doesn't play well with isolinux. I got it to work, but you can't pass any additional kernel paramaters on the boot line. I would have to go to iso/grub, which makes it harder to boot with custom options
00:57|-|konoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
00:58<fo0bar>(IE, instead of typing "finnix toram", you would do esc,e,down,down,e," toram<enter>",b
00:59<fo0bar>#4, cosmetic, but the hypervisor lines screw up the ability to run "quiet" mode on the kernel, and I haven't figured out how to make the hypervisor "quiet"
00:59<fo0bar>#5, pie
01:00<taupehat>hmm
01:00<taupehat>pie is legal in the US
01:01<@caker>fo0bar: #2, if you don't specify, it uses all avail, and balloons down on subsequent domU launches, down to the min default, which is somwehre around 128 or 256M
01:02<@caker>like you said, if I were you, I wouldn't even consider replacing Finnix with a xen+xenlinux combo
01:02<@caker>maybe a side-release, but not for the main thing
01:04<fo0bar>yeah, this was more of a "what if" exercise
01:04<@caker>AFA #4, I *think* you can add com1=null or somesuch on the xen.gz bootloader line
01:06<fo0bar>oh, #6, the xen tools package requires udev, which I'm not quite ready to move to yet (mainly because udev takes FOREVER to start up)
01:06<@caker>oh, it's console=none, from what I read in an ML archive
01:06<fo0bar>on a distro that takes 20 seconds on average to boot, adding 30 seconds to populate /dev really sucks
01:07<gpd>q: if you were to experiment with BSD - which would you choose: free|open|other ?
01:07<fo0bar>gpd: depends on your goal
01:07<@caker>another really weird thing is the hotplug scripts for starting a domain try to get major/minor numbers for devices you're going to bring up inside the domU from the dom0's /dev .. even if they don't exist (stupid)
01:07<fo0bar>"general/desktop", I would say free
01:07<fo0bar>for server, open
01:07<@caker>like, our hosts don't have any ide devices, so no /dev/hd**
01:08<@caker>it works, but just generates a warning in xend.log
01:08<fo0bar>if you want bsd on your toaster, net
01:08<fo0bar>caker: weird
01:08<gpd>fo0bar: my live disk network card issues prompt experimentation with BSD -- never used it b4... but a friend is a FREE fan--- just checking if he is crazy (box is to be apache server)
01:10<fo0bar>caker: oh, I also got rid of the requirement to use a special initrd for uml/xenu booting. now you can just copy the main /cdrom/isolinux/minirt initrd and it "just works" under any environment
01:10<@caker>minirt?
01:10<@caker>miniroot?
01:10<fo0bar>the great thing is at one point after some changes, I looked at the initrd and thought "hey, this will work with uml/xenu now!"
01:11<fo0bar>caker: it's a throwback from that short period when finnix was based on knoppix
01:11<fo0bar>dunno why it's named that
01:11[~]caker wants netbootable finnix from tftp or somesuch, or at least a useful initrd via tftp
01:12<fo0bar>I keep meaning to look into that
01:12<@caker>since netbooting requires tftp anyhow, any initrds move across via that, no nfsroot nonesense needed
01:12<@caker>s/any/and/
01:12<fo0bar>well, where would you get the compressed filesystem from?
01:13<@caker>what I've done in the past is just an initrd with my tools on it, along with some ramdisk_size=xxx cludging
01:13<@caker>DEFAULT pxelinux.cfg/vmlinuz-2.6.15.4-bigmem-skas8.2-skas3-v8.2 ramdisk_size=61440 initrd=initrd-lvm2.gz lvm2rescue root=/dev/md0 elevator=cfq console=tty0 console=ttyS0,115200
01:13<fo0bar>well, you could do that with the existing finnix initrd (just pass the "debug" parameter and one of the first things you get is an ash prompt with busybox available)
01:14<@caker>ahh
01:14<@caker>good stuff
01:14<Spads>http://linuxmafia.com/faq/Security/fish-protocol.html <-- nice stuff
01:14<@caker>yeah, I built one that's basically busybox
01:15<fo0bar>but I was thinking initrd via tftp, which does an nfsmount, then mounts the squashfs filesystem (seen via nfs) like normal
01:15<@caker>right
01:16<fo0bar>however, I sense pain there, because it seems every couple weeks there's a "unionfs over nfs or nfs over unionfs or something craps out" thread on the unionfs lists
01:16<Spads>FISH
01:16<fo0bar>GO
01:17<fo0bar>> What's wrong with sftp?
01:17<fo0bar>1. Still seldom enabled on the server end. 2. Supported by very few SSH packages and on few host operating systems on the client end.
01:17<fo0bar>nope, the email stamp says 2003, not 1999
01:17<fo0bar>hmm
01:18<Spads>only tatu ylonen's implementation and the openssh guys have it
01:19<fo0bar>OH NOES DROPBEAR CAN'T DO SFTP
01:19<Spads>and it's a good idea to just leave it off, since the sort of people who want it are the sort of people who want a windows GUI thing
01:19<Spads>and that's the other problem with sftp
01:19<Spads>you're enabling a weird protocol thing without any real benefit
01:20|-|konoko [~Vampire@fhso-hst-240-67.hst.fhso.ch] has joined #linode
01:20<TheFirst>ftp w/ tls works fine....and there are plenty of windows clients that do sftp + ftp w/ tls
01:21<Spads>tls or no, FTP is the goatse.cx of the Internet
01:21<Spads>it bends over. it opens wide.
01:21[~]fo0bar installs telnets
01:21<TheFirst>at least with sftp/tls your passwords aren't sent cleartext
01:22<Spads>and at least you can't see the goatse guy's *face*
01:26|-|dsoul [darksoul@vice.ii.uj.edu.pl] has joined #linode
01:36<warewolf>ftp/tls breaks natting firewalls
01:36<warewolf>because the natting firewalls can't dick with the PORT x.y.z.p.d.q commands
01:37<warewolf>sftp or scp over ssh is much better in that respect
01:37<warewolf>also, if you're looking for a good windows ftp/sftp/scp/rsync client, use FileZilla.
01:37<Spads>FISH is even better
01:37<Spads>it's like scp++
01:38<warewolf>my dad (who is 66) uses FileZilla. It is extremely similar visually to WS_FTP, so he likes it a lot.
01:57<TheFirst>warewolf: i've not had any problem with ftp/tls behind a nat
01:57<Eman>ftp needs to die
01:58|-|womble [~mpalmer@106.135.233.220.exetel.com.au] has quit [Quit: What's behind the round window...]
01:59<warewolf>TheFirst: then your control channel is not encrypted.
02:00<warewolf>TheFirst: and you are probally not really using tls.
02:00<TheFirst>uhhh yah it is....
02:01<warewolf>are you using active mode ftp or passive mode ftp?
02:01<TheFirst>[R] Connected. Negotiating TLSv1 session..
02:01<TheFirst>[R] 150 Opening ASCII mode data connection for directory listing using SSL/TLS.
02:01<TheFirst>[R] TLSv1 negotiation successful...
02:02<TheFirst>[R] TLSv1 encrypted session using cipher DHE-DSS-AES256-SHA (256 bits)
02:02<TheFirst>:)
02:02<TheFirst>that's with pasv...never need to use port
02:03<TheFirst>but it does work using port also ...
02:03<warewolf>http://www.isaserver.org/articles/FTPTLS_Friendly_Firewalls.html
02:03<warewolf> 7.3. Network Address Translators
02:03<warewolf> NAT firewalls will not work for secure FTP if the NAT will affect
02:03<warewolf> the PORT address or the PASV response address.
02:04<warewolf>that's an IETF RFC btw.
02:04<TheFirst>sure...but you can always specify that option in the ftp conf so that that is not the case
02:05<TheFirst>for nat it's a matter of port forward, set the ip in the ftp conf and a passive port range...works every time
02:06|-|kriby_ [~kvandivo@ppp-70-225-161-199.dsl.chmpil.ameritech.net] has joined #linode
02:13|-|kriby [~kvandivo@ppp-70-225-162-153.dsl.chmpil.ameritech.net] has quit [Ping timeout: 480 seconds]
04:49<linbot>New news from forums: Crappy Cron in General Discussion <http://www.linode.com/forums/viewtopic.php?t=2283>
05:16|-|jekil [~alessandr@pcport.univ.trieste.it] has joined #linode
06:33|-|jekil [~alessandr@pcport.univ.trieste.it] has quit [Quit: Leaving]
06:55|-|Dreamr3 [~Dreamer3@0-1pool107-167.nas33.chicago3.il.us.da.qwest.net] has joined #linode
07:00|-|Dreamer3 [~Dreamer3@0-1pool106-157.nas33.chicago3.il.us.da.qwest.net] has quit [Ping timeout: 480 seconds]
08:56|-|linville [~linville@nat-pool-rdu.redhat.com] has joined #linode
09:38|-|yjmwrk [ident_name@bzq-84-108-241-136.cablep.bezeqint.net] has joined #linode
09:41|-|odedra [ident_name@bzq-84-108-241-136.cablep.bezeqint.net] has quit [Ping timeout: 480 seconds]
09:59|-|jekil [~alessandr@82.52.171.198] has joined #linode
10:00|-|spr [~spr@ramona.cs.byu.edu] has joined #linode
10:40|-|Damaja1 [revans@62-64-223-81.dynamic.dial.as9105.com] has joined #linode
10:56|-|konoko [~Vampire@fhso-hst-240-67.hst.fhso.ch] has quit [Ping timeout: 480 seconds]
11:26|-|konoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has joined #linode
11:33|-|Netsplit oxygen.oftc.net <-> xenon.oftc.net quits: dc0e, jekil, efudd, Quizzer, Redgore, emcnabb, spr, TheFirst, kriby_, luis, (+1 more, use /NETSPLIT to show all of them)
11:34|-|Netsplit over, joins: Damaja1
11:34|-|Netsplit over, joins: spr
11:34|-|Netsplit over, joins: jekil
11:34|-|Netsplit over, joins: kriby_
11:34|-|Netsplit over, joins: luis
11:34|-|Netsplit over, joins: dc0e
11:34|-|Netsplit over, joins: TheFirst
11:34|-|Netsplit over, joins: efudd
11:34|-|Netsplit over, joins: Quizzer
11:34|-|Netsplit over, joins: emcnabb
11:34|-|Netsplit over, joins: Redgore
11:35|-|iggy [~iggy@12.45.184.235] has quit [Remote host closed the connection]
11:35|-|luis_ [~fortytwo@84.90.17.8] has joined #linode
11:37|-|jekil [~alessandr@82.52.171.198] has quit [Ping timeout: 480 seconds]
11:37|-|luis [~fortytwo@84.90.17.8] has quit [Ping timeout: 480 seconds]
11:43|-|iggy [~iggy@12.45.184.235] has joined #linode
12:02|-|Damaja1 [revans@62-64-223-81.dynamic.dial.as9105.com] has quit [Quit: ]
12:38<@mikegrb>mmm cake
12:38<npmr>man of cake?
12:40<npmr>http://dorothy.movealong.org/munin/movealong.org/dorothy.movealong.org-memory.html <-- caker, see memory usage by month. sharp rise in "committed" april 20th corresponds to reboot using 2.6.16.1-linode18-bb1
12:41<npmr>caker, is this a known change of calculation in that kernel vs. previous versions or is this a behavioral change in userspace under the new kernel
12:47|-|FireSlash [~FireSlash@0-1pool106-33.nas23.kansas-city2.mo.us.da.qwest.net] has joined #linode
12:53|-|sonorous [simon@attenuate.org] has left #linode []
13:18|-|_GoldwiN_ [~KLBKvTurk@85.100.212.66] has joined #linode
13:18<_GoldwiN_>greeting
13:29|-|jekil [~alessandr@host198-171.pool8252.interbusiness.it] has joined #linode
13:37|-|FireSlash [~FireSlash@0-1pool106-33.nas23.kansas-city2.mo.us.da.qwest.net] has quit [Quit: Leaving]
13:37[~]_GoldwiN_ FROM YOU, TURKISH, KNOW, THERE,
13:41|-|_GoldwiN_ [~KLBKvTurk@85.100.212.66] has quit [Quit: Kelebek v.Türkçe Madem Türksün Kelebek v.Türkçe Kullan Herkes Ürksün.. Download : http://www.kelebek.gen.tr Adresinde...]
13:58<@mikegrb>npmr: change in actual amount or change in %?
13:59<@mikegrb>if the new kernel has more stuff enabled, it would use more ram
13:59<@mikegrb>though if the kernel memory is counted in the amount, then it could change the value even if not shown in %
14:16<npmr>mikegrb, %?
14:16<@mikegrb>well % commited vs X MB committed
14:16<@mikegrb>sirry, I didn't look at the graph
14:16<npmr>mikegrb, i guess you could say committed memory went from ~90% of mem+swap to about 160%
14:17<npmr>and i really doubt the kernel is using that much, even if it is now counting itself
14:17<npmr>you know what committed is, right?
14:18|-|jekil [~alessandr@host198-171.pool8252.interbusiness.it] has quit [Ping timeout: 480 seconds]
14:19<npmr>earlier today, when the swap usage dropped, i had restarted a bunch of userspace
14:19<npmr>that's all that was
14:19<npmr>i had expected committed to drop too, but no
14:19|-|kokoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has joined #linode
14:19<npmr>not by much anyway
14:19|-|konoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
14:21<Spads>npmr: do you remember the emad kernel module?
14:21<npmr>there was an emad module?
14:21<Spads>http://72.14.207.104/search?q=cache:sfEwzSSibE8J:psax.org/pipermail/psax-commit/2005-May.txt+falling+emads&hl=en&gl=us&ct=clnk&cd=4&client=firefox-a
14:21<Spads>we're trying to find the actual source
14:22<npmr>no se
14:22<npmr>emad.c, right?
14:23<Spads>I guess
14:23<Spads>it's hard to parse that message
14:23<Spads>because it swallowed whitespace
14:23<npmr>i don't remember it, but i'll search anyway
14:23<Spads>RCS file: /mnt/fat/cvs/emad/dave.c,v
14:24<Spads>odd
14:24<Spads>emad mentioned it at troncon
14:24<Spads>and none of us believed him
14:24<npmr>In directory doublethink:/mnt/fat/home/dave/work/emad Added Files: emad.c
14:25<npmr>Makefile.2.4 DELETED --- --- dave.c DELETED --- --- old-dave.c DELETED
14:25<npmr>so yeah, emad.c
14:26<Spads>aha
14:26|-|gsf [BiZEq5Lb6m@panix3.panix.com] has quit [Ping timeout: 480 seconds]
14:26<npmr>[inkblot@goose:~]$ find / -name emad.c 2> /dev/null
14:26<npmr>[inkblot@goose:~]$
14:27<Spads>damb
14:27|-|anderiv [~anderiv@207.67.87.34] has quit [Ping timeout: 480 seconds]
14:28<npmr>likewise on dorothy
14:41|-|anderiv [~anderiv@207.67.87.34] has joined #linode
14:49|-|spr [~spr@ramona.cs.byu.edu] has quit [Remote host closed the connection]
14:50|-|spr [~spr@ramona.cs.byu.edu] has joined #linode
14:51|-|gsf [dQe3KoiQCq@panix3.panix.com] has joined #linode
14:51|-|afv-13 [~adrianvv@rkdp-ip-nas-1-p435.telkom-ipnet.co.za] has joined #linode
15:01|-|Battousai [~bryan@216-164-28-158.c3-0.drf-ubr1.atw-drf.pa.cable.rcn.com] has joined #linode
15:36|-|coumbes [~coumbes@12.43.95.191] has joined #linode
15:52|-|linville [~linville@nat-pool-rdu.redhat.com] has quit [Quit: Leaving]
16:10|-|jekil [~alessandr@host198-171.pool8252.interbusiness.it] has joined #linode
16:10|-|Marcel [~marcel@jharkema.xs4all.nl] has quit [Quit: Leaving]
16:54|-|jekil [~alessandr@host198-171.pool8252.interbusiness.it] has quit [Ping timeout: 480 seconds]
17:14|-|nybble [~nybble@d36-38-252.home1.cgocable.net] has joined #linode
17:35|-|spr [~spr@ramona.cs.byu.edu] has quit [Quit: "Ah, the bounce has gone from his bungie"]
17:44|-|thoth39 [~hm@200216077231.user.veloxzone.com.br] has joined #linode
17:51<linbot>New news from forums: Reboot: host50 in System and Network Status <http://www.linode.com/forums/viewtopic.php?t=2284>
18:04|-|djayc [~40caa041@webuser.linode.com] has joined #linode
18:05<djayc>Is there something wrong with host50? I noticed it was rebooted earlier.. but it seems down again
18:06<thoth39><linbot> New news from forums: Reboot: host50 in System and Network Status <http://www.linode.com/forums/viewtopic.php?t=2284>
18:07<djayc>what was the timestamp on that?
18:07<djayc>the forum says 5:48PM, but I'm not sure what timezone
18:07<Eman>[6:51:31PM] EST
18:08<Battousai>yeah, roughly 20 mins ago on irc
18:08<djayc>ah ok .. so thats new..
18:08<djayc>oh I see.. GMT - 5
18:08<darkbeholder>15 mins before you joined
18:08<djayc>great.. thanks guys
18:08<djayc>my linode hasnt come up yet, so I didnt know if I should manually issue a boot or not
18:26|-|pyrokenx [~pyrokenx@67-20-173-43.chvlva.adelphia.net] has joined #linode
18:27|-|pyrokenx [~pyrokenx@67-20-173-43.chvlva.adelphia.net] has quit [Quit: ]
18:47|-|jekil [~alessandr@host236-170.pool8250.interbusiness.it] has joined #linode
18:56|-|Dreamrr_3 [~Dreamer3@0-3pool252-227.nas33.chicago3.il.us.da.qwest.net] has joined #linode
18:59|-|coumbes [~coumbes@12.43.95.191] has quit [Quit: Leaving]
19:02|-|Dreamr3 [~Dreamer3@0-1pool107-167.nas33.chicago3.il.us.da.qwest.net] has quit [Ping timeout: 480 seconds]
19:17|-|jekil [~alessandr@host236-170.pool8250.interbusiness.it] has quit [Quit: Leaving]
19:25|-|Battousai [~bryan@216-164-28-158.c3-0.drf-ubr1.atw-drf.pa.cable.rcn.com] has quit [Quit: Leaving]
19:29|-|Battousai [~bryan@216-164-28-158.c3-0.drf-ubr1.atw-drf.pa.cable.rcn.com] has joined #linode
20:59|-|okokok [~Vampire@fhso-hst-240-67.hst.fhso.ch] has joined #linode
21:00|-|konoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has joined #linode
21:06|-|kokoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
21:07|-|okokok [~Vampire@fhso-hst-240-67.hst.fhso.ch] has quit [Ping timeout: 480 seconds]
21:11[~]caker discovers partprobe
21:11<@caker>^-- lets the kernel know of partition changes of in-use devices *without* rebooting
21:12<Spads>nice
21:23|-|kokoko [~Vampire@fhso-hst-240-67.hst.fhso.ch] has joined #linode
21:28<Dreamrr_3>caker: is it a big leap to go from 160 to 240? i have to move servers, right?
21:28|-|okokok [~Vampire@80-218-157-65.dclient.hispeed.ch] has joined #linode
21:28<@caker>Dreamrr_3: pretty easy -- after we configure a migration, you shutdown, click the migrate button, and $minutes later, boot back up (after resizing or whatever)
21:29<Dreamrr_3>caker: what if something terrible breaks? is my data still around?
21:29<Dreamrr_3>(ie, on the original host)
21:29<@caker>yes
21:29|-|konoko [~Vampire@80-218-157-65.dclient.hispeed.ch] has quit [Ping timeout: 480 seconds]
21:30<Dreamrr_3>will i still be in CA?
21:30<@caker>yeah
21:30<Dreamrr_3>any idea on time? 30 minutes? 1 hour?
21:30<@caker>just submit a ticket to get the ball rolling
21:30<@caker>for a migration?
21:30<Dreamrr_3>yeah
21:30<@caker>a few minutes per gig
21:33<Dreamrr_3>can you get me on another host that never goes down like the one i'm on now? :-)
21:34<@caker>none of the CA hosts are affected
21:34<Dreamrr_3>huh?
21:35|-|kokoko [~Vampire@fhso-hst-240-67.hst.fhso.ch] has quit [Ping timeout: 480 seconds]
21:36<Dreamrr_3>caker: we could do this tonight?
21:37<@caker>Dreamrr_3: send me a ticket
21:38|-|spr [~spr@c-67-169-251-207.hsd1.ut.comcast.net] has joined #linode
21:38<Dreamrr_3>yeah, i'm decided if i want to proceed
22:10|-|coumbes [~coumbes@12.43.95.191] has joined #linode
22:16<Dreamrr_3>i will if this shared hosting situation doesn't improve
22:17<Dreamrr_3>funny my VPS is approaching the cost of a dedicated
22:30<TheFirst>what's the problem if you dont mind sharing?
22:59|-|VS_ChanLog [~stats@ns.theshore.net] has left #linode [Rotating Logs]
22:59|-|VS_ChanLog [~stats@ns.theshore.net] has joined #linode
23:09|-|djayc [~40caa041@webuser.linode.com] has quit [Quit: CGI:IRC (Session timeout)]
23:20|-|thoth39 [~hm@200216077231.user.veloxzone.com.br] has quit [Quit: Leaving]
23:50|-|afv-13 [~adrianvv@rkdp-ip-nas-1-p435.telkom-ipnet.co.za] has quit [Quit: leaving]
---Logclosed Wed May 17 00:00:59 2006