=== fisted_ is now known as fisted
=== FireCrotch is now known as nickmoeck
=== a7ndrew_ is now known as a7ndrew
tohuwI just installed mysql. How do I run "mysql" if the user mysql has no password?03:55
tohuwoh, mysql -u root -p04:00
uvirtbotNew bug: #734645 in openssh (main) "package openssh-client 1:5.3p1-3ubuntu5 failed to install/upgrade: " [Undecided,New] https://launchpad.net/bugs/73464505:02
CherishWhy I use auth_digest can't login my web site?06:29
overriderI am having trouble with my 10.04 samba installation. How can i completely remove it, then reinstall it? apt-get purge does not do it, because it leaves /etc/samba and other files in place. Manually removing those also bad, because then a reinstall will not recreate those default files. Any clue? Thanks06:40
whalesaladAnyone know how to get the latest nginx package running on karmic? I need at least 8.4 and would rather not build from source (I enjoy using apt-get, not from ease of use, but for portability)06:57
EvilPhoenixwhalesalad:  no backports?07:10
whalesaladEvilPhoenix: ?07:10
EvilPhoenixs/no/did you check the/07:10
whalesaladah, not sure what those are07:10
EvilPhoenixthey're not-completely-supported updates07:10
whalesaladwhere can I look for something like that?07:11
EvilPhoenixpushed to an older distribution from a newer build07:11
EvilPhoenixafaik, it'd be added to your sources.list file07:11
EvilPhoenixyou were looking for nginx 8.4 or greater?07:11
whalesaladim doing an "apt-get upgrade" right now07:12
EvilPhoenixhate to say it, but according to packages.ubuntu.com (the site for the package lists), there's no nginx >= 0.7.x available in anything07:13
EvilPhoenixcept the natty repos and natty isnt released07:13
whalesaladyea that's retarded to be honest07:14
whalesalad8.54 is the current nginx stable07:14
EvilPhoenixwell let me explain it like the QA people explained it to me about backports07:14
EvilPhoenix"there's nobody willing to put the time nor the effort into making something that's proven stable on the newer builds compatible with the older builds"07:15
EvilPhoenixand also07:15
EvilPhoenixfrom experience07:16
EvilPhoenix"Something proven to work on a newer build might just completely fubar your computer"07:16
EvilPhoenix^  true statement07:16
EvilPhoenixso yeah, you might be stuck compiling from source ;/07:16
EvilPhoenix:/ *07:16
uvirtbotEvilPhoenix: Error: "true" is not a valid command.07:16
EvilPhoenixshut up bot07:16
EvilPhoenixgod, this bot's almost as bad as my loco team's bot >.>07:17
EvilPhoenixbut yeah, that's one reason backports arent really commonly done07:17
EvilPhoenixonly time i've seen a DECENT backport is with the 2.6.35 kernel backport from maverick => lucid07:18
whalesaladhow painless is it to do full distribution upgrades btw?07:18
whalesaladlike karmic -> lucid07:18
whalesaladnot that it helps in this specific case..07:18
EvilPhoenixdo you want my brutally honest opinion, or a somewhat honest opinion?07:18
EvilPhoenixor both07:18
whalesaladI have a production server in sweden that I am in control of07:18
whalesaladI can't afford a ton of crap, because I only have one to play with (they only have one prod server)07:19
whalesaladso I can't stage everything (updates)07:19
EvilPhoenixbasically you cant fubar it07:19
whalesaladbut at the same time, I have zero patience07:19
whalesaladI frequently ignore major updates to linux boxes07:19
EvilPhoenixlets put it this way07:19
whalesaladand I'd like ot get into the habit of keeping them fresh07:19
EvilPhoenixin my honest opinion...07:20
EvilPhoenixthe best upgrade is a fresh install and you restore data from backups07:20
EvilPhoenixthat's because there's ******ALWAYS****** the chance something goes wrong during upgrading07:20
whalesaladah, okay07:20
whalesaladthat makes sense07:20
EvilPhoenixi've successfully went from jaunty -> karmic upgrade07:21
EvilPhoenixthen the subsequent karmic -> lucid upgrade failed badly07:21
EvilPhoenixhundreds of broken things07:21
EvilPhoenixended up fresh installing anyways :P07:21
whalesaladright on07:21
whalesaladI need to start my own hosting company07:21
EvilPhoenixso yeah, you could do dist-upgrade if ya want07:21
EvilPhoenixbut be aware: you should   A L W A Y S   have a backup copy of everythign07:22
whalesaladwell the db server is external, and the code is in git07:22
EvilPhoenixJUST IN CASE the upgrade fails miserably07:22
whalesaladonly thing I need is config info07:22
whalesaladdo you happen to have any kindof backup in place for your /etc dir or anything like that?07:22
EvilPhoenixfor my boxes?07:23
whalesaladalso, hmm. this was my fail. I tried to install the 8.53 deb directly from natty or whatever07:24
whalesaladran into some issues... so tried to sorta undo all that now nginx appears as uninstalled07:24
whalesaladbut nginx is still installed (the karmic one), and it's still running07:24
EvilPhoenixRAID6 across 6 disks + local-net rsync on a weekly basis + monthly physical cloning of the drive07:24
EvilPhoenixthat's my setup ;P07:24
whalesaladsounds pretty tasty... I haven't graduated from VPS hosting yet07:25
EvilPhoenix(yes, i've got some mean hardware here :P)07:25
EvilPhoenixyeah, i've got 2 VPSes for my business07:25
EvilPhoenixthis box here though with the crazy setup like that...07:25
EvilPhoenixno, that box is my personal server box :P07:26
EvilPhoenixthat thing's too good for hosting things cept the stuff i personally need >:)07:27
EvilPhoenixbut yeah, VPS hosting is decent07:27
EvilPhoenixi personally cba to set it up07:27
EvilPhoenix(hence why I dont have a dedi box somewhere :P)07:27
lifelesskirkland: hi09:17
lifelesskirkland: (i realise its nuts-am for you). I'll ping you tomorrow - I want to chat about a lp bug on blueprints you filed09:17
=== jjohansen is now known as jj-afk
Cherishapache2-mpm-worker doesn't support php5 ?09:32
CherishI must use apache2-mpm-prefork ?09:32
_rubeniirc, yes09:38
sorenCherish: It's sort of the other way around, but the effect is the same.09:48
sorenCherish: It's not apache-mpm-worker that doesn't support php5. It's php5 that doesn't support apache-mpm-worker. php on UNIX isn't guaranteed to be thread safe, so you're stuck with prefork.09:49
Cherishsoren: maybe can use fastcgi10:01
sorenCherish: Sure.10:27
jamespagettx: ping10:33
ttxjamespage: pong10:34
HackeMatei can access my webdav server but i cant write on it11:12
Tophatanyone mind helping me setup an SNTP server?11:43
jetoleHey guys. I have a problem with libvirt failing to autostart some domains in 10.04.02 on x86_64. I traced both issues to one common factor, it seems libvirt is starting before networking is setup. Does anyone know how I can change when it boots using the new-ish /etc/init (since it no longer uses the init.d folder)?11:50
sorenjetole: start on network-device-up IFACE=eth011:51
sorenjetole: For instance.11:51
sorenjetole: Eerr..11:51
sorenjetole: net-device-up11:51
sorenjetole: Not network.11:51
jetoleso I replace start on runlevel [2345] with start on network-device-up IFACE=br0 ?11:52
sorenjetole: More likely you want IFACE=br011:52
sorennet-device-up, but yes.11:52
* jetole gives it a try11:52
jetolecan I specify two nics?11:52
sorenjetole: You might ant to do "start on runlevel [2345] and net-device-up IFACE=br0"11:52
jetoleI have NFS on SAN and DNS on WAN and both are relating to two different VMs11:53
sorenjetole: You might ant to do "start on runlevel [2345] and net-device-up IFACE=br0 and net-device-up IFACE=br1"11:53
jetolecan I add another and?11:53
sorenSorry, that came out weird.11:53
jetoleI knew what you meant11:53
sorenYeah, just like I did in the last one.11:53
jetolestart on runlevel [2345] and net-device-up IFACE=br0 and net-device-up IFACE=br111:54
jetolethat should run?... ah lemme reboot and see11:54
jetole*let me11:54
sorenLooks ok to me.11:55
jetolerebooting the server now11:55
* jetole thanks god for KVM over IP11:55
Tophatneeding to setup SNTP server to serve my routers, wanting to use ubuntu-server -- want to advertise local machine time instead of relaying public SNTP servers.  is this possible?11:56
jetoleAfter I get this running I need to go over to the data center and pluck a failed disk on a RAID 5 that used to have a hot spare11:56
jetoleTophat: don't don't about local machine time but have you considered NTP vs. SNTP (afaik they are different protocols)11:57
jetolesoren: well it worked though I still have one VM which didn't boot. Looking into the _why_ now11:58
jetolesoren: Do you know if there is a way to run it after either a) all mounts have finished or b) NFS mounts have finished?11:58
RoyKTophat: apt-get install ntp11:59
jetolethe I have _netdev as an option for the NFS mount in fstab but it seems like libvirt is still starting too soon11:59
RoyKjetole: sntp is a subset of ntp, an sntp client can query an ntp server12:00
* jetole choose openntp since it doesn't rely on every NIC the way ntp does12:00
RoyKrely on every nic?12:00
jetoleplus openntp was written by Theo from obsd and the obsd guys and they write good and secure software12:00
sorenjetole: Sure.12:00
sorenjetole: All mounts: filesystems12:00
sorenjetole: All nfs mounts: remote-filesystems12:01
jetolesoren: I appreciate it but don't get it. How would I re-write the following: start on runlevel [2345] and net-device-up IFACE=br0 and net-device-up IFACE=br112:01
sorenjetole: start on net-device-up IFACE=br0 and net-device-up IFACE=br1 and filesystems12:02
jetolesoren: thanks again12:02
jetoleRoyK: ntp seems linked to every nic on a system and if starts / stops on every nic etc and is honestly messy on systems with a lot of nics though I hear that has been sorta fixed if you edit the ntp.conf, it's still a default12:03
jetoleRoyK: openntp doesn't seem to have that issue12:03
jetolesoren: I now understand what your comment meant after you showed the example12:03
zul_good morning....everyone except for soren in which case good day12:04
=== zul_ is now known as zul
jetolesoren: I used the following string and libvirt didn't start. I only added " and filesystems". Do you know what might have happened?: start on runlevel [2345] and net-device-up IFACE=br0 and net-device-up IFACE=br1 and filesystems12:07
pmatulisjetole: yes, /etc/network/if-up.d/ntpdate is strange to have12:09
jetoleDoes anyone have a good doc on upstart?12:10
pmatulisjetole: i believe the upstream docs are best right now12:10
jetolesoren: crap. nevermind. it did start. it was just a late start but it seems like everything is working fine12:11
jetolesoren: thanks again for the help12:11
jetolepmatulis: don't know what "the upstream docs" is referring to but I am specifically looking for start on conditions12:12
sorenjetole: Sure.12:12
jetoleAh I am going to the data center for a couple of hours. bbl and thanks for all the help12:14
patdk-wkupstream would mean, debian, or the manainter of the program itself12:15
patdk-wkstupid bot12:16
uvirtbotNew bug: #734804 in squid (main) "package squid 2.7.STABLE7-1ubuntu12.2 failed to install/upgrade: el subproceso instalado el script post-installation devolvió el código de salida de error 1" [Undecided,New] https://launchpad.net/bugs/73480412:16
RoAkSoAxmorning all13:01
=== ogra is now known as Guest42034
=== Guest42034 is now known as ogra_
kirklandlifeless: howdy, here now13:47
polyvisualCan anyone suggest how I can monitor the amount of traffic between my web server and a range of IP addresses?13:51
polyvisualI more lightweight the better. I looked at vnstat but that's 'per interface'.13:53
RoyKI don't think that allows for filtering per IP address or range, though13:55
polyvisualCheers, but awstats is a bit too heavyweight.13:57
polyvisualThe server connects to a bunch of proxy servers, pulls a bunch of data, then closes the connection. All I need is the bandwidth for the range of ip addresses.13:58
kim0hallyn: hi there .. kvm-spice is crashing consistently with "kvm-spice: /build/buildd/qemu-kvm-spice-0.14.0+noroms/qemu-kvm.c:1724: kvm_mutex_unlock: Assertion `!cpu_single_env' failed"14:01
speakmanNot quite the right channel maybe, but do any of you have any idea how to trace what's causing this: http://pastebin.ubuntu.com/580065/ ?14:03
Caribou1speakman, MCE are most often caused by hardware issues. If the problem repeats itself frequently, you could enable crash dump capture to get a vmcore for analysis14:06
screen-xHi all. Installed lucid on new server, adding an nfs mount to fstab causes boot to fail. I booted with init=/bin/bash, and commented out the nfs line. Server then boots fine including networking.14:07
Caribou1You could also check with your hardware vendor if you have any kind of hardware support14:07
geekbriscreen-x: is it possible it just needs an option like nobootwait?14:07
screen-xgeekbri: very possible :)14:07
screen-xI havent had to use that option before, new requirement in lucid?14:08
geekbriscreen-x: its just an option that says the system does not have to report an error if the partition isn't found.14:08
screen-xgeekbri: thing is, its a necessary partition, so really I do want it to be present and mounted before  the system considers itself booted.14:09
screen-xs/partition/nfs mount/14:09
geekbriscreen-x: hrm yes i see14:09
patdk-wkuse _netdev instead of nobootwait14:09
kirklandlifeless: saw your comment on https://bugs.launchpad.net/bugs/48733714:10
uvirtbotLaunchpad bug 487337 in launchpad "bug status for bugs linked in a blueprint is not shown" [Critical,In progress]14:10
kirklandlifeless: your reading is fine;  that would be nice14:10
kirklandlifeless: thanks14:10
screen-xpatdk-wk: can't see _netdev in man page for fstab or mountall, use it as an option in fstab?14:11
patdk-wkit's even in the howto: https://help.ubuntu.com/community/SettingUpNFSHowTo14:11
geekbriscreen-x: it basically says "this device needs networking"14:11
screen-xgeekbri: isnt that implicit in type=nfs?14:12
uvirtbotNew bug: #565578 in chkrootkit (main) "chkutmp crashed with SIGSEGV in _Unwind_Backtrace()" [Medium,Incomplete] https://launchpad.net/bugs/56557814:12
screen-xI've put _netdev in fstab, rebooting now to see if it works.14:12
geekbriscreen-x: I guess not :)14:13
screen-xstill failing with "mount.nfs: DNS resolution failed for blah.blah.blah"14:15
speakmanCaribou1: It's completely new stuff with a fresh install of Ubuntu 10.10. Intel Xeon W3680 with 24GB Kingston ECC RAM on a Asus WS Revolution motherboard, 850W Cooler Master PSU and a CPU temp not above 45 degrees celcius.14:15
geekbriscreen-x:  hrm well thats strange.  just out of curiosity could you temporarily use the IP instead of the DNS address14:16
speakmanCaribou1: I have got mcelog installed now, but what do you suggest for further investigation? It happens randomly and has nothing to do with any load (yesterday, it happened while I was logged out and there is no server daemons or such running == completely idleing)14:16
patdk-wkor add the dns name and ip to /etc/hosts14:16
Caribou1By hardware issue, I did not mean hardware failure. I recently investigated an MCE which was caused by the F/W of the 10 GbE NIC14:17
geekbriyup thats only dangerous if you forget to remove it eventually :(14:17
speakmanCaribou1: I'm all ears! And since all parts are new, there should be no problems returning anything disfuctional!14:17
patdk-wkgeekbri, remove it? don't update it :)14:17
screen-xI'll try adding a hosts entry.14:17
geekbripatdk-wk: well if its something that has an actual DNS entry you dont want to have it always be over written on manually edited, especially if its a file system critical to the system :(  what if somebody changes the DNS entry but the host entry is still there!14:18
Caribou1Well, you might want to configure crash dump capture (aptitude install linux-crashdump I think) then, once you get a valid vmcore, have a look at it with the crash utility so see if there is any more information in the kernel ring buffer14:18
* patdk-wk just doesn't use dns at all for nfs14:19
patdk-wkor iscsi14:19
geekbriyeah i dont either14:19
patdk-wkbut for vmware it pretty much forced me too14:19
geekbrii started using KVM and im hoping we will no longer need vmware ;)14:20
patdk-wkI can't grow enough with kvm14:20
patdk-wkplus, the whole xen -> kvm switch really didn't sit with me well14:21
patdk-wkand I am not going go through another switch, kvm -> x14:21
BCS-SatoriIs there a way to refresh my partitions on /dev/md0 (mdadm drive) without rebooting so that the OS sees the drive?  If I do "parted /dev/md0" and print I can see the partition but the disk is missing when I do "ls /dev/md0*"14:21
geekbrihey fair enough, we aren't too big of an operation and the KVM stuff is only internal so hopefully it will do for us14:21
patdk-wkya, kvm just doesn't support memory overcommit well enough yet14:21
=== Caribou1 is now known as Caribou
screen-xBCS-Satori: partprobe14:22
BCS-Satoriscreen-x: I tried already, it did not detect the new partition14:23
BCS-Satoriscreen-x: I think its a mdadm refresh issue maybe?14:23
hallynkim0: still working on setting up a test env.  does every guest os do thta for you?  And only since version -ppa1?14:24
screen-xBCS-Satori: I'm not sure what else to do, maybe could restart the array if you arent using any of the other partitions?14:25
BCS-Satoriscreen-x: ya I have iscsitarget running on several partitions on /dev/md0.  Its in a cluster I could reboot figured there has to be a way to not reboot14:28
BCS-Satoriscreen-x: seems blockdev and hdparm give this error "BLKRRPART: Device or resource busy" which is probably why partprobe doesn't work14:29
pikenDoes anyone know of a way to get a list of all package dependencies including the dependencies of those dependencies?14:37
pikenfor example, build-essential package list and their dependencies.14:37
pikenI am trying to build a vmware studio image but it has no access to outside repos and I must have all the deps on the VMWare studio server.14:37
screen-xgeekbri: patdk-wk my server boots ok, with the nfs mount as an IP, rather than DNS name.14:38
screen-xSeems like a bug14:38
geekbriscreen-x: i honestly would avoid DNS names for NFS mounts always anyway... it just adds another possible layer for failure :)14:39
uvirtbotNew bug: #668295 in bind9 (main) "I just insert the CD of Ubuntu and waited for him to load the system and the error appeared." [Medium,Incomplete] https://launchpad.net/bugs/66829514:39
patdk-wkthat sounds like it could be a *fun* bug :)14:40
screen-xheh, in bind9??14:40
raubvogelupstart question: if I have "start on" in the config file for a given service with nothing after it, would I then have stopped this service  from  starting up?14:40
geekbriscreen-x: see just another reason to avoid DNS for nfs right ;)14:40
screen-xgeekbri: hmm ;-)14:41
uvirtbotNew bug: #633309 in bind9 (main) "host crashed with SIGSEGV" [Medium,Incomplete] https://launchpad.net/bugs/63330914:41
uvirtbotNew bug: #566697 in bind9 (main) "host crashed with SIGSEGV in lh_doall_arg()" [Medium,Incomplete] https://launchpad.net/bugs/56669714:42
nimrod10piken, the closest that I can think of your use case is to use          apt-cache rdepends  ~Dpackage_name14:48
kim0hallyn: I'm only only testing the win7 guest. AFAIK, that bug is related to qxl driver. It seemed to have started with kvm-spice separate binary14:54
nimrod10piken, also see if you can find something useful here : http://www.debian.org/doc/manuals/reference/ch02.en.html#_basic_package_management_operations14:56
madmax_xso i have a race condition here with my network ...after a power failure.... my ubuntu box boots before the dhcp server does...is there a way to have nm-applet or /etc/networking auto retry dhcp after a timed delay?15:00
patdk-wkdhcp attempts to for several min15:01
patdk-wkmaybe adjust the dhcp timeout in dhcp?15:01
madmax_xim using nm-applet to manage the network connection from with X....it tries fails and sits there15:01
madmax_xif i wait for the server to come up its as simeple as telling it to use "auto eth0"15:02
patdk-wkmaybe try #ubuntu?15:02
patdk-wknetwork-manager isn't part of server15:02
madmax_xsorry to bug you...thnx15:03
nimrod10madmax_x, just restart the  network-manager service15:03
madmax_xthe problem is w/o an ip i cant get to the box...i have to go to the physical machine and reset the network15:04
nimrod10a I didn't realize that !15:05
madmax_xover the weekend the power company decided we didnt need power for 8 hours here at the university ... look at all the cool new 'features'...my system has a built in exercise function15:06
madmax_xi tried the retry setting in dhclienet.conf but it doesnt seem to apply to my nm-appplet setup15:08
hallynkim0: well, i don't think i can spend time on that bug.  There are much higher prio ones to be dealth with.  like grub2.vs.-vga-std.  We should keep talking to dev-zero about packaging into universe.  We can deal with the win7 client bug later15:19
dschuettCould anyone point me towards a tutorial that explains how to configure ssl for multiple virutalhosts on the same ip in ubuntu server 10.04 LTS?15:24
joschidschuett: http://wiki.apache.org/httpd/NameBasedSSLVHostsWithSNI15:28
mandad06if someone can help me to instal the terminl of ubuntu in my win715:28
dschuettthanks joschi! - just what i was looking for!15:29
Kimbif someone can help me to instal the terminl of ubuntu in my win715:33
raubvogelIf I am using update-rc.d to disable, not remove, a script why it wants to leave a startup script in rc1.d?15:34
=== jj-afk is now known as jjohansen
BCS-SatoriI have several volumes on /dev/md0 carved which are offered via iSCSI to VMware, so the volumes contain VMFS Partition.  Is it possible to grow the volume in parted (or something else) by adding raw disk space on the end of the volume so that ESX can increase the partition table?16:09
uvirtbotNew bug: #728385 in php5 (main) "apache2 crashed with SIGSEGV in xdebug_stack_element_dtor()" [Low,Incomplete] https://launchpad.net/bugs/72838516:12
RoAkSoAxkirkland: howdy!! quick question :). For orchestra, is it planned to do automatic key exhcnage between puppetmaster and client?16:22
SpamapSzul: "Opinion"?! come on.. thats at least Confirmed/Wishlist . ;)16:23
kirklandRoAkSoAx: yeah, that's the idea;  i don't think we have an agreed upon solution for that yet16:23
kirklandRoAkSoAx: i think that might be an oneiric target16:23
zulSpamapS: what are you talking about16:23
RoAkSoAxkirkland: I see, cause I'm just setting up puppet for the first time here and it indeed is tedious the whole signing process16:24
joschiRoAkSoAx: look into autosigning16:25
kirklandRoAkSoAx: right;  the end goal is for Orchestra-deployed Ubuntu systems would be automatically linked to the puppetmaster, and vice versa16:25
joschikirkland: what's orchestra?16:25
SpamapSzul: bug 70569116:25
uvirtbotLaunchpad bug 705691 in cobbler "cobbler-web should include a working configuration and a README file detailing the steps necessary" [Wishlist,Opinion] https://launchpad.net/bugs/70569116:25
RoAkSoAxjoschi: Ubuntu's installation and infrastructure service16:25
joschiRoAkSoAx: any URL on that thing available for the public? it's kind of hard to google for it ;)16:26
zulSpamapS: heh yeah16:26
RoAkSoAxkirkland: right, so will the orchestra package pull puppet and set custom config files or will puppet itself will be modified for such purpose?16:26
RoAkSoAxjoschi: note that it is just the integration of different technologies to achieve our purpose :) https://launchpad.net/ubuntu/+spec/cloud-server-n-install-service16:27
joschiRoAkSoAx: just interested in it. it sounds a lot like foreman16:28
hggdhjamespage, got a problem with hudson-slave16:29
kmanHello.  Anyone have a fool (that's me) proof way of resetting a user password?  I can't remember the login password to log on.  I've tried the esc at grub loader > recovery but it is always asking me to login in and I have forgotten the password.16:32
patdk-wkmount the fs on a different computer16:32
patdk-wkor use like a livecd16:33
kmanif I use ubuntu live do I then edit some file to allow the reset?16:33
joschikman: https://help.ubuntu.com/community/LostPassword16:33
RoAkSoAxkirkland: so i was thinking and something like what was done with euca can be done in this case using puppet's maste-side client certificate generation16:35
RoyKhm... I have an issue here with Lucid and nfs/nis - a user can't chgrp a directory to another group which he also is a member of. Locally on the NFS server, this works, and testing with Maverick as the NFS server works16:37
kmanThanks will try the second way.  The first did not work for me.16:37
jamespagehggdh - whats the issue?16:44
kmanI have tried all three suggested solutions in the link and none of them are working for me.  Can't find /bin/bash.  I think I will have to go with the live cd method.16:46
hggdhjamespage, http://paste.ubuntu.com/580189/16:47
hggdhjamespage, this was after a server restart16:48
hggdhup-to-date natty16:48
jamespagehggdh - looks like something might not be right on the hudson instance itself16:49
hggdhjamespage, hum. Reboot as an option?16:50
jamespagehggdh: might be a good first step16:50
hggdhjamespage, will do it now16:50
jamespage hggdh: around but on a off for the next hour or so - ping me if this does not work16:51
hggdhjamespage, ack16:54
jamespagehggdh: full filesystem - now that I can see the log....17:02
hggdhoh bloody hell17:07
jamespagehggdh: suggest we shutdown hudson and clear up some of the older d-i logs17:08
kirklandSpamapS: re: https://bugs.launchpad.net/ubuntu/+source/upstart/+bug/70747917:08
uvirtbotLaunchpad bug 707479 in upstart "service <service> restart does not use an updated job configuration" [Medium,Triaged]17:08
kirklandSpamapS: i was just about to patchpilot that one17:08
kirklandSpamapS: i see cjwatson's comment from a week ago, though17:09
kirklandSpamapS: and i also saw that jhunt has a big merge for natty too17:09
SpamapSkirkland: yeah, jhunt is almost done.. I'd hold off a bit longer.17:09
kirklandokay, thanks17:09
SpamapSkirkland: tho I definitely want that in natty.17:09
SpamapSI understand why restart works how it works.. but I fail to see where its actually ever useful that way. :-P17:10
SpamapSkeybuk has had a lot more time to think this through and has a few use cases I'm sure.17:10
kirklandSpamapS: is he just going to commit your change supstream?17:11
kirklandSpamapS: when ready?17:11
SpamapSkirkland: no upstart-job doesn't exist upstream IIRC17:12
SpamapSupstream all thats going to happen is the restart command will be documented more completely17:13
kirklandSpamapS: got it, thanks17:15
jamespagehggdh: hmmm - looks like alot of the jobs don't clear down the workspace before they start17:21
jamespageso they just get bigger and bigger over time.17:21
Orfeoushaving some issues with SSL and apache2, i have generated a self-signed certificate but got some problems when restarting apache2, CN does not match servername...17:22
Orfeouswhat configs can i show to help resolv this?17:23
SpamapSkirkland: and thanks for looking into it.17:24
Orfeousi followed the certificate making guide here: https://help.ubuntu.com/10.10/serverguide/C/certificates-and-security.html17:26
hggdhjamespage, yes. I am considering only leaving the last -- say -- 10 runs17:32
jamespagehggdh: thats not really the problem17:33
jamespagehggdh: http://pastebin.com/0dyNu4hm17:34
jamespagehggdh: that is for job 31 - it contains the archive for all previous jobs as well as they where present in the workspace.17:35
hggdhjamespage, oh17:35
jamespagehggdh: unless the job clears its workspace first you get this incremental build up17:35
jamespagelemme check which ones are doing this - I did fix a load automagically17:35
jamespagehggdh: looks like the TBench and DBench (for natty) tests did not have the updated script; all others looked OK to me.17:47
hggdhjamespage, indeed, the D|Tbench were old, I wrote them very early in the process17:52
hggdhjamespage, seems they are all set now. I will monitor and see if we missed any18:15
uvirtbotNew bug: #303059 in samba (main) "Samba segfaults four-five times a day on Ubuntu Server (both hardy and intrepid)" [Undecided,Invalid] https://launchpad.net/bugs/30305918:22
ubottuFor an introduction to IPv6 and information on tunneling IPv6 through IPv4 connections, see https://wiki.ubuntu.com/IPv6 | To disable IPv6 see https://help.ubuntu.com/community/WebBrowsingSlowIPv6IPv418:25
axisysi can ping my ipv4 ip using ping .. if i ping6 ipv6 ip i get "invalid argument" .. any idea why?18:30
lifelesskirkland: hi18:34
lifelesskirkland: so you're saying on https://bugs.launchpad.net/bugs/487337 that it wasn't the freeform text you cared about, it was the list of explicitly linked bugs?18:34
uvirtbotLaunchpad bug 487337 in launchpad "bug status for bugs linked in a blueprint is not shown" [Critical,In progress]18:34
axisyslsmod | grep ipv6 does not show anything18:37
kmanOK I just tried doing the password change by esc at grub menu then recovery but I get the prompt for a root password.  then I tried the edit mode and still get prompted.  I just tried the LiveCD method where it suggests one changes the shadow file and the password is already set to the passwordless simple hash.  What to do now?18:49
kmanI have forgotten the password.  When I went looking in the home directory I expected to see more than just a Ubuntu account. but that is all I can see.18:50
kmanAnyone have any further suggestions before I reinstall?18:50
uvirtbotNew bug: #735017 in php5 (main) "package php5-fpm 5.3.5-1ubuntu2 failed to install/upgrade: подпроцесс новый сценарий pre-removal return code error 1" [Undecided,New] https://launchpad.net/bugs/73501718:52
crohakonWhat is a good, free, alternative to cPanel to manage shared hosting?18:56
MuuuhWe'd like to setup a Ubuntu server box between our LAN and existing Sonicwall box to do some packet filtering/logging. I'm gathering information at this stage so I can convince the IT manager to let me use Ubuntu. Has anyone done such a setup ?19:29
b0gatyrHi can someone shed some light, I have a monitor plugged into an ubuntu box who's max res is 1280x1024 i plugged it into a KVM switch and now the resolution is 1024x768 and it gives me a max res of 1360x768 which is incorrect.. can anyone help19:31
b0gatyrhow can i make it detect the right res?19:32
pmatulisMuuuh: i thought sonicwall *was* a packet filtering thing19:32
Muuuhpmatulis: indeed, but it's reporting facilities are not giving us what we want in terms of detailed traffic to the internet19:33
RoyKMuuuh: try snort19:34
Muuuhpmatulis: thus the idea to use a linux box in between19:34
speakmanIf anyone new has arrived; I'd like to help trace what's causing this; http://pastebin.ubuntu.com/580065/ on my Xeon W3680 + 6x4GB ECC RAM (QVD) on a Asus WS Revolution?19:34
RoyKMuuuh: just install snort on another machine connected to a sniffer port or hub on the internet - don't set an IP address on the NIC connected there, and you're in business19:34
MuuuhRoyK: ok, googling snort... you mean it will "snort" without being actually in the "pipeline" ?19:35
RoyKMuuuh: with a hub or a port configured as a sniffer/monitor port on the switch (most managed switches supports that), all the data will be sent to that port19:37
RoyKspeakman: either a mobo or cpu failure, at least that's my guess19:37
pmatulisMuuuh: what kind of information are you after?19:38
RoyKmcelog has saved me some hours :P19:38
RoyKpmatulis: I would guess some sort of IDS19:38
Muuuhpmatulis: trying to nail down internet abusers19:39
pmatulisMuuuh: and abuse means what?19:39
pmatulisMuuuh: content?19:39
RoyKwhat's that? am I abusing the internet if I use the bandwidth I pay for? :)19:39
Muuuhpmatulis: content is one aspect, yes19:40
MuuuhRoyK: well, you are if you're doing it on the job without your boss' authorization ;-)19:40
bigbangwhat you guys think of zimbra ? worth looking into it ?..19:40
RoyKMuuuh: spying on people's use of the web at work is bad, it's just BAD19:41
Muuuhsince when ?19:41
RoyKbigbang: been using it for a couple of years - works splendidly :)19:41
pmatulisMuuuh: well we can't recommend a solution if we don't know all the criteria.  but snort can filter on pretty much anything19:41
bigbangawesome thanks RoyK .. :)19:42
Muuuhpmatulis: well, we're basically trying to gather evidence as to people wasting their time on internet instead of working19:42
RoyKMuuuh: if it's internal web traffic you're trying to monitor, just setup a proxy, squid works well19:42
RoyKbigbang: most hard-core ubuntu freaks don't like zimbra, since it's not really very open, but I do, even so19:42
MuuuhManagement doesn't want to lock it down tight. It's just that they suspect some users of abuse, and wants to gather evidence before accusing19:44
pmatulisMuuuh: but i think there must be an easier tool like a kind of passive content filter19:44
pmatulis(monitoring instead of blocking)19:45
Muuuhas I said, we don.'t want to block, just gather useage data19:45
RoyKMuuuh: using squid as a transparent proxy will help you log these things quite easily19:45
speakmanRoyK: based on..? All parts are completely new, and has a fresh Ubuntu 10.10 install. The failures is random with about a couple of days between (but as well a couple of times a day some days).19:45
RoyKspeakman: sorry, you lost me - what are you talking about?19:46
speakmanRoyK: Another issue which might be related, is the CPU temperature not showing on some sessions. Sometimes it does, sometimes it doesn't.19:46
speakmanRoyK: me MCE above19:46
RoyKspeakman: according to the MCE dump, it seems to be a CPU issue, but I cannot say for sure - it may be the mobo as well. Are the logs showing different addresses for the issue?19:48
pmatulisMuuuh: see #24 here: http://contentfilter.futuragts.com/wiki/doku.php?id=faq#general_faq19:49
RoyKI have a similar issue on a server - mcelog logs every now and then - it used to hang the server during memory check, but doesn't after we got a new mobo, but still, errors keep ticking in. The supplier told me to swap the CPUs to see if the address was the same on the next error19:50
speakmanRoyK: I have only got one log so far :(19:50
speakmanRoyK: I have been running memtest86 for two hours (at least one complete pass) and no complains19:51
RoyKspeakman: didn't you say it was crashing repeatedly?19:51
RoyKif so, use syslog, and setup remote logging to another box19:51
speakmanRoyK: My problem is that the CPU and mobo is from two different suppliers. But there's probably no problems changing both of them.19:51
RoyKmcelog/syslog can't write to the disks if the system is panicing19:51
speakmanRoyK: yes, but I didn't know what caused it earlier. Since I'm on X, I had to set up a netconsole to grab that message.19:52
speakmanRoyK: Now I've setup mcelog as well19:52
speakmannetconsole is the s*it :)19:52
RoyKjust configure syslog to log to another box as well19:52
RoyK*.* @some.ip.in.house19:53
RoyKand remember to allow remote logs on that box19:53
speakmanwhy really?19:54
RoyKbecause then you can catch logs that can't be written to disk if the system is panicing and i/o is blocked19:54
speakmanI can read /var/log/syslog afterwords, and still I havn't found anything there yet.19:54
* RoyK always uses a separate syslog server19:54
speakmanRoyK: isn't kernel messages enough you think..?19:54
speakmanI can of course...19:55
speakmanhttp://www.mjmwired.net/kernel/Documentation/networking/netconsole.txt btw19:55
RoyKspeakman: probably, but it won't hurt to send it all to another box with sufficient storage19:55
RoyKit's not like it's a lot - a few megs or so19:55
speakmanI'll just send it to the same machine receiving netconsole logs19:56
RoyKspeakman: a good old serial cable will do well too :)19:56
speakmanRoyK: if there were only a good old serial connector on the mobo ;)19:56
speakmanNot even an internal one available.. :/19:56
speakmanis Ubuntu running rsyslog instead of plain old syslogd nowadays?19:58
RoyKrsyslog iirc19:59
ScottKYes.  It's been the default for a few releases.19:59
RoyKbut opening for remote syslog is trivial19:59
speakmanbut setting up syslog for sending?20:01
RoyKspeakman: just add *.* @somehost in /etc/syslog.conf20:02
speakmanthanks :)20:02
RoyKthen restart syslog and see if something is written to the remote host20:03
speakmanyes, but can I set port?20:03
RoyKdunno - usually it uses the standard (514/udp)20:04
speakmanI'm already listening on another port, I'll just RTFM ;)20:04
RoyKif you're worried about other hosts flooding, setup a ufw filter20:04
RoyKspeakman: adding :port might help20:05
speakmanUp and running...20:14
speakmannow netconsole again20:14
KB1JWQWhat's the polite way to tell apt-get to go with either "no config" or "local only" for postfix programmatically20:18
KB1JWQI'm trying to script a template creation and that's breaking my automation unfortunately.20:18
=== nxvl_ is now known as nxvl
david5345If I run the "kvm-ok" command and the output is "INFO: your CPU does not support KVM extension". Does this mean I can't use KVM or does it mean I can use KVM but it'll work slowly ?20:38
EvilPhoenixi dont think you can use KVM if your CPU doesnt support it20:38
EvilPhoenixhence the error message20:38
hallyndavid5345: 'kvm' will just end up running qemu without kvm extensions.  so like you said, slowly20:39
david5345Thanks :D20:39
david5345I'll update the Wiki20:39
speakman...and now I can't even get netconsole working :/20:52
lifelesskirkland: thanks21:11
RoAkSoAxkirkland: ping21:26
RoAkSoAxDaviey: ping?21:36
SpamapSRoAkSoAx: yo21:50
RoAkSoAxSpamapS: yo!! what's up21:53
SpamapSRoAkSoAx: writing upstart docs. You?21:54
RoAkSoAxSpamapS: testing powernap's new changes for eucalyptus xD21:55
RoAkSoAxhow was the weekend?21:56
SpamapSRoAkSoAx: busy as usual. work is so much more relaxed. ;)21:59
RoAkSoAxSpamapS: indeed!! I finally got the chance to go to the beach yesterday after more than 6 months :)22:00
RoAkSoAxSpamapS: too bad I'm not in spring break though as everybody else22:00
DavieyRoAkSoAx, o/22:03
RoAkSoAxDaviey: htough u were sleeping man22:03
RoAkSoAxDaviey: anyways, I commited the changes to powernap for eucalyptus. I simplified it to only Monitor related changes22:06
SpamapSDaviey never sleeps .. he is always there.. watching.. waiting.. for the next eucalyptus bug report.22:06
RoAkSoAxDaviey: so anyways, I guess we can make a config for powernap to be shipped with eucalyptus22:08
DavieyRoAkSoAx, i never sleep!22:08
RoAkSoAxSpamapS: btw... upstart's restart is not intended to load the config again then?22:08
DavieyRoAkSoAx, That sounds great!  Have you documented the syntax for the file?22:09
RoAkSoAxDaviey: well it is the same syntax as the regular config file i.e.22:09
RoAkSoAx[ProcessMonitor]\n init = "^/sbin/init"22:10
DavieyRoAkSoAx, Ahh!22:10
SpamapSRoAkSoAx: it should stop/start the daemon without re-loading the job configuration.22:10
SpamapSRoAkSoAx: in essence, it forces a respawn.22:10
DavieyRoAkSoAx, Please can i have a config for mythtv, to monitor if i'm watching something or if it is idle :)22:10
Daviey(actually, could probably make a good guess based on load avg)22:11
RoAkSoAxDaviey: it will just override the default config for the Monitors (if any) or add new monitors. I'll first think which one would be best22:11
RoAkSoAxDaviey: yeah either that or ProcessMonitor22:11
RoAkSoAxi think kirkland had similar for ProcessMonitor22:11
DavieyIIRC, kirkland's just monitored if mythfrontend was running22:11
DavieyI never exit out, so that is not a win for me.22:12
kirklandRoAkSoAx: hey there22:12
kirklandRoAkSoAx: okay, i have time to talk about powernap/eucalyptus now, if you like :-)22:12
RoAkSoAxSpamapS: ahh I see now, so still the pre-start issue for restartin remains22:12
RoAkSoAxkirkland: cool :). I just commited a branch22:12
RoAkSoAxkirkland: that will allow to load configs from /etc/powernap/config.d/22:13
DavieyRoAkSoAx, Did you carry the priority idea?22:14
DavieySo i can nullify or extent settings in a higher value one, for example?22:15
SpamapSRoAkSoAx: restart also doesn't work if there is a pre-stop22:16
RoAkSoAxthat will override the config from /etc/powernap/config, but *only* for the monitors22:17
RoAkSoAxDaviey: yeah, python handles that22:18
DavieyRoAkSoAx, Splendid!22:18
RoAkSoAxSpamapS: only pre-stop or pre-start too22:18
SpamapSRoAkSoAx: I'm not aware of any pre-start issues22:19
RoAkSoAxoh ok so only pre-stop22:20
SpamapSepic home-row fail22:21
=== ogra_ is now known as ogra
BCS-SatoriHow would I go about making an apache2 website be only accessible by its servername or serveralias?  I don't want the website to appear if someone types in an IP address.22:23
RoAkSoAxDaviey: yeah so if you place 80etc.whatever and you have [LoadMonitor] load1 = 1 and then you have a 90test.whatever with [LoadMonitor] load1 = 2 , then the 90 file has greater priority and overrides the 80 one22:24
RoAkSoAx(i was on the phone, sry :) )22:24
SpamapSBCS-Satori: you just need to configure set NameVirtualHost ip:port22:26
SpamapSBCS-Satori: and then wrap the site in <VirtualHost ip:port> with a ServerName / ServerAlias22:26
BCS-SatoriSpampaS: internal or external ip or both?22:27
SpamapSBCS-Satori: If you can, *:port works too22:27
SpamapSBCS-Satori: you may need to creat a "default" virtual host that is empty for people who reach the site via the IP directly.22:27
RoAkSoAxkirkland: so yeah the idea is to ship a config with eucalyptus for powernap in config.d/ that adds monitors or overrides the ones already defined, so that we it works out of the box and we don't have to tweak anything because of the new changes in PowerNap. Though the override is only for the monitors, I'm not touching the general config22:28
SpamapSIIRC apache uss the first virtualhost definition as the default22:28
BCS-SatoriSpamapS: I have all of that but the default...which may be why it doesnt work22:28
kirklandDaviey: mine use to look for a set of processes, like mplayer, vlc, etc.  not mythfrontend22:29
Davieykirkland, I was sure you had a myth entry there :/22:29
kirklandRoAkSoAx: yeah, i liked that idea when Daviey or whoever suggested it22:30
kirklandDaviey: i don't think so22:30
kirklandDaviey: but that's long since been replaced with RoAkSoAx's magical new powernap :-)22:30
Davieykirkland, Your blog says otherwise :P22:30
kirklandDaviey: heh22:31
kirklandDaviey: what's the date on that entry?  :-)22:31
RoAkSoAxkirkland: indeed Daviey.  So I guess it could pass as a bug fix for bug #711587 or shall I request a FFe? Changes are these: http://bazaar.launchpad.net/~powernap/powernap/trunk/revision/22022:31
uvirtbotLaunchpad bug 711587 in powernap "powernap and Eucalyptus seem unable to reach an understanding" [Medium,Fix committed] https://launchpad.net/bugs/71158722:31
RoAkSoAxDaviey: Iblogposted a week ago or so about the new changes :) :P22:32
kirklandRoAkSoAx: i don't think this needs an ffe, imho22:32
kirklandRoAkSoAx: this is the correct fix to a known bug, a regression that affects eucalyptus22:32
* kirkland reviews the code22:32
kirklandDaviey: what's your opinion?  needs an ffe, or not?22:33
kirklandRoAkSoAx: we should convert powernap's debian/rules to a dh7 style rules file, after this cycle :-)22:33
Davieykirkland, Hmm, it fixes a bug :)22:34
RoAkSoAxkirkland: yeah wanted to do that too22:34
kirklandRoAkSoAx: diff looks nice and clean to me22:34
kirklandRoAkSoAx: low impact22:34
kirklandRoAkSoAx: let's get this merged and uploaded22:34
kirklandRoAkSoAx: i'm happy to sponsor, if you like22:34
RoAkSoAxkirkland: alrighty then. Adn yes please, go ahead an upload it :)22:34
Davieykirkland, It's not a new feature, it's handling a change in one of it's reverse depends :)22:34
kirklandRoAkSoAx: speaking of, you should apply for upload rights for powernap :-)22:34
RoAkSoAxkirkland: yeah I want this cycle to finish and I'll apply either for PPU for cluster stack+powernap or the Ubuntu Server package set22:35
kirklandRoAkSoAx: cool :-)22:35
DavieyRoAkSoAx, Server set sounds good... that needs more members. :)22:35
RoAkSoAxDaviey: indeed!! I'm just too lazy to pass the results from the script to the wikipage :P22:37
RoAkSoAxkirkland: btw.. how was London/england?? I always wanted to go there since I was a kid.... hope I get the chagne to go there soon :)22:38
kirklandRoAkSoAx: uploaded22:38
kirklandRoAkSoAx: i'm glad to be back in Austin, that's for sure :-)22:38
kirklandRoAkSoAx: i'm working outside, in the sun, 75 degrees right now22:38
kirklandRoAkSoAx: it was 33F when i got off of the plane in London :-)22:39
DavieyRoAkSoAx, :P22:39
RoAkSoAxkirkland: haha yeah I bet!! it is nice here too... I went to the beach yest after probably 8 months22:39
Davieykirkland, I think we need an Austin sprint. :)22:40
RoAkSoAxDaviey: we need a sprint in Miami >P22:40
kirklandDaviey: yeah, we do :-)22:40
hallynDaviey: hey - how is the open-vm-tools merge review going?22:40
RoAkSoAxDaviey: from a beach facing hotel22:40
hallynmiami sprint - +122:40
kirklandDaviey: robbiew should get on that :-)22:40
* RoAkSoAx wishes he was working from the beach...22:43
Davieyhallyn, How did we last leave it?22:43
Davieyhallyn, You raised a FFe, right?22:43
Davieyhallyn, Hmm.. i thought you had.. seems not?22:45
hallynDaviey: no, i was waiting for your review22:47
Davieyhallyn, Oh, sorry.22:47
hallyni thought it was decided ffee wasn't needed?22:47
hallynsince this is not a feature?22:47
Davieyhallyn, I was a MASSIVE diff iirc.. not really easy to review :/22:47
hallyni could be wrong22:47
Daviey294925 lines (+100634/-169192) 750 files modified22:48
hallynwell, maybe we should find someone in community who actually uses/cares about this to review and consider maintaining the package?22:49
hallyni would think you would want tofocus on the changes under debian/22:49
hallynthe rest is just the reuslt of upstream churn22:49
DavieyThat is what i am diffing atm :)22:50
hallynDaviey: i wasn't asking you to do it right now :)  Just wanted to make sure you weren't waiting on some action from me22:51
Davieyhallyn, I'm doing that now... but if you can go through the FFe dance, that would be good.. The size of the diff alone, makes me think we should go through that.22:53
Davieyhallyn, $(KVERS) really does == $(shell uname -r)?  You checked that?23:05
DavieyOh forgive me, that is a dropped patch23:05
Davieyhallyn, Actually, debian/patches/01-kvers.patch - seems to be converted to a bzr diff style patch... did it need regenerating - looks like it would have cleanly applied?23:08
=== jjohansen is now known as jj-afk
DavieyRoAkSoAx, bug #732759, do you have a build log, copy and paste of apt-get install handling and comment on testing you have done that it works?23:34
uvirtbotLaunchpad bug 732759 in ubuntu "[FFe] [needs-packaging] python-ethtool" [Wishlist,New] https://launchpad.net/bugs/73275923:34
jiboumanshey folks, hudson has an apt-repo here: http://hudson-ci.org/debian/binary/ but it only has a Packages file, not Packages.gz and that makes apt-get update claim '404' -- what's the right way to use this repository?23:44
RoAkSoAxDaviey: yeah will copy it in a bit, gotta run for a while23:44
Davieyjiboumans, Firstly, have you tried our PPA?  jamespage has been working hard making that usable23:48
jiboumansDaviey: no i haven't yet - i was told by folks that hudson-ci was the right way to go23:49
jiboumans'folks' being.. well... you know, 'them' :)23:49
SpamapSjiboumans: isn't it jenkins now? ;)23:49
jiboumansDaviey: i'm happy to give that a look as well, but right now I do need to get this repo working first. any thoughts?23:49
Davieyjiboumans, well, i know we would appreciate additional testing... :)23:50
Davieyjiboumans, I have an apt repo without a gz'd Packages, and that seems to work.. :/23:50
jiboumansDaviey, here's what I'm using:23:51
jiboumanscurl http://hudson-ci.org/debian/hudson-ci.org.key | apt-key add -; echo "deb http://hudson-ci.org/debian binary/" > /etc/apt/sources.list.d/hudson.list; apt-get update23:51
Davieyjiboumans, You might aswell just wget the deb :/23:55
jiboumansi'd like puppet to manage the updates and dependencies23:55
jiboumansDaviey: ^23:55
SpamapSYou mean you'd like hudson-ci.org to manage the upgrades ;)23:56
SpamapSman.. I was just curious.. there's no good place to find "the apt repository format"23:58
Davieyjiboumans, I'm not sure i trust them.... based on http://issues.hudson-ci.org/browse/HUDSON-2472?page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel23:58

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!