=== Wandrewvious is now known as WALoeIII [00:16] These are good news: http://en.community.dell.com/dell-blogs/enterprise/b/tech-center/archive/2010/07/27/dell-openmanage-6-3-for-ubuntu.aspx [00:37] anyone use rtorrent here? i'm having issues with the watch directory feature & pausing/temporarily stopping torrents... [00:45] mathiaz, actually most of that is in uec-run-instances, which SpamapS worked on recently. it definitely needs some work, but it is there. [00:45] its also, i tihnk, sort of part of mr.awsome [00:52] Trying to learn how to create a private package repository: which directories does reprepro needs and what do they do (my google-fu has only found how-tos, but not whys) [01:15] *cough* [01:17] New bug: #611101 in mysql-dfsg-5.1 (main) "upstart config does not sleep between pings" [Undecided,New] https://launchpad.net/bugs/611101 [01:21] New bug: #611102 in mysql-dfsg-5.1 (main) "mysqld does not start due to typo in upstart config" [Undecided,New] https://launchpad.net/bugs/611102 [02:02] exit [02:10] im broken? [02:45] how can I force a 6.4LTS server to do a dist upgrade? I've tried "apt-get dist-update" and it claims there is no updates [02:46] 6.04 isn't supported is it? [02:46] do-release-upgrade [02:47] arrrghhh, I had to install 6.04 in order to get these IBM x335 boxes to play nice with Ubuntu [02:48] are you sure it's a good idea to upgrade them then? lol [02:48] arrrghhh, they're fresh installs, so if it tanks, I'll just reinstall [02:48] there's emergency broacast boxes that run redhat distro's from 1996 becuase it works. [02:48] arrrghhh, what's the syntax I need? [02:48] if it works, it works. [02:48] "sudo do-release-upgrade" what do you mean? [02:49] arrrghhh, that's what I needed to know. I was trying "sudo apt-get do-release-upgrade" and it wasn't working. LOL [02:49] oh... crap. sorry haha. [02:49] arrrghhh, your syntax didn't work. it just line feeds [02:49] hrm... [02:49] maybe that was an advent with 8.04 [02:50] other ideas? [02:50] aptitude? [02:50] aptitude safe-upgrade i think [02:51] nana [02:51] *nada [02:51] http://www.ubuntu.com/desktop/get-ubuntu/upgrade [02:52] go to the "network upgrade for ubuntu servers" section [02:52] arrrghhh, that's desktop [02:52] c'mon now! [02:52] i know it says 8.04 -> 10.04 [02:52] yep [02:52] but do it [02:52] apt-get install update-manager-core [02:53] edit /ect/update-manager/release-upgrades and set prompt=lts [02:53] or to =normal, whatever [02:53] and sudo do-release-upgrade --devel-release [02:53] sudo apt-get install update-manager-core [02:53] sudo do-release-upgrade [02:54] i was going to keep you in LTS trains [02:54] but whatever mang [02:54] it's all there in that doc! [02:54] thanks [02:57] ...and off to hardy we go [02:57] have fun [03:42] If you don't have plymouth installed, does "splash" in the kernel parameter do antyhing? [04:52] Hi [04:53] I am facing some problem regarding discovering nodes in eucalyptus http://paste.ubuntu.com/470196/ [04:53] When it is discovering a node it is identifying with its inet6 address and not the inet4 address [04:54] can someone please help resolve this issue [05:17] i am using the ec2 image, i added a new user via adduser, except I can't connect as that user using its password. Do I have to pass in a pem file when I ssh ? [05:51] Adman65: its likely that password auth is just turned off for the EC2 images. [05:59] Daviey, awake ? [06:00] Adman65, password auth in ssh is disabled in ec2 images. [06:00] to "fix": [06:01] sed -i "s/#PasswordAuthentication/PasswordAuthentication/" /etc/ssh/sshd_config [06:01] sudo sed -i "s/#PasswordAuthentication/PasswordAuthentication/" /etc/ssh/sshd_config [06:01] sudo restart ssh [06:24] I am facing some problem regarding discovering nodes in eucalyptus (UEC) http://paste.ubuntu.com/470196/ [06:24] When it is discovering a node it is identifying with its inet6 address and not the inet4 address [06:25] Do you actually use ipv6 at all? [06:25] If not, suggest turning it off at the boot prompt [06:27] twb: nope [06:27] twb: by boot prompt do you mean bios? [06:27] No, I mean in your bootloader: grub or pxelinux or whatever [06:28] ok, I am using grub, can you plz give me a pointer to how I might go about achieving it twb? [06:29] grub2 or grub legacy [06:29] twb: the one provided with Lucid Lynx [06:29] In /etc/default/grub add ipv6.disable=1 to GRUB_CMDLINE_LINUX_DEFAULT and run "update-grub" [06:31] twb: currently the value is GRUB_CMDLINE_LINUX_DEFAULT="quiet" changing it to GRUB_CMDLINE_LINUX_DEFAULT="quiet ipv6.disable=1" and running update-grub [07:17] RoAkSoAx: pong [07:41] ttx: I attached a debdiff for rhcs [07:41] bug #600984 [07:41] Launchpad bug 600984 in redhat-cluster "redhat-cluster-suite fails to build from source in maverick" [High,Confirmed] https://launchpad.net/bugs/600984 [07:42] RoAkSoAx: ok [07:59] smoser, o/ [07:59] hey [07:59] que pasa senor? [07:59] smoser, oh joy, about ftbfs [07:59] i've a merge for you to look at for eucalyptus. [07:59] yeah [08:00] once it bfs, i have some changes that i need. [08:00] they'rre not well tested, thoguh, as i can't install [08:00] :) [08:00] but fairly well tested given that slight difficulty [08:00] HAH [08:01] i need some serious help [08:01] for some reason my server seems to be stuck in an infinite loop [08:01] it boots up gets to the login then shuts down and restarts [08:02] if you care to review, Daviey https://code.launchpad.net/~smoser/ubuntu/maverick/eucalyptus/maverick.bug611144/+merge/31249 [08:05] smoser, visually, it looks sane [08:05] well, it builds. [08:07] smoser, in lucid or maverick? [08:08] in maverick [08:08] i hand installed the libjibx into an otherwise clean schroot [08:08] the dpkg-buildpackage [08:09] ahh [08:09] sweet :) === TeLLuS_ is now known as TeLLuS [08:26] New bug: #611177 in samba (main) "samba as pdc, w2k3 as memberserver, winxp domainclients" [Undecided,New] https://launchpad.net/bugs/611177 === remix_auei is now known as remix_tj [08:46] vhey guys i need some big time help i have a server down and its kind of in an infinite boot up and restart loop [08:46] im trying to boot onto the live cd and it wont read it for some reason and with out any networking i can login but after logging in it restarts itself [08:54] eagles0513875, not much you can do until you get it booted to a live environment, or load the disk in another computer [08:54] sigh :( [08:55] qman__: woudl a network boot work? [08:55] have you tried the "recovery mode" option? [08:55] only if you have a server hosting PXE images [08:55] qman__: what do i press to get to grub and kernel selection [08:55] for grub1, escape, for grub2, hold left shift [08:55] hold on [08:56] recovery mode boots it into single user, and then gives you a few options besides just dropping to a root shell [08:56] i know [08:56] i think this loop has something to do with watchdog [08:57] im holding shift and im getting no response [08:57] its like its ignoring the keyboard [08:57] USB? [08:58] ya [08:58] also the cd drive si a bit wonky [08:59] its a sata cd rom drive [08:59] if so, try playing with the legacy USB keyboard options in the BIOS [08:59] or try PS/2 [08:59] make sure it's in "native IDE" mode [09:00] some things can handle AHCI, others can't [09:00] qman__: what doesnt make sense is this was working just fine yesterday [09:01] it's a big mess, so if you don't need to hot plug your drive, just use native IDE [09:01] blarg this is a mess [09:01] now, hard drives can really benefit from AHCI, but disc drives, not a big deal [09:02] man this is getting frustrating [09:03] i am wondering if it was the time as the time in the bios was 2 hrs behind [09:03] eagles0513875, by default, linux sets the hardware clock to UTC [09:04] humm [09:04] ok [09:04] that didnt do it [09:04] and i cant change to legacy usb [09:04] Can you not trying setting your first boot device to CDROM and then putting in a recovery disk [09:04] just to see if you even get usb support? [09:06] SuperLag: already have that [09:06] setup like that and it still doesnt boot off the live cd [09:06] of ubuntu server [09:06] ahh [09:06] sorry couldnt scroll back that far [09:06] its ok [09:07] i am running out of ideas [09:07] i think a reinstall might be in order [09:07] what happens when you try to boot it up? [09:07] starts up i login and it shutdown and reboots [09:08] and in single user mode the same? [09:08] his keyboard isn't working to interrupt grub and get to single [09:08] you're going to have to fix one or the other, the CD or the keyboard, in any case [09:09] can't exactly reinstall if you can't boot the live environment [09:09] ah have you tried booting from a USB device? [09:09] thing is i dont have a spare usb device on me atm [09:09] the one i have is 16 gb of data on it [09:09] actually brb [09:09] cuz the office was supposed to order some for work [09:10] hold on i might not need it [09:10] ok nm [09:10] i do need it [09:10] sure could it be ahrd ware issue if the kb isnt working and its in a boot loop? [09:10] it seems like all processes are crashing from what im seeing during shutdown [09:10] the server was workign juts fine yesterday [09:11] brb [09:11] that doesn't rule out a hardware issue [09:11] has anything changed? plugged in sumthink [09:15] Does it crash ont he same process every time? [09:15] on the [09:16] it seems like its something with udev [09:16] im gonna use my pendrive and create a bootable usb [09:16] using unetbootin [09:17] just have to get all my data off it first [09:17] Sure one other suggestion, when you do remove all the hardware you dont need, including some of the ram and the CD ROM drive if you can [09:18] SuperPetRalf: ? [09:18] Just incase it is a hardware issue, it maybe on the the external devices [09:19] or an internal device such as a broken pin in the ide cable or faulty RAM [09:20] ieverything is sata [09:20] from cd drive to hdd drives [09:20] i unplugged the network [09:20] which helped some [09:20] before with it plugged in [09:20] it was an instant reboot [09:21] without network it at least stays up for a bit then i login and it goes down [09:22] i think im getting ddosed or something [09:22] is it possibly 2 seperate issues, for example you cant use the kb in grub becuase USB support isnt there and there is a corrupt file cousing your system to reboot? [09:23] SuperPetRalf: all i did was modify something in the watchdog conf file and restarted watchdog [09:23] and after that thats when it was in the infinite loop [09:23] prior to that it was fine [09:24] can you resotre the old file [09:24] i know what i modified [09:24] just have to comment out one thing [09:24] idea [09:25] can you turn it on but not logon [09:25] then login remotly with ssh? [09:25] i could try that [09:25] also take a look at this [09:25] http://pastebin.com/ebmPZndP <---does that look like someone trying to hack or dos the network [09:26] there all outbound though [09:26] not all [09:26] but the majority are and I dont htink that could be anough for DOS conditions [09:27] only one is inbound [09:27] looks like your server's been compromised to me [09:27] unless you're legitimately scanning those IPs yourself [09:27] im not [09:27] im protected by router firewall [09:28] only ports i have open and forwarded to the server are 22 and 80 [09:28] that's more than enough [09:28] so i need to get iptables up and running on the server as well as a proxy [09:28] you need to fix the hole that they got in with, if they got in [09:29] ya its server side i think [09:29] leave it off the network, and check for evidence of a break in [09:29] thing is i removed the server from the connection via router to the internet [09:29] breakin in what sense [09:29] yeah but the traffic is outbound [09:29] suggesting if it is its allready comprimised [09:29] unauthorized SSH login, most likely [09:29] with a subsequent rooting [09:30] qman__: i use an alpha numeric password which isnt easy to crack [09:30] doesn't matter [09:30] using password authentication at all is risky [09:30] brute force, but i still dont think its likly to be that [09:30] thing is i cant even check the logs [09:31] and most of the attacks are on closed ports [09:31] whats the IP address of the server? [09:31] those are outbound connections [09:31] .4 or .7? [09:31] none of those [09:31] then its not the server [09:31] server has a static internal ip [09:31] look at the addresses [09:31] blocking inbound ports has no effect on them [09:31] those i believe are wifi addresses [09:32] ahh but all the traffic goes no where near the server so to speak [09:32] look its all outbound .2.4 and 2.7 [09:32] yeah [09:32] looks like you had a couple zombies on your wifi then [09:32] or someone playing around with nmap [09:32] i'm finally got working suexec with fcgid, but it uses a wrapper that contains http://pastebin.com/KCsLN1dR, now, the wrapper must be owned by the user, so he can change it and put whatever command he wants, how can i solve this? [09:33] try and ssh into your own server [09:33] those are wifi ips from .2 to 100 [09:33] via router dhcp [09:33] if your server is on the same subnet, you definitely need to put up a firewall [09:34] but that's down the road [09:34] if none of those addresses point to your server, then it's irrelevant [09:34] and there are time log inconsistences too [09:34] and this is sounding more like hardware failure by the minute [09:34] ssh seems to be down [09:34] see no 11 then 12 and you see what i mean [09:35] cant ssh in [09:35] anyway ssh is down just checking are you re plugged in [09:35] morning [09:35] can you ping? [09:35] Morning [09:35] wait hold on [09:36] are you using an onboard NIC? [09:37] no [09:37] that was faulty [09:37] its a pci nic [09:37] i was gonna get a 2nd pci nic [09:38] no ping reply is that? [09:38] put one on the dmz and route the traffic to the other one after content filtering etc [09:38] well, that's even more suspect [09:38] server is offline atm [09:38] qman__: ? [09:38] if the onboard was already dead, the motherboard is likely crapping out [09:38] qman__: thing is i had taken this server home and installed server just fine using the onboard nic [09:38] came to work and it didnt work here [09:39] you mean during that time you installed fisrt and powered up now its been moved!? [09:40] let me see if i can boot onto a bootable pendrive [09:40] SuperPetRalf: yes [09:40] i reinstalled with this other nic card i bought [09:40] ahh [09:40] a failing southbridge chip would explain everything [09:40] hardware, or maybe disloged component [09:40] udev crashing, NIC failure [09:40] O_O [09:40] im with qman on that [09:40] qman__: this has been online for at least 2 three weeks [09:41] irrelevant [09:41] all hardware fails eventually [09:41] when is a matter of chance [09:41] most of the time it jsut goes [09:41] no warning [09:41] especailly if youve moved it [09:42] thing is this is only a 1 yr old machine [09:42] also irrelevant [09:42] is it on a UPS? [09:42] yes [09:42] is the UPS rated hight enough? [09:42] could be the power supply as well, but it's a bit too consistent [09:43] when they go, they either usually just pop [09:43] or fail under a heavy load [09:43] Daviey: around ? [09:43] which I guess booting up could case [09:43] not if its a 250w USP and 350 PSU [09:44] ttx: o/ [09:45] ttx: Going through my back log, wanted to catch up with you. [09:45] Daviey: I had a few questions on the status of euca 2.0 regressions wrt alpha3 [09:45] ttx: I am ontop of hggdh's bugs. [09:45] great [09:45] ttx: Well at the moment euca 2.0 FTBFS on maverick [09:45] qman__: and SuperPetRalf i have anothe rmachine here at the office i can setup if all else fails [09:45] (archive version) [09:45] Daviey: smoser told me. The jibx transition was half done on our side [09:45] if youve got backups of the old one i would go for it [09:45] eagles0513875, run some hardware diagnostics, boot live and put it under a heavy load [09:46] Daviey: just wondering why it hits so late [09:46] try taking it off the UPS or swapping the power supply, see if it fixes the problem [09:46] if it doesn't, it's probably the motherboard [09:46] ttx: The old version was only removed from the archive yesterday [09:46] arh. [09:46] qman__: making live usb of kubuntu atm [09:46] only iso i have laying around on this laptop atm [09:47] Daviey: ok, keep me posted, and don't hasitate to ask me/Dustin for support if needed [09:47] sure [09:47] ttx: Personally, i'm not convinced it needs a MIR.. but i guess i should - but MIR's currently seem to be operating slowly [09:47] MIR ? For jibx ? [09:48] ttx: yeah [09:48] it's just a package split... so it's should be a formality === ogra_ is now known as ogra [09:48] the apckage was already reviewed, it's just a matter of promoting it [09:48] ttx: the new package isn't yet in Ubuntu.. There isn't ANY jibx in Maverick at the moment, i don't think [09:49] but there was one :) Once in main, can return to main. [09:49] Yeah.. it was for this, i thought it could bypass a MIR. [09:49] Daviey: I'd really much like to have the basic UEC installer work out of the box for A3, like it used to [09:50] ttx: Agreed. [09:50] I don't care so much about the 10% instance fail rate [09:50] ttx: I need to do some investigation to the registration issue. [09:50] that we can fix after [09:50] ttx: It seems it's inconsistent.. :/ [09:50] Daviey: Ideally we need to identify the upstream issues [09:50] if any [09:51] to push them upstream today [09:51] yeah.. There is a call with them today, so i'll try and make sure as much is ready for that. [09:51] inconsistent ? I thought euca_conf --list-nodes always failed [09:51] qman__: and SuperPetRalf live usb is working booting into single user mode [09:51] what should i run [09:52] should i drop down to a root shell with networking ? or a normal root shell SuperPetRalf or qman__ [09:52] eagles0513875, with the live environment, you really should load the normal full system [09:52] GUI and all, and load some heavy application to stress the hardware [09:52] ok [09:53] ttx: That does seem to, but describe-avaliability-zones verbose, sometimes returns [09:53] qman__: normal system is loaded [09:53] and try editing your watchdog file back to norman [09:53] normal [09:53] Daviey: registration.log should leave a clear trail of what was detected and called [09:53] ping out see if your nic is working as it should [09:54] Daviey: to check of the absence of reg is due to an announce issue or a reg issue [09:56] SuperPetRalf: how can i mount the partition of the cd drive cuz on the live usb its only seeing the pendrive partition [09:57] use the mount command [09:57] go to command line and type man mount [09:58] SuperPetRalf: i know how to use it but i cant seem to find the partition listed when doign fdisk -l [09:59] has it deteced it? [09:59] how can i tell [09:59] lspci it [10:00] detected the har ddrive [10:01] ah, qman_ maybe better at this than me [10:01] ahhh here we go [10:01] go it? [10:01] had to drop down to runlevel 1 [10:01] ahh [10:01] humm but cant mount it as its not listed in the fstab [10:01] qman__: any suggestions [10:02] besides adding it to the fstab [10:04] eagles0513875, use sudo [10:05] i did and still wasnt getting listed [10:05] sudo mount /dev/sd?? /media/disk [10:06] isnt a cd "scd"? [10:06] ahh grep it [10:06] a CD could be scd or sr, but I don't know why you'd be mounting a CD [10:07] well it gives you sumthing to do :) I think eagles needs someting off the cd [10:07] my CD used to be hda [10:07] changed in an update [10:07] question [10:07] how can i get to the etc of the mounted drive [10:07] wait there it is [10:08] lol my cd used to be /dev/null [10:08] lol [10:09] lets see if that fixes it [10:09] * eagles0513875 crosses fingers [10:10] server is back up SuperPetRalf and qman__ [10:10] * eagles0513875 makes not to self not to modify watchdog.conf file [10:10] WEY! [10:10] congrats mate [10:10] ty [10:10] now to firewall it for extra protection [10:10] !iptables | eagles0513875 [10:10] eagles0513875, please see my private message [10:12] yeah, just keep in mind that there's no such thing as a secure wireless network [10:12] if you design and configure with that in mind, you won't have any problems [10:12] sure there is its when you turn it off :) [10:12] treat wireless like you treat the internet [10:12] ya [10:12] i hear ya [10:12] its encrypted but cant be too safe [10:13] :) [10:13] qman__: and SuperPetRalf would you guys recommend me getting a 2nd nic and putting that on the dmz then routing traffic to the internal nic or no need [10:13] or can i still provide content filtering with a single nic interface [10:13] i wonder if i coudl get the onboard working [10:13] eagles0513875: yes, you can. [10:14] ok cuz i want to do a lil content filtering for users of the wifi [10:14] you shouldn't need multiple NICs anywhere but at your router [10:14] is there a limit to the number of IP addresses that you can associate with a NIC in /etc/network/interfaces? [10:14] I would suggest moving your wireless to its own DMZ if you can [10:14] qman__: so you wouldnt recommend putting one nic on the dmz and then routing traffic to a nic which isn ton the dmz [10:14] but that's not always practical [10:14] we want it secure [10:14] when you saccrifice usability for security you're being too secure ;) [10:15] lol ShadeS dont get me started with this goverment fiber line we have here at this clinic [10:15] way too bloody restricted [10:15] eagles0513875, splitting your network in multiple places isn't necessary [10:15] you should split it at the router [10:15] qman__: ok [10:15] i added a second IP to /etc/network/interfaces using an "iface eth0:1 inet static" etc. line [10:15] second that [10:15] "Those that give up essential useability for security, deserve neither useability or security!" - Benjamin franklin [10:15] which works [10:15] youd end up over complicating things [10:15] but adding a third IP using "iface eth0:2 inet static" doesn't seem to work [10:16] lol [10:16] i can't ping it from the outside [10:16] jordanl, IME, if you want to use subinterfaces you have to de-configure the main interface [10:16] actually, i can ping it from something else on the same switch [10:16] but not from outside that switch [10:17] brb [10:17] qman__: de-configure the main interface? what do you mean by that? [10:18] jordanl, I have never managed a working configuration where 'eth0' and 'eth0:1' both worked simultaneously, always had to change over to a 'eth0:1', 'eth0:2' setup [10:19] i see [10:19] so don't specify anything at all in the file for regular eth0? [10:19] just `auto eth0` so it brings the hardware online [10:19] currently i have a section for "iface eth0 inet static" [10:19] I would simply change the 'eth0' in that line to 'eth0:0' [10:23] * eagles0513875 is happy that server is back online [10:24] qman__: is it possible to make this change without restarting the system? [10:25] can i just /etc/init.d/networking restart [10:25] jordanl, yes [10:25] or service networking restart [10:27] didn't quite work as expected [10:27] "ip addr show" still lists an IP for eth0 [10:27] and nothing for eth0:0, in fact, i got an error when it tried to assign an address to eth0:0 [10:28] http://paste.pocoo.org/show/242999/ [10:30] question qman__ or SuperPetRalf if i am using godaddy's dns and pointing the domain i have with them at the server ip woudl i need to have port 53 open on the server? [10:31] http://paste.pocoo.org/show/243000/ [10:31] eagles0513875, no [10:31] ok [10:31] qman__: would you suggest changing the port ssh uses? [10:31] eagles0513875, no [10:32] I would suggest using key-based authentication instead of passwords [10:32] ok will do that now gonna need some help with that but let me get the firewall up and running [10:32] jordanl, I haven't seen that one before [10:33] http://paste.pocoo.org/show/243001/ [10:33] maybe they've changed some things to make it work [10:33] that's my conf file [10:33] oh, I bet I know what it is [10:33] the ubuntu server guide made it seem so easy [10:33] try removing the IP assigned to eth0 manually [10:34] then restarting networking [10:34] ok [10:34] ip addr del 10.2.2.154/26 dev eth0 [10:34] like that? [10:35] What is the normal amount of load on a server (dual proc) with 2gb ram and a website that has ~ 20,000 pg views/day? [10:35] if its a full blown server have you ever conisred sumthink like webmin [10:35] I don't know how to do it with the ip command, I always use ifconfig, even though I guess that's the "wrong" way now [10:36] theres a ip command? [10:36] still no luck [10:36] very strange that i can't ping the .160 from the outside [10:36] more IOCTL errors? [10:37] or just the ping not working [10:37] no, the IOCTL errors went away [10:37] and it looks okay in ip show [10:37] ok [10:37] how smart is the switch? [10:38] http://paste.pocoo.org/show/243006/ [10:38] the switch is supposed to be smart [10:38] 'supposed' [10:39] that might be what's breaking it [10:39] yes, it's a cisco [10:39] if the switch is programmed to only allow X IPs per MAC [10:39] or if it's doing something like STP and just dropping the ball [10:39] i've encountered this before [10:40] and i *think* rebooting fixed it [10:45] Hi [10:45] I need some help regarding eucalyptus [10:46] i disabled the other secondary IP [10:46] still not working :( [10:48] I am doing $ sudo service eucalyptus-nc status and getting start/running, but if I do euca-describe-availability-zones verbose the max column shows all 0. When I try sudo euca_conf --no-rsync --discover-nodes it does not discover any nodes :( can someone please help me? output of the commands are http://paste.ubuntu.com/470622/ [10:48] i'll try to pick up on it tomorrow [10:48] it's late here [10:48] thanks for the help [10:48] jordanl, good luck with it [11:06] New bug: #611226 in dovecot (main) "Dovecot-common Install Error" [Undecided,New] https://launchpad.net/bugs/611226 [11:06] I am doing service eucalyptus-nc status and getting start/running, but if I do euca-describe-availability-zones verbose the max column shows all 0. When I try euca_conf --no-rsync --discover-nodes it does not discover any nodes :( can someone please help me? output of the commands are http://paste.ubuntu.com/470622/ (re-run) [11:20] Anyone who uses schroot here? [11:20] Any thoughts on why it might be running its chroot setup-stop scripts *twice* on chroot exit? [11:23] is possible to restore database from mysql files? i mean, i have just the filesystem, not standard sql backup [11:29] twb`: if you have time would you kindly have a look at http://paste.ubuntu.com/470622/ ? I set the ipv6 setting after that, discover-nodes does not even discover the node :( [11:29] !squid | eagles0513875 [11:29] eagles0513875, please see my private message [11:30] is there a wiki page for setting up squid on the ubuntu wiki [11:30] imyousuf: I don't provide support for eucalyptus [11:30] oh ok twb`, thanks, can you please point me to someone who does, I have been sitting with this for over a week :( [11:33] I can't, sorry. [11:33] twb`: np, thanks [11:44] quick question about iptables [11:44] i have port 80 open on the inbound connection how am i able to get a webpage with the outbound port being blocked? === jo-erlend_ is now known as jo-erlend [12:58] What's the current best practice for doing PPPoE on lucid? [12:59] I see that a stock install has pulled in "pppoeconf", which depends on "ppp | pppoe" -- that seems to suggest those packages are interchangable. [12:59] !pppoe [12:59] Setting up an ADSL/PPPoE connection? Look at https://help.ubuntu.com/community/ADSLPPPoE [13:00] twb`: is that of any help for ya [13:00] eagles0513875: well, reading random shit on the internet is plan B. [13:00] doesn't look random to me [13:00] Plan A is to get a recommendation from a denizen here that I trust to have a clue [13:01] (Though that recommendation could be "ubuntu-serverguide's way is the Right Way") [13:03] pmatulis: the community part of the wiki can be pretty random :-( [13:10] twb`, If you are happy to follow that, and report how it works out - if it looks good, we can move it to the offical part. [13:10] Daviey: fairy nuff. [13:11] I can't actually do that TODAY, because manglement hasn't arranged for a spare DSL modem and account to test against. [13:11] :( [13:12] I should also grovel through the C4 host that's currently doing our PPPoE [13:15] I have a shell script that writes stuff in a lofile, I want to execute the script in a way that it is restarted if nothing is written in the logfile for 30 mins, ideas? [13:19] That concept is called a "watchdog" [13:19] If you go through apt, there's probably a couple of tools to facilitate it [13:22] pseudo-code would be smth like this: if (not writing to logfile) then (restart process) unless (process finished) [13:22] I guess I want a software watchdog, not the kernel one, I'll look through apt, thnx twb` [13:22] eduardo_f: right [13:31] New bug: #611272 in tomcat6 (main) "clean tomcat6 install causes load to go up" [Undecided,New] https://launchpad.net/bugs/611272 [13:32] Of course it does, it's Java >.> [14:04] !watchdog | eagles0513875 [14:12] hi [14:13] guys how can i see what services are running on ubuntu? [14:16] diogo_79: you can look at open ports (listening) or processes running [14:17] what is the command ? [14:17] why not just do a top? [14:18] thats: [14:18] $ top [14:19] thanks [14:19] np hope thats what your looking for [14:20] i am having trouble installing 10.04 lts on a raid 1 can anyone help? [14:20] when i do sudo mysql -e to grant privileges to a user gives me access denied dont undersand why if i execute the command with sudo [14:21] Doesnt MySQL have a different user database to you linux box [14:21] diogo_79, why do you need to do sudo mysql ? [14:22] and whats the RAID issue? [14:22] diogo_79, it should work fine without suda [14:22] *sudo [14:22] because without sudo gives me access denied [14:22] where is the command [14:22] it asks me if i want to activate the raid and i say yes, then it fails to write the file system to the drive [14:23] sudo mysql -e "grant all privileges on zabbix.* to zabbix@localhost identified by 'senha';" [14:23] gives access denied to root [14:24] i get to the point of choosing a guided partitioning and i have chosen each of the options seperately and still get no where [14:26] ah afraid im not going to be much use with that sorry, but there should be someone else in the community that can have you tries the ubunut forums? [14:28] not yet, i figured i would try this place first [14:28] thanks though [14:45] diogo_79, you should be able to execute the mysql command without sudo access [14:45] as in mysql -e "grant ...." [14:45] but gives me access denied [14:45] then run it with mysql -uyouruser -p -e "command " [14:47] diaog_79: mysql command has nothing to do with sudo...mysql contains its own username/password in order to manipulate mysql stuff. [14:47] the user that you give to mysql should be the user you have set up in mysql and NOT a linux user [14:49] ttx: howdy! [14:50] mathiaz: yo [14:50] ttx: is there a blueprint to track sponsoring work as work items? [14:50] mathiaz: no [14:50] ttx: IIRC there is a blueprint for New,Undecided bugs work [14:51] ttx: what do you think about creating a server-maverick-sponsoring blueprint? [14:51] mathiaz: as in one work item per week per core-dev/motu person ? [14:51] ttx: yes [14:51] ttx: modeled after the bug-triagging blueprint [14:51] New bug: #611305 in php5 (main) "No debug symbols for libapache2-mod-php5" [Undecided,New] https://launchpad.net/bugs/611305 [14:51] hmm [14:51] ttx: server-maverick-dailytriage [14:52] ttx: the intent is the same IMO [14:52] I'm not totally convinced by that one tbh [14:52] hey mathiaz [14:52] ttx: why? [14:52] ttx: it shows up on the work item list and help to burn wi down [14:52] ttx / mathiaz: Keep in mind package sets can also sponsor, not just motu/coredev [14:53] mathiaz: having plenty of work items that you cannot do anything about until it's time to act on them... [14:53] Daviey: right - I'd suggest the WI to be named: sponsor stuff [14:53] mathiaz: a calendar sounds more appropriate [14:53] mathiaz, a BIG +1 [14:53] otherwise you'll duplicate your vacation notices [14:54] ttx: right - I'd still have to mark it as completed [14:54] completed ? or dropped ? [14:54] ttx: things I've done - so that it shows on the graph [14:54] yeah.. and as it is supposed to be a dedicated thing - it should still show on our WI tracker [14:54] Also at the start of a subcycle you'll end up having 50 WIs with only 10 you can actually work on [14:55] which will make looking for them a bit counter-productive [14:55] ttx, But the %'s increase of comepleted would then match the milestones? [14:55] so I wanted to try the dailytriage one for beta and see how well it flies [14:55] ttx: ok [14:56] ttx: my goal is to have the sponsoring work show up somehow [14:56] I think it generates noise and duplication [14:56] ttx: it seems that integrating it in the personal page is better then [14:56] I can be convinced otherwise though :) [14:56] ttx: including a calendar feed on the personal page [14:56] Well we could speak with dholbach about duplicating the 5 a day project to sponsoring stuff [14:57] another way to fix it is to do "sponsor+triage" days (community role days) [14:57] extracting sponsorship information via the LP API isn't complex [14:57] ttx: so that things that needs to be done on a specific day show up automatically [14:57] then it does not generate so much additional wi [14:58] Daviey: right - what I'd like to have is a report based on the sponsoring page but filtered for the ubuntu-server team [14:58] another thing I fear is that if 60% of the WIs are actually things that will get burned in all cases, being late in the remaining 40% might not be easy to spot [14:58] mathiaz, a TODO list? [14:58] Daviey: yes - it's all about generating TODO lists [14:59] Daviey: which are small enough to not scare people away [14:59] in a nutshell, I'm not sure we need work items to show recurrent work that we can skip in case of absence [14:59] Daviey: and getting them small is about providing views [14:59] mathiaz, Getting that should be reasonably easy.. limited to packages in the packageset [14:59] I prefer work items to be things that need to be completed [14:59] (but should we only be sponosring stuff in our set - or generic)? [14:59] rather than a weekly task reminder [14:59] Daviey: start with set, then go on generic [15:00] Daviey: the idea is to provide multiple lists to process by order of priority [15:00] ttx: fair enough [15:00] Daviey: for example my sponsoring work consists of: [15:01] Daviey: 1. https://code.launchpad.net/~mathiaz/+activereviews [15:01] mathiaz: but that's just me. That's what I mean by "not entirely convinced" and wanting to try out with dailytriage first [15:01] Daviey: 2. https://code.launchpad.net/~canonical-server/+activereviews [15:01] Daviey: 3. https://code.launchpad.net/~ubuntu-server/+activereviews [15:01] mathiaz: otherwise what's the next step, one work item with every weekly meeting ? [15:01] Daviey: 4. https://bugs.launchpad.net/~ubuntu-server/+patches [15:02] ttx: hm - I don't understand what you mean [15:02] Daviey: process each queue in that order [15:02] mathiaz: well, meetings also take time. Having one work item for each of your weekly meeting would account for that [15:03] Daviey: once arrived at the end of one queue, jump to the next one [15:03] Daviey: provided you still have time [15:03] mathiaz: if you have one to remember your "sponsoring hour" why not to remember your "weekly meeting" ? [15:03] ahh, i see mathiaz - is your process documented? [15:03] Daviey: the other part of that is that sponsoring is time-boxed [15:04] ttx, We don't have a problem attending meetings - it's also very obvious if people aren't there [15:04] Daviey: spend 1/2 a day maximum [15:04] sponsorship seems to be slipping by, hence dholbach's email [15:04] Daviey: nope - it's in my head - and I'm still experimenting - that's my personal workflow [15:04] Daviey: we shouldn't have problems attending sponsoring hours. There is a schedule for them [15:04] https://wiki.ubuntu.com/UbuntuDevelopment/CodeReviews [15:05] Daviey: to me sponsoring time is very close to meetign time. [15:05] it occurs at a specific time and takes one hour. [15:05] I cannot have opther meetings conflicting with it. === freeflyi1g is now known as freeflying [15:05] ttx: granted - if WI is defined as work to be done ASAP and sponosring/meeting as recurring event at a specific time/day, then meeting/sponosring should go on the calendar [15:05] mathiaz, Talking about last week - i really think it is awesome if people share personal workflows - doesn't need to be an "offical" one [15:06] Daviey: agreed - I'm still experimenting with my workflow and improving on it [15:06] Daviey: I plan to blog about it - and/or share with others [15:06] mathiaz, personally, i'd really appreciate it [15:06] Daviey: it's just in experimentation mode now - I'd like to wait until I can report that this is working well for me [15:07] ttx: how about server-maverick-isotesting [15:07] oh sure [15:08] ttx: this is also something that can only happen at a very specific time [15:08] mathiaz: rigth.. I like that one because it accounts for work that needs to be done on release week [15:08] ttx: so having a WI also helps to calculate the workload? [15:09] mathiaz: not the workload. An accurate completion % [15:10] if you do 10 WI per week, you should only plan 9 on release week, since one of them is taken by ISO testing :) [15:10] ttx: so how is that different from sponsoring work? [15:10] adding 1 work item every week doesn't help in that area [15:10] ttx: gotcha [15:10] again, I'm open to experiment with it [15:11] I just fear to dilute the % work to be done into recurring work that will get done anyway. [15:11] * Daviey thinks about some pretty graphs! [15:12] ttx: BTW when is the archive frozen for alpha3? [15:12] ttx: today Thursday or next Tuesday? [15:12] if you have 120 normal WI and 100 recurrent ones... The last 100 will get done anyway. But spotting that you'll be late on the first 120 gets harder [15:12] soft freeze, normally today [15:13] ttx: ok [15:13] That's why I asked for papercuts to be fixed before today, btw [15:13] hi guys, ive follwoed this tutorial here, https://help.ubuntu.com/community/Mailman, when i send an email to the new list i get the following error "pipe_transport unset in system_aliases router" in exim4 mainlog...any ideas please? [15:13] * ttx pauses [15:13] exim works to send email to system accounts and receives from remote addresses... [15:15] but when i send a email to the new list, i get that error ^ [15:16] New bug: #611316 in php5 (main) "Segmentation fault in php5-sybase" [Undecided,New] https://launchpad.net/bugs/611316 [15:22] ttx, Are you free for a chat in a few mins? Currently in a call, and would be useful to talk about some of it. [15:24] no ideas? [15:33] Daviey: I'm available now [15:33] rocking [15:34] ttx, mumble? [15:34] Daviey: I'm on [15:42] ScottK: looks like the kolab patches are on the radar now for PHP [15:46] New bug: #611330 in openssh (main) "ssh hangs after login when using broadcom wifi" [Undecided,New] https://launchpad.net/bugs/611330 === 14WAAZ29V is now known as foxbuntu [16:05] SpamapS: Yes. I've been following the discussion. Thank you. [16:08] ScottK: Hopefully the Kolab guy can help push it through when he gets back. ;) [16:08] SpamapS: I've looked at the ceph package [16:08] SpamapS: Do we need the configure stuff they mentioned? [16:09] SpamapS: and curbed LP to be able to use a merge proposal to track the discussion [16:09] SpamapS: so you may have received a bunch of uncessary emails while I was experimenting with LP to get it to do what I wanted [16:11] unnecessary emails from LP ? [16:11] what ? [16:11] [16:17] mathiaz: I think I got one or two. :) [16:18] SpamapS: let me know when you get to ceph - and I'll explain what I did to get the proposal going [16:18] SpamapS: the *merge* proposal setup === sergevn is now known as sergevn|dinner [16:20] mathiaz: I'm looking at the proposal right now.. [16:21] SpamapS: ok - so I've setup a project in LP with an empty-branch, branch it and copied the content of the source package from the PPA [16:21] mathiaz: http://code.launchpad.net/~clint-fewbar/+junk/ceph-packaging this one at least fixes the lintian report W's and E's [16:22] SpamapS: and then I pushed it to LP to create a merge proposal [16:22] mathiaz: sage has been selectively pulling changes from that branch into his own packaging [16:22] SpamapS: right - so could you rebase your +junk branch from lp:review-new-branches [16:23] mathiaz: I don't know how to use rebase [16:23] SpamapS: well - rebase was wrong choice [16:23] SpamapS: basically restart based on lp:review-new-branches [16:23] lol ok.. and cherry pick diff's in? [16:23] SpamapS: so that we can create a merge proposal and keep the discussion there [16:24] SpamapS: yeah - if you wanna keep the history [16:24] SpamapS: this is a workaround the fact that LP doesn't support commenting on branches [16:24] mathiaz: I think its appropriate that a merge proposal is where discussion goes [16:25] mathiaz: but its difficult when there's nothing to merge to.. ;) [16:25] SpamapS: yeah - that's the issue ;) [16:25] SpamapS: if you don't base your branch from an existing branch you can't create a merge proposal [16:26] SpamapS: see bug 564391 [16:26] Launchpad bug 564391 in launchpad-code "Enable commenting on a branch the same way as a merge proposal" [Undecided,Won't fix] https://launchpad.net/bugs/564391 [16:27] SpamapS: and bug 575104 [16:27] Launchpad bug 575104 in launchpad-code "No way to get review on the addition of an official branch" [Medium,Triaged] https://launchpad.net/bugs/575104 [16:27] mathiaz: so should I branch lp:~mathiaz/review-new-branches/ceph-new-pkg , or the trunk? [16:28] * mathiaz thinks [16:28] SpamapS: you can branch lp:~mathiaz/review-new-branches/ceph-new-pkg and push it to your own LP account [16:28] SpamapS: we'll start over the merge proposal then [16:28] right, and then propose to merge with yours [16:29] I don't think you have to start over [16:29] SpamapS: with lp:review-new-branches [16:29] SpamapS: my branch doesn't have anything special [16:29] SpamapS: it's just a copy of the package. [16:30] SpamapS: once we get the merge proposal rolling we'll be able to track fixes in new revisions [16:30] clint@ubuntu:~/pkg/ceph/bzr$ diff -ur ceph-new-pkg ceph | wc -l [16:30] 74 [16:30] My branch doesn't exactly have a ton of changes. :) [16:31] SpamapS: this whole thing is an experiment for me as well - we'll problem run into some bumps along the way [16:32] mathiaz: right, we usually don't have this problem because of auto-sync from debian. ;) [16:37] bzr: ERROR: Branches have no common ancestor, and no merge base revision was specified. [16:37] I'm trying to remember how to resolve this.. have done so before [16:38] Delegate the task to someone lower in the food chain. [16:39] ScottK: ++ [16:39] crap [16:39] Only amoebas and Cucumbers below me in the food chain [16:46] mathiaz: ok, pushed, merge proposed, I'm heading out for a few minutes [17:06] hi all [17:08] I just spoke to the ppl @ #httpd, and they are telling me that my hostname or host domain can't match the name of my apache2 virtual servers.. and that is the reason I'm getting a wierd error message when restarting apache2... does this make sense, and if so, is there an easy way to change the host or domain name of the box? [17:09] I get this message when restarting apache2 "[warn] NameVirtualHost *:80 has no VirtualHosts" [17:14] Daviey, working euca ? [17:15] smoser, should be - working libjibx should be in the archive - as of a few mins ago [17:20] Yosi123: your sections need xxx to be *:80 [17:23] Daviey, so there is a libjibx ? [17:24] not just a libjibx1.1 ? [17:24] do you need to rebuild euca ? [17:24] smoser, I think it should just work - there is a meta package [17:24] ah. [17:24] ok. [17:24] libjibx-java [17:24] ^^ meta package to versioned package [17:24] Daviey: Error: "^" is not a valid command. [17:24] not had a chance to confirm yet.. but it should have landed within the hour [17:25] can i patch this manually http://code.google.com/p/modwsgi/issues/detail?id=197 [17:28] SpamapS - des this look fine this is my apache2.conf http://pastebin.ca/1911072 [17:29] is it possible to make dig show all the a records for a domain? [17:29] webPragmatist: its a harmless message [17:29] SpamapS: that doesn't mean it should be allowed to fill up my logs with garbage [17:31] webPragmatist: if its happening a lot, then you can most certainly patch it. ;) [17:31] yea it does [17:31] webPragmatist: looks like 3.3 came out last week [17:32] http://cl.ly/48ee77b2abc8d80a11af [17:32] every second for a period of times sometimes [17:33] time* [17:33] anyway my dig question is more important [17:33] webPragmatist: looks like the watch file on the mod-wsgi file is a tad broken... [17:33] webPragmatist: so the debian maintainer may not know that 3.3 is available [17:33] webPragmatist: you can certainly build your own 3.3 package and install it, which is what I'd suggest over manually patching [17:34] webPragmatist: unless you have a patch from somebody else that you know will work. [17:34] webPragmatist: dig can do an AXFR if the server allows it, but most servers do not. [17:34] webPragmatist: AXFR == zone transfer == "show me all your records" [17:34] oh right [17:35] for using failover dns [17:35] or secondary dns [17:35] rather [17:35] yes [17:35] tho I tend to agree with Dan Bernstein on this that AXFR is stupid and rsync over ssh is probably a better method. [17:35] well it's not my dns server [17:36] which is why the admin probably doesn't want to show you all the records. ;) [17:36] yea… i mean i could setup axfr… but for some reason i was thinking i could look up all the a records [17:36] without some authoritative access [17:37] well some sites still do allow axfr to anybody [17:37] well ours you have to setup and acl === dendrobates is now known as dendro-afk [17:47] I cant connect to my ldap server I'm a newbie [17:49] 4sure [17:49] mathiaz: ping [17:49] zul: p/ [17:51] SpamapS: can i not just replace python with python3 from the repo/ [17:52] webPragmatist: definitely not [17:52] webPragmatist: python3 is not entirely compatible with python2 [17:52] mathiaz; do we want smbk5pwd it looks like pulling in heimdal-dev then [17:52] zul: smbk5pwd is not enabled in openldap now IIRC [17:53] zul: are you looking at enabling the overlay? [17:53] mathiaz: it is [17:53] SpamapS: ah i see [17:53] mathiaz: in debian it is [17:53] zul: hm - it may have been a change in debian then [17:54] zul: we'd probably have to remove that build dependency then [17:54] zul: or check if it can be build with mit-dev instead [17:54] mathiaz: k ill remove it [17:55] Im trying to make a local install source for netboots from a lucid dvd I have mounted. in my apt-mirror config file I have deb file:/media/cdrom lucid main - when I run apt-mirror It says Proceed indexes: [Psh: cannot open file:/media/cdrom//dists/lucid/main/binary-i386/Packages.gz: No such file but that file exists and I can gunzip -c Packages.gz and see all the meta data for packages. any idea? [17:55] also when I copied the dvd source and shared it via http as an install source I get the same problem. corrupt maybe? but when I install from the same dvd no problems. === twister004_ is now known as twister004 [18:06] what version of ubuntu does vmbuilder use when making a VM? [18:06] nvm is that the suite option? [18:10] papertigers, vmbuilder runs on the host [18:10] you can specify the suite, and it will debootstrap a version of that suite. === dendro-afk is now known as dendrobates [18:11] UEC, Eucalyptus in Static mode, do I set the cloud storage control to use the bridge interface, or the physical interface included in the bridge? [18:11] smoser: okay sweet, how do you give the vm a name? [18:12] i dont know what you mean by name [18:12] "bobby" is a pretty good name [18:12] as in virsh list shows me the vm's [18:12] how do i set that name [18:12] i dont know. i know its in the libvirt xml. [18:12] but i dont' know if vmbuilder allows you to set that or not. [18:13] smoser: foudn it [18:13] in the man [18:13] thanks [18:17] maybe im going about this the wrong way. how do I take a dvd of lucid and convert it into a source I can use for network installs? [18:27] zul: where is the daily builds PPA for php? [18:28] SpamapS: not there yet [18:30] zul: oh? that would be nice. ;) [18:30] SpamapS: yes yes :) [18:31] maek, cd /var/www; mkdir ubuntu;mount /dev/sdc /mnt/iso;cp -r /mnt/iso/* /var/www/ubuntu [18:31] maek, then http://localhost/ubuntu and bliss [18:38] does anyone know or smoser how to tell the vm disk where to go, example I want it to be on my nfs server [18:39] Hi, i have a pc with ubuntu server 10.04, it ran for month (before 10.04 with 8.04) but now the pc halt at the boot start. If i boot with a liveusb it just halt after selecting language and starting the installation. I'm sure the cpu temperature is right and the rams are ok. what can be the problem? === dendrobates is now known as dendro-afk [18:50] ugh.. I hate when people ask and bail in < 10 minutes === dendro-afk is now known as dendrobates [19:09] SpamapS: welcome to ubuntu ;) [19:14] SpamapS: can you keep an eye on those imap patches when it goes into php's svn repo? [19:15] SpamapS: its going to be a while before debian upstream is going to accept them...they havent even though of moving to 5.3.3 yet [19:16] zul: yeah, I am subscribed to the php bug [19:16] SpamapS: k....so am i [19:16] just one? :) [19:17] well....yeah...ubuntu-server-bugs gets all of it so I see it anyhow [19:17] * zul goes back to fixing php....yay [19:18] zul: the annotations patch is particularly sticky [19:18] well if its all imap stuff then its not going into the the proper php package anyways [19:19] zul: Apparently upstream (uw-imap) won't accept annotation support until rfc5464 is ratified by the ietf [19:19] SpamapS: meh [19:21] zul: no movement since 12/2008 .. This seems like Kolab's fight.. not ours. [19:21] SpamapS: indeed...we just integrate [19:21] ScottK: did you see the update regarding c-client not supporting annotations? [19:22] SpamapS: I saw that. Is c-client part of php? [19:22] SpamapS: well if we can help poke, we should - which is what you have done :) [19:22] * ScottK mostly knows about MTAs, not IMAP. [19:23] ScottK: no its the uw-imap client lib for IMAP [19:23] Right. [19:23] That patch is on kolab's list. [19:24] I'm happy to poke whoever you guys think I should poke, but I don't know if I'm sexy enough for the IETF ;) [19:25] http://kolab.org/cgi-bin/viewcvs-kolab.cgi/server/patches/imap/ [19:25] SpamapS: What does "ratified" mean? [19:25] That's not an IETF term. [19:25] If it's got an RFC number, then it's been published. [19:26] ScottK: meaning instead of just a draft that people are working on, it is published as an official IETF RFC [19:26] hm [19:26] so maybe it got published and nobody told uw-imap ;) [19:27] indeed it has been published [19:27] http://ietfreport.isoc.org/idref/rfc5464/ [19:27] Hey guys. I'm trying to resize a 2.5tb GPT partition into a 5tb one without destroying my data. The filesystem is OCFS2, so parted blows up when I try resize. Is there another tool I can use? [19:29] SpamapS: It's a propsed standard, http://www.rfc-editor.org/categories/rfc-proposed.html, which in IETF terms is pretty standard. [19:30] SpamapS: For comparison, that's as standard as RFC 2821/22. [19:32] ScottK: got a good primer on rfc procedures? i'm pretty ignorant on them [19:32] SpamapS: I saw comments (I think in the Kolab vcs) that uw-imap upstream is pretty dead. If the patches are "OK" with php upstream except needing polishing for configure time check for c-client, I think it might be ~OK for us to go ahead. [19:32] ScottK: agreed [19:33] SpamapS: Not one that reflects reality (IETF is big on theory). Bottom line is if "proposed standard" isn't enough for uw-imap, that's just a fancy way of saying "we aren't going to do it." [19:34] SpamapS: The other question this brings up is, "If php needs c-client at build time for this to work, how are we going to build it in php5 without putting uw-imap in Main?" [19:34] That may be a lot harder. [19:37] Source: php-imap [19:38] Ah. In Universe. Cool then. [19:38] Thanks. [19:52] Hmm, I wonder if I can make a collectd-plugins package like this php-imap pakage so we don't have to have all of collectd's dependencies in main [19:56] SpamapS: You might also look at clamav and libclamunrar for example. [19:57] SpamapS: What's the next step wrt the php packages? If I get Kolab to commit to updating the patch to provide configure time checks, is that sufficient? [20:01] ScottK: I think that would go a long way to helping the developers move forward. Getting c-client's upstream to wake up and take the patch would be good too. [20:01] I'm shocked to hear that uw-imap's development is that dead [20:02] if so.. then it might make sense to simply fork it and suggest to PHP that they use the kolab-c-client [20:02] I'll let the upstream's sort that out. I don't feel guilty about patching our uw-imap in the meantime though. [20:06] seems we might abandon our new 16 and 24 core machines in favour of GPUs [20:06] it's bad - what are we to do with that old iron :D [20:08] I'm running a couple servers with fresh installs, and having some issues with tftp-hpa [20:08] it doesn't seem to log anything [20:09] any idea? [20:13] it doesn't log anything [20:13] check /etc/default/tftpd-hpa [20:13] you can add options to the daemon [20:14] iirc, -v is for verbose loging [20:14] SpamapS: I sent mail to the Kolab people. Thanks for your help in this. [20:15] right, -vvvv should do the trick :) [20:17] ScottK: no problem its been fun. :) [20:18] RoyK: You can send it to me if you want. [20:18] hehe [20:44] invoks was helpful. [20:45] although apparently it is picky about where you place the -vvvv [20:45] have to do it right after the file name on the exec line, or else it gets ignored [20:49] hi, I want to automatically blacklist / remove a kernel module nf_nat_sip at startup [20:49] I created /etc/modprobe.d/blacklist-mylis.conf [20:50] remove nf_nat_sip /sbin/modprobe -r nf_nat_sip [20:50] is that the right way to force the kernel not to load that module ever ? [20:50] (i am running lucid) [20:52] No. [20:52] You probably want "blacklist nf_nat_sip" [20:52] What you did told modprobe to remove nf_nat_sip every time you remove nf_nat_sip :) [20:53] hello soren I already tried this one but after a reboot the module was loaded :( [20:53] is bind9 required? [20:53] required? [20:54] For what? [20:57] judgement call question: i have a server running 9.10. is there much benefit from waiting for 10.04.1 vs upgrading now? [20:57] pwnguin: yes :) [20:57] and that would be? [20:58] pwnguin: a number of bugs have been fixed mostly. [20:59] dont i get them from a dist-upgrade anyways? [21:00] pwnguin: If you have some project that is blocked on moving forward, then yeah you should update, but if you just want the latest crack.. I'm not sure there's a compelling reason to push forward. [21:00] 10.04 is like 3 months old now [21:00] pwnguin: https://launchpad.net/ubuntu/+milestone/ubuntu-10.04.1 [21:00] Take a look at the bugs listed there. [21:02] all the bugs in the list above marked as Fix Released are already in lucid now [21:02] the one marked as Fix Committed are currently sitting in lucid-proposed [21:03] pwnguin: so if you're not waiting for any bugs *not* marked as Fix Released then you can update to lucid now === lau is now known as Guest89521 [21:03] SpamapS: 10.04.1 is just a reroll of the isos to include the current updates in lucid [21:06] pwnguin: and BTW if you've kept your system already up-to-date then you'll already have all the fixes [21:06] mathiaz: right, I guess my point is that the .1 release will have a number of bugs fixed, and if any of them might affect you, wait. :) [21:07] mathiaz: he will? [21:07] SpamapS: well - .1 will have all the fixes included *now* in lucid [21:07] SpamapS: yes - we don't publish new fixes on 10.04.1 [21:07] SRU's go back to all releases? [21:07] SpamapS: Point releases are just an iso respin that include all updates published in lucid [21:07] I guess it makes sense. [21:08] SpamapS: when 10.04.1 is released there aren't any new updates that show up from nowhere [21:09] SpamapS: if you've installed 10.04 and upgrades as security/SRU have been rolled out, your system will be at 10.04.1 automatically [21:09] no [21:09] he has 9.10 [21:09] SpamapS: point releases are really about iso (ie insallation media) [21:09] I was offering an opinion that may not be shared by all.. [21:10] SpamapS: well - upgrades from 9.10 to lucid include SRU and updates IIRC [21:11] I guess to me, point releases are rallying points for bugs to be fixed by. [21:11] But I concede the point that ultimately, you can upgrade at any time and get *at least* the bugs that have been released up until now. [21:11] SpamapS: right. [21:12] I think we should probably *brace* for the impact of bugs from people rolling out 10.04.1 [21:12] SpamapS: a point release in Ubuntu LTS is a rallying points for new installation [21:12] SpamapS: what will also happen at 10.04.1 is that upgrades from Hardy the last LTS will be enabled [21:13] does that mean that hardy users who do dist-upgrade will get moved up to lucid? [21:13] SpamapS: not dist-upgrade - via do-release-upgrade/update-manager [21:14] SpamapS: dist-upgrade is not recommended to perform an upgrade in Ubuntu [21:14] SpamapS: and dist-upgrade doesn't know about new releases [21:14] SpamapS: (ie sources.list won't be automatically modified) [21:14] thats what I thought.. wasn't sure. :) [21:14] have never used 'do-release-upgrade' [21:15] mathiaz: thanks. i thought as much but figured theres an angle i havent considered [21:15] In Debian land we always had to go into sources.list and change "slink" to "potato" and dist-upgrade. ;) [21:15] SpamapS: right - that's what do-release-upgrade does [21:16] and a few other things [21:16] SpamapS: it also has code to handle upgrade issues that are not supported by dist-upgrade [21:16] SpamapS: do-release-upgrade (and update-manager) can be thought as "executable release notes" [21:16] theres a pool of scripts it runs for no obvious place scripts [21:18] ah cool [21:19] somewhere during my webops experience I let go of ever upgrading a server again [21:19] current OS not doing what you want? Spin up latest stable OS that does do what you want, update config-mgmt to adapt to any changes, deploy onto new server, deprecate or reuse old server. [21:20] SpamapS: Velcoume 2 Ze Kloud... [21:20] came up with a server naming scheme that supported it too.... clustername-revname-id ... so static-b-04 meant static cluster, OS rev b (CentOS 5.3 in this case), number 4 [21:21] mathiaz: I was doing this with throw-away 1U's .. but yes.. ze kloud ist der makink it fashter [21:22] mathiaz: so .. collectd plugins.. what would you say to splitting it into its own source package, much like php-imap does? [21:22] mathiaz: at issue is the list of MIR-needing libs .. if we can just leave the plugins that need those libs out of main .. *win* [21:31] i'm having trouble with nfs shares that mount at boot causing my system to hang [21:34] they are mounted at /home and /usr/local, and it looks like mountall stops the boot process when they fail to mount [21:41] New bug: #607665 in libaio "Invalid test case on ARM" [Medium,Confirmed] https://launchpad.net/bugs/607665 [21:53] mathiaz: https://bugs.edge.launchpad.net/ubuntu/+source/mysql-dfsg-5.1/+bug/551130 [21:53] Launchpad bug 551130 in mysql-dfsg-5.1 "[SRU] infinite loop in /etc/init/mysql.conf if mysqld is not running." [High,Fix committed] [21:53] you know if that one's due to customization, or does it include default installs? [22:03] Can someone recommend a better guide than https://help.ubuntu.com/community/UbuntuBonding for ethernet bonding in ubuntu === dendrobates is now known as dendro-afk === unreal_ is now known as unreal === dendro-afk is now known as dendrobates [22:34] hello, anyobody has experience with installing ubuntu 9.10 on vmware esx? [22:35] performance is rather slow and i wonder if anybody encountred an issue [22:36] doronba, I run it paravirtual in Xen 4.0 with good results [22:38] ChmEarl , there are probably big enough differences that it wont compare, but did you tweak any setting on it? specifically Apache is running really slow, and cant get why. [22:39] doronba, what choices in tasksel? I only run lamp+ubuntu server [22:41] can you run an pv kernel in esxi? [22:41] ChmEarl basic, lamp, openssh [22:42] lowridah i dont actually know [22:42] i prefer xen or xenserver since i can run most linuxes in pv [22:42] xenserver is crazy slick [22:43] lowridah i should try it, however our production farm is vmware based. [22:43] and generally it rocks, i just cant figure if the problem is lamp based or host based [22:44] SpamapS: I've never had an Ubuntu Server upgrade go bad where I used do-release-upgrade. [22:45] doronba, create this file in your www root and run apache bench http://paste.ubuntu.com/470847/ [22:45] doronba, $ab -n 2000 http://localhost/test.php [22:46] doronba, once it runs redirect results to a file and pastebin.ubuntu.com [22:48] ChmEarl http://paste.ubuntu.com/470848/ [22:49] doronba, I get this http://paste.ubuntu.com/470850/ on a C2D 3.0 Ghz with LVM on Sata [22:50] doronba, your document length should be 90kB, not 280b [22:51] ChmEarl where does it get this value then? [22:51] doronba, if the doc length is 280b, then you are getting 404 [22:52] right :) [22:52] doronba: running in KVM runs fine [22:52] sorry one sec [22:54] ChmEarl http://paste.ubuntu.com/470853/ [22:55] doronba, the transfer rate is on par with 100Mb network and # of requests is OK [22:57] but it doesnt engage the network device right? [22:57] doronba, but that ran on localhost... across the network will be lower [22:57] ScottK: Yeah, I'm not saying upgrades are bad for servers. I'm saying, in a high churn webops environment, where I've spent a large amount of my time managing servers, upgrades are sort of pointless. [22:58] ChmEarl i should test it via network see what kind of result i get [22:58] SpamapS: I think it tends to go better in Ubuntu than other distros. For most I wouldn't even attempt it. [22:58] doronba, next you can run apache bench on another VM [22:58] i think once it passes through the NIC all kinds of weird stuff happens [22:59] ScottK: yes, every debian shop I ever managed, (ubuntu server didn't exist back then) would always just upgrade when the stable releases arrived. [22:59] doronba, run ab on another VM and point it back to your target VM [22:59] ChmEarl, yup trying to set this up [23:00] ScottK: but when you're scaling up and out.. and moores law means leaving a 3 year old server on *costs* you money.. upgrades are pointless. [23:00] Right. Generally with servers as long as what you have is still supported for security, unless you need something new, staying with what's working is best. [23:01] ScottK: case in point, last company bought a pair of $30kUS servers for MySQL in 2005, and then in 2008, a pair of $30kUS servers for mysql.. that had *6 times the RAM* and CPU's that were 3x faster, and 2x the drive bays.. oh, and that used about 10% less power. [23:01] Right. [23:02] But, if you have a flat load and are more focused around having a stable server that never goes away.. upgrades that go well are as sweet as honey. [23:03] I have a development server that I upgrade every 6 months. The rest I just run on LTS (I have already upgraded to Lucid though). [23:06] yeah, traditional shops will upgrade. especially if they do something like install 6 - 9 months before LTS.. [23:06] If I had been evaluationg Ubuntu Server this time last year, I'd have chosen Karmic, not Hardy [23:07] figuring that an upgrade would be smooth and I could get the newer features that it already had. [23:07] ChmEarl http://paste.ubuntu.com/470855 [23:07] these are results from a machine on the same lan [23:08] its a jump on Timeper request [23:08] doronba, 3.4 MB/s transfer is poor, but thee are no errors [23:09] yes, so my hunch is with the negotiation with the switch where the problem is [23:09] doronba, for pv to pv VM I get about 28MB/s on a gb network [23:10] doronba, apache2 is not the problem if a straight file transfer speed is similar [23:12] ChmEarl it is on a gb network, something is wrong with network engotiation, ill look deeper. thank you for all the help [23:14] doronba, np [23:21] doronba, can you tell what ethernel driver karmic uses? lsmod === sergevn|dinner is now known as sergevn [23:24] Hi, I used the alternate install disc for meerkat to get a minimal install [23:24] ChmEarl http://paste.ubuntu.com/470859/ results [23:24] but it is installing a bunch of stuff. i wasn't asked for a software selection [23:24] i am getting compiz, erlang, a bunch of stuff i don't want [23:25] doronba, pcnet32 [23:25] is that the way the alternate install disc is supposed to work? [23:26] i really just want miminal + xorg + lxde [23:27] doronba, sometimes there are other drivers which work better in karmic [23:28] ChmEarl nay tips how do i go testing this? [23:28] Nwalins: why not install server+ubuntu-desktop? [23:29] oh [23:29] lxde [23:29] doronba, if you can specify a driver in esx, then as long as karmic supports it, try it [23:29] hello people [23:29] will give it a try [23:30] is there any GUI for managing apache and samba ? === erichammond1 is now known as erichammond [23:36] gidday, I'm just doing some maintenance on friends dead pc and thought I'd boot a live 10.04 cd. It brings up a 10.04 login prompt, username of ubuntu and nothing does not log it in. Anyone seen this before? [23:40] identify quiky* [23:41] doronba, see post #7 here: http://newyork.ubuntuforums.org/showthread.php?t=1421399 they got vmxnet3 driver working