[00:19] i need to run a process at boot. is there i guide somewhere that can help me determine the appropriate combination of runlevel and sequence numbers to feed init? [00:24] also, is there a log that records the success/failure of init scripts at boot time? [00:34] hi guys -- this is prolly a super dumb question --but im stressed and not able to figure it out --- i just compiled charybdis-IRC on a ubuntu server -- i assumed it was going to put an init script in which it didnt -- now im a bit unsure how to start the thing -- can someone lend a hand? [00:36] nm -- found it [00:36] sorry for the noise [01:27] how easy (or difficult) is it to relocate ubuntu server to new hardware.... or should i build a virtual machine as server [01:27] i need to build a new temporary server to be relocated to new hardware in the future [01:28] is it easy enought to do a direct install and relocate, or should i use a virtual machine platform? [01:29] captainkirk: Easy enough. When I've had older hardware that woulnd't boot a CD, I install in one machine and move the hard drive. I've never had a problem with it. [01:31] scottk: ok, so the 'old' hardware i will be using is PATA hdd, but i will be relocating to SCSI system... does this become a problem? [01:32] You'd need to move the image to a new drive, but I'd guess not. No guarantees from me though since I've never done it. [01:56] scottk: thanks for the advice. i think i will do a normal system install and config, with all shared data on seperated drive and system and config etc on system PATA drive [02:00] where do i configure samba data shares in the samba.conf file? === Lord_Devi is now known as Darth_Devi [02:11] New bug: #368323 in dovecot (main) "dovecot refuses connection after upgrade" [Undecided,New] https://launchpad.net/bugs/368323 [02:20] whats the best way to fix broken packages? [02:20] sudo apt-get install -f wont work :s [02:22] It depends a lot on how and why it's broken. There isn't a one size fits all answer. [02:29] anyone familiar with kimsufi? [02:30] hello all [02:30] ok, i just was informed in #ubuntu that i can install dovecot-postfix for my mailsystem, is anyone here familiar with the 9.04 mailserver metapackage dovecot-postfix? [02:31] i installed the package [02:31] now what? :P [02:32] how do i manage my users? is there a gui say phpmyadmin-like gui frontend for postfix or anything like that? thats easy to manage mail settings and users and such? or am i dreaming [02:42] billybigrigger: for a very small server you can just manage users manually. login name/password for email matches local username/password. you can disable the login shell in /etc/passwd if you don't want the user to be able to login. [03:11] question, i have an old decrepit dell 2350 that i just added a pci sata controller to. the bios on this thing is a pita. i am having trouble booting off the hard disk. i tried booting off the 1T sata drive, no such luck. put a PATA drive back on teh ide controller to give it sth easy to boot off of, and no luck there either [03:12] when i boot w a livecd, grub identifies the sata drive as hd(0) and the pata as hd(1) [03:12] i get a grub error 2 when i try to boot [03:12] I have a 2450, but I boot it off SCSI. my SATA drives are for my network shares. [03:12] I haven't tried booting it off SATA, but the SATA "RAID" menu pops up during POST, so I'm guessing it's possible. [03:13] I assume you mean a poweredge 2350. [03:13] i wish i was so luck, no dimension 2350, CL find [03:14] ahh. [03:14] it actually just worked. wth. [03:14] well, that's a good sign. [03:14] if its gonna break, at leastbe consistent [03:15] make sure everything is tightly seated. press down any chip or card that is socketed/removable. [03:15] PhotoJim, so just using local system users for mail will have their username @ hostname? ie my username would be billybigrigger@thefrozencanuck.ca [03:15] ? [03:16] billybigrigger: yup, although you can use different addresses if you have multiple domains. e.g. for user photojim@metol.ca, I create a photojim_metol account on my server and create a forwarding entry to forward all mail from photojim@metol.ca to photojim_metol@photojim.ca [03:16] there may be better ways to do it but that's how I do it [03:17] technically they'll be billybigrigger@yourmachinename.thefrozencanuck.ca but if you make machinename.thefrozencanuck.ca the mail DX for thefrozencanuck.ca, that address will work too. that's how I do it. [03:18] brb [03:18] gotta switch to gnome [03:28] hey guys can anybody tell me how big of a hard drive I should use for a email server? [03:29] that depends on a lot of things. [03:29] how many users, how much traffic they will generate, whether you're permitting imap... [03:30] well lets say for small business with 20 or so employees [03:31] a few gigabytes ought to be tons [03:32] but again, depends on how much traffic they generate... [03:32] and how often they clean out their inboxes [03:32] would 250gb hard drive will work? [03:32] and whether you have IMAP which lets them save messages in folders on your server [03:32] depends what else you have running on it, but as long as you don't eat up too much of the space on other stuff, sure [03:32] alright that is all I needed to know [03:33] if in doubt, guess high :) [03:34] I have only 3 users on my machine and 250 gigs was enough for everything including our entire music library, until I put about 300 gigs of video files up (on my new server) :) [03:34] and I have imap, and all my mail is stored on the server (that I have archived) [03:34] so 250 gigs should be plenty [03:34] now could can I do quotes on email boxes also [03:34] a friend has 150 users onto his system, uses 110 gigs. [03:36] also what is the best anti virus program to use on the email server? [03:36] I don't run any AV on mine, I run them on client Windows machines only [03:37] there are a few options, but I'm in no position to recommend [03:37] well like which ones because I only know one and that is clamav [03:38] that's a well-known one [03:39] any others? [03:39] I think so but I can't think of them off the top of my head. [03:40] well thanks for your help photojim [03:41] New bug: #368153 in kerberos-configs (universe) "Kerberos, NFS4 and autofs issue" [Undecided,New] https://launchpad.net/bugs/368153 [03:41] np Tim [03:45] anyone using screen-profiles? does it have capability to present running screens to the user on reconnect? [03:52] Hi guys, I have a little question on Jaunty and VMs... anyone able to help on that? [03:56] I have the server CD (64bit) here, and want to create a VM in VMWare Fusion. I did that with 8.04 (separate JeOS CD) and 8.10 (use F4 at CD boot time), but there seems to be no such option anymore in Jaunty. [03:57] I could install the minimal install, and then use vm-builder to make the vm, but that seems kind of odd and cumbersome. [04:16] barty, I dont think VMWare has been updated to support the 2.6.28 kernel yet. [04:17] Aj. thats a pity. [04:18] but why did they remove the option at all? The server guide still mentions the F4 option... [04:35] hey all, what part of samba.conf do i edit / add to share folders? [04:44] [nameofshare] [04:44] hey all, what part of samba.conf do i edit / add to share folders? - anyone [04:45] See above. [04:45] [homes] for individual home directories. and specific discreet shares by whatever name you like, e.g. my /public is [public] comment=Public Stuff path = /public etc [04:48] photojim: hi, thanks i will try that now [04:48] np [04:53] what command do i enter to restart the samba d after editing the smb.conf? [04:55] why cant i type anything. [04:56] i entered /etc/init.d/samba restart and seems to have worked [04:56] im having difficulty mounting 2 xfs filesystems. they are not /, i hav ext3 for that. i have installed xfsprogs, and have then addressed by UUID, only option is defaults [04:57] can anyone suggest a web based gui for manageing samba shares and users? [04:57] may be fixed, im rebooting [04:59] what does 'etc' stand for when refering to the etc folder...? [05:00] etcetera, i would assume [05:00] hmmm sounds plausable... [05:00] captainkirk: swat maaybe [05:01] slestak: swat.... samba web admin tool... looking into it now [05:02] captainkirk: for restarting services, did you use sudo? also i have started using invoke-rc.d recently, little easier to bash complete that the path to /etc/init.d [05:03] i have enabled and logged in as root... so no need for sudo...( i am still testing ) [05:06] sudo is a good habit to pickup. once you type your passwrod once, it is cached for a time. it is good in the work environement not to get in the habit of everyone runnning around as root, and leaving root sessions unatttended [05:07] slestak... i agree totally but while testing it is a right royal u know what.... :) [05:08] i didnt see your initial question, are you trying to make a smb share,our mount an existing share? [05:10] slestak: if you asking me... i was trying to share an existing folder /data/public through samba to windows workstations.... i have succeeded in that [05:10] slestak: now I am looking for a web gui (like swat) to manage samba users and shares === pace_t_zulu_ is now known as pace_t_zulu [05:11] when setting up a mailserver do i have to forward my 25(outgoing) and 110(incoming) ports to my mailserver? [05:12] billybigrigger: that doesn't make any sense. [05:13] well i can recieve mail but can't send mail [05:13] billybrigger: no [05:13] You need to set up an MTA on every server (and ideally, every workstation) you run. [05:13] These MTAs need to be configured to relay all outbound mail to the "smarthost", i.e. what you call your "mailserver" [05:14] nmap shows my smtp port 25 and my pop3 110 ports are open [05:14] i used ubuntu server's metapackage dovecot-postfix to get my mailserver running [05:14] all seems to be well except for sending mail, so what step did i miss? [05:14] what do i need to install? mta mail transfer agent right? [05:15] dovecot is an IMAP server, you shouldn't be using POP3 [05:16] MTA is the Mail Transport Agent, e.g. postfix or exim4. [05:16] dovecot-postfix opened up ports for smtp/imap/pop3 [05:16] so dovecot-postfix is for sending mail only then or what? [05:16] postfix is an MTA. [05:17] dovecot is an IMAP server, and apparently also a POP3 server. [05:17] looking in the dovecot.conf i see this for protocols....protocols = imap imaps pop3 pop3s [05:17] so ya apparently it does support both [05:18] trying to install swat, but getting unmet dependencies errors.... anyone help with this issue? [05:19] apt-get install swat openbsd-inetd and inet-superserver is not installable [05:19] so postfix does all the sending/recieving of mail? [05:20] Postfix sends mail to, and receives mail from, other mail servers. [05:20] Dovecot gives users access to their mailboxes. [05:23] so where would my problem be then? [05:23] i can recieve mail but not send it, should i start looking at postfix for my problems? [05:24] Postfix has very good logs. Look in /var/log/mail.log [05:29] cant copy files to a samba share.... probably a permission issue... can anyone advise [05:30] hmm [05:30] looking through postfix logs doesn't give me any errors as to why they wont send [05:39] how do i install a .deb file in ubuntu? [05:40] billybigrigger: Use tail -f /var/log/mail.log and then try to send. That will show you the new log entries from the sending attempt as they happen. [05:40] i did [05:40] they are timing out [05:40] Can you pastebin something? [05:43] can someone help me with my apt sources.list file.... i seem to have problems accessing and installing basic things [05:43] k ill paste from the moment i press the send message button in thunderbird [05:44] http://pastebin.ubuntu.com/159806/ [05:46] billybigrigger: What happens if you do "ping mx4.hotmail.com" from the server? [05:47] If that works, try "telnet mx4.hotmail.com 25" (if that connects, just type quit to get out of the connection). [05:48] billybigrigger@alixandria:/etc/postfix$ ping mx4.hotmail.com [05:48] PING mx4.hotmail.com (65.54.244.104) 56(84) bytes of data. [05:48] nope [05:49] just sits there [05:49] billybigrigger: Both of them? [05:49] why can't i send to hotmail?? this is odd, i can login to hotmail and send/receive no problem through hotmail [05:49] telnet times out too [05:49] i can telnet to my mailserver if that helps at all [05:49] billybigrigger: Did both ping and telnet fail or just the telnet? [05:50] both time out [05:51] You have either a basic networking problem on that box (obviously not too basic as you're connecting to it) or there is a firewall between you and hotmail for those ports. [05:51] so this is a problem with my outbound mail only, so postfix/smtp are the culprit here no? [05:51] Nothing to do with your postfix/mail server setup. [05:51] Absolutely not. You just replicated the problem with postfix not at all involved. [05:51] right, by pinging and telneting to the hotmail mailserver [05:51] ok i get ya [05:52] Yes. Thos both work from here. [05:52] Thos/Those [05:52] So leave the mail server along and work on the more basic networking problem. [05:53] maybe my isp is blocking my port 25 outbound? [05:53] Yep [05:53] If you're on a residential connection there's a good chance they are. [05:55] bah [05:55] sob [05:55] hehe well that sucks [05:55] that is like the only port i've ever had blocked, is there a surefire way of finding out? [05:55] they let me run port 21/80 [05:56] where can i find an apt sources.list file for ubuntu server 8.04.2? [05:57] my machine hardly ever finds whatever i want to install [06:00] billybigrigger: Blocking port 25 is specificaly done as an anti-spam measure. Blocking the other ones are relevant for that. [06:00] ahh [06:00] Bottom line is if you want a real internet connection, you'll have to pay for it. [06:00] can i run smtp on port 26? [06:00] You can, but no one listens on port 26, so it wouldn't help [06:01] Which is why blocking port 25 works. [06:01] but dont i just need to run my outgoing mail on port 26? [06:01] why would it matter if everyone else is on 25? [06:01] never mind [06:01] dumb question :P [06:01] :P [06:01] lol [06:02] its time for bed me thinks [06:04] almost ready to strangle ubuntu..... apt-get install (any simple package) not working - no installation candidate - what is going wrong [06:08] captainkirk: insufficient data. [06:08] twb: what do u mean? [06:11] captainkirk: There are two possibilities: 1. Installing new packages is horribly broken in Hardy and no one can do it. 2. Something is wrong with your system. [06:11] Personally, I'll go with 2. [06:20] scottk: thanks for your invaluable advice :P i have added some new sites to the sources.list and i am having some success now [06:21] captainkirk: Great. My main point is until you give some specifics, we really have no idea what's going on with your machine. [06:21] Good luck. [06:21] I'm going to bed. [06:21] scottk.... im still at work.... cheers :) [06:22] Hi [06:22] stupid question you guys, why is this wrong? EDITOR=Nano visudo [06:26] !u [06:26] U is the 21st letter of the modern latin alphabet. Neither 'U' or 'Ur' are words in the English language. Nor are 'R', 'Y', 'l8', 'Ne1' or 'Bcuz'. Mangled English is hard for non-native English speakers. Please see http://geekosophical.net/random/abbreviations/ for more information. [06:27] !ur [06:27] Aw, I like dpkg's "ur" entry [06:27] ar: Nano doesn't exist; nano does. [06:27] ar: also, strictly speaking, you should set VISUAL=nano not EDITOR=nano [06:28] hee hee, someone just pointed that out on slicehost =o) Thank you twb, I appreciate it [06:28] * twb knows all! Tell your friends! [06:28] LOL [06:37] cant write to a public samba share... access is denied [06:37] new install and i guess it is not configured correctly. can anyone help [07:32] anyone around? [07:32] wanna dance? [07:33] I try to SSH into my distant remote server and get this - debug2: we sent a publickey packet, wait for reply Connection closed by 12.345.678.90 - It's because something is eating up 100% of my 4GB ram. [07:33] What am I supposed to do? [07:34] ask support where you rent your server [07:35] um? [07:35] my house [07:35] OOM kill, increase swap, turn on ulimits [07:35] http://pastebin.ubuntu.com/159844/ [07:35] Uninstall the offending program [07:35] then go onto your server physically and see what's goign on [07:35] twb: you missed the remote part [07:36] Master_Taco: ring up the NOC monkey [07:36] huh> [07:36] That part was implied [07:37] I'm the NOC monkey [07:37] but I'm hours away from that system [07:37] If SSH doesn't let you in and you have no other remote acces mechanism (such as KVM over IP or remote power cycling), then you have to get physical access to the system. [07:38] is there any way to make ssh hold a connection longer so it doesn't time out waiting for the reply? [07:38] Master_Taco: stop typing into it [07:38] -o BatchMode=yes might also help [07:38] That turns on TCP keepalives [07:39] I'll try it [07:39] I doubt it will help in your case. [07:40] ssh is hanging up on the other end because e.g. it can't spawn the login shell [07:40] I'm 99.99% sure it's hanging because of 100% mem usage [07:41] which is probably keeping it from being able to do that - or something else similar [07:45] Master_Taco: are you saying there's *nobody* on site -- you can't even call someone and tell them to hard-reboot it? [07:46] nope [07:46] then it will wait for your return... [07:46] I just tried to call the only dick that could do it for me [07:46] Can someone help me set up ny LAMP server? I've mucked up /etc/hosts, and I don't know if /etc/resolv.conf is set up proper. [07:47] J if it not then you cannot ping a domain name [07:47] Master_Taco: you should insist some more on getting in with ssh ;) [07:47] cemc: how? [07:47] J-_: I don't know why you would need to mess with either of those files. [07:48] Master_Taco: try it some more [07:48] I'm trying with what twb said and I'm waiting for [07:48] "debug1: Checking blacklist file /etc/ssh/blacklist.DSA-1024" right now [07:48] twb: static IP, and using afraid.org's nameservers. [07:48] ssh -vvv will not tell you anything useful about the server side [07:48] Because that [wc]ould constitute disclosure of information to potentially untrusted parties. [07:49] twb: I know - I know what's going on over there though [07:49] It's just useful to see where it dies [07:49] or how far it is [07:50] to go and have no energy for the class I'm failing in tomorrow - or not to go [07:52] I can't pass it anyway.... I'm going [07:52] I'd need 101% in the final exam and final project [07:53] I can ping both my internal IP, and my domain name. But, I can't ping google from the server. What can I do to fix it? [07:55] J- can you ping grc.com [07:55] If I can't ping google, I can't ping grc.com [07:55] of sry can't ping google [07:56] then put the dns namerserver in /etc/resolv.conf [07:56] I have. [07:56] nameserver IP-of-your-DNS [07:56] what about the routes [07:57] netstat -rn [07:57] Should I include the DNS of my ISP, and DNS from afraid.org? [07:57] J whatever DNS as long as you can ping its IP [07:58] ping 4.2.2.2 gives what? === Barre_ is now known as Barre [08:00] 100% packet loss [08:00] then you do not have access at the internet at all. check your routes [08:01] can you ping your router? [08:02] Yes [08:02] My routes are messed up [08:12] bah, I don't know [08:16] well - I hard reset. I'm going to restore some stuff to a very distant backup. reinstall the package. and hope to god things work [08:16] perhaps an at job to reboot the thing at some point tomorrow [08:17] If I was smart enough to start things in screen I could be getting back already for some sleep. Even though I'm going to fail that class [08:18] ssh on a mac is such a pain too [08:19] GR! I'm getting 120K/s from the other server and it should be throwing up 2M/s [08:43] I HATE SERVER DOWNTIME! [08:44] I just made a heck of a strong white cactus [08:45] What's a white cactus? [08:45] New bug: #368420 in samba (main) "You have 1 broken package on your system!" [Undecided,New] https://launchpad.net/bugs/368420 [08:46] MTecknology: Trial by fire rocks. [08:47] cybersplice: white tequila, ginger ale, lime juice, usually sweetener [08:47] MTecknology: sans the sweetner: om nom nom nom [08:48] cybersplice: can be, I have really cheap tequila in it [08:54] MTecknology: With really cheap vodka, you can run it thorugh one of those water filter jugs to clean it up. Works a treat. [08:54] * cybersplice checks on the status of his bug. [08:54] I hate vodka [08:55] * cybersplice is alarmed to find his bug has disappeared. [09:06] New bug: #367878 in samba (main) "segfault in pam_smbpass.so (reported in syslog) (dup-of: 303458)" [Undecided,Incomplete] https://launchpad.net/bugs/367878 [10:10] does anyone else use Ubuntu for clustering? [10:15] <_ruben> clustering in which sense? [10:25] does aptitude have the ability to email me a list of available upgrades after doing a scheduled update of the repos? [10:28] frippz: apticron [10:29] atomic__: great, thanks! [10:29] yw [10:41] well, hmm... something went missing I think. got a local mail to my account saying "Mailing to remote domains not supported". guess something needs configuring? [11:35] New bug: #368510 in mysql-dfsg-5.0 (main) "the daemons don't work when are launched in startup, but they do when u restart them manually" [Undecided,New] https://launchpad.net/bugs/368510 === cjwatson_ is now known as cjwatson [11:53] !libltdl3 [11:53] Sorry, I don't know anything about libltdl3 [11:53] dfs [11:53] !info libltdl3 [11:53] Package libltdl3 does not exist in jaunty [11:53] !info libltdl7 [11:53] libltdl7 (source: libtool): A system independent dlopen wrapper for GNU libtool. In component main, is optional. Version 2.2.6a-1ubuntu1 (jaunty), package size 286 kB, installed size 376 kB [12:03] New bug: #363275 in postfix (main) "package postfix 2.5.5-1.1 failed to install/upgrade: " [Undecided,Incomplete] https://launchpad.net/bugs/363275 [13:28] Is it possible to use vmbuilder with apt-cacher? I use --install-mirror= and it seems to want to look for dists_jaunty_Release - which apt-cacher doesnt have [14:11] New bug: #368141 in openvpn (universe) "openvpn init script contains invalid variable declarations (dup-of: 271777)" [Undecided,New] https://launchpad.net/bugs/368141 === ssd7 is now known as Guest10436 [14:42] Hi all, what do people use for backup tools on their servers [14:43] TheOpenSourcerer: rsync [14:43] nothing \o/ (I love having a machine where every byte is disposable) [14:44] I have used rdiff-backup at home but I want something for a customer that is "easy" and the backup data can be dumped to a remote machine which might not be Linux. [14:45] for the package selection in jaunty jeos install, how can I tell what is included in each selection? Particularly "Basic Ubuntu Server"? [14:45] TheOpenSourcerer, unison-gtk perhaps (depending on complexity), or backuppc [14:45] i have looked in teh server docs at Ubuntu.com [14:46] Thanks Kamping_Kaiser I will investigate those. [14:46] and tried google. [14:46] TheOpenSourcerer, backuppc does (or did) have quite a setup cost (eg, effort), but after that its easy. [14:47] slestak, check its tasksel task (sorry, cant recall how off the top of my head though) [14:47] Kamping_Kaiser: ty [14:48] can I do that during the install, or at packages.ubuntu.com? [14:48] Kamping_Kaiser: nm, you just said you couldnt recall :) [14:48] slestak, during install perhaps, but not sure. p.u.c probably not. [14:55] found a link, the tasksel selections are stored in /usr/share/ (forgot the rest) and that dialog can be brought up after install with sudo tasksel === jorge_ is now known as jcastro === mogul218 is now known as mogul218- === mogul218_ is now known as mogul218 === Darth_Devi is now known as Lord_Devi [15:25] i'm having trouble mounting an iSCSI device during boot time. during boot i get "waiting for iscsi devices" and then times out. once up, the device is present and ready to be mounted. plus, putting an entry in fstab crashes everything (i drop to a root shell). i'm running Jaunty. any ideas? [16:01] do many people here use pidgin for irc? I'm using it now, for the first time, and liking it so far. [16:01] New bug: #283811 in cups (main) "cups says "can't promp for authorization" when printing to my samba printer" [High,Confirmed] https://launchpad.net/bugs/283811 [16:11] kirkland: have you seen my comment on your blog? === ssd7 is now known as Guest38527 [16:14] ivoks: just saw it [16:14] ivoks: i agree 100% [16:15] ivoks: i'm planning on pushing this hard for karmic [16:15] kirkland: great! [16:15] ivoks: as much as ubuntu=gnome, kubuntu=kde, xubuntu=xfce, server should =screen [16:15] ivoks: we should use it in the same way [16:16] server=screen|ebox|some-cool-ldap-desktop-app-for-workstation [16:16] :) [16:16] ivoks: sure [16:16] ivoks: those are welcome to play too === Guest38527 is now known as ssd7 [16:17] I'm trying to instal ubuntu on a USB stick. After install of both desktop or server versions, after some/one reboots the filesystem is always corrupted, leading to unreadable system files. Is this a known issue ? some fix for this? some clue ? === ssd7 is now known as Guest98068 [16:18] ivoks: https://blueprints.edge.launchpad.net/ubuntu/+spec/server-karmic-screen-window-manager [16:18] ivoks: you're welcome to add your thoughts/support to the whiteboard [16:18] sure [16:19] kirkland: but i wouldn't launch screen by default on login... [16:19] launching screen by default... for which user? [16:19] kirkland: i would rather launch it by default when loged in via ssh [16:21] ivoks: so not on the tty? [16:21] hm... no :/ [16:21] ivoks: actually, keybuk has been talking about replacing the tty manager with screen [16:21] it should be configurable [16:21] ivoks: that's fair enough [16:22] some programs won't start if screen isn't 80x24 [16:22] so, if we push 'panel' in the window... and we can't get out of it... then we have a problem [16:23] ivoks: like what programs? [16:24] some ncurses based [16:24] i don't recall exactly which one, but they do exist [16:24] maybe sniffit [16:25] ivoks: then i think we should bug those programs and fix them [16:25] ivoks: 80x24 is pretty draconian in modern linux, honestly [16:26] ivoks: i mean, i don't mind writing code that fits in 80 columns [16:26] nope, it's not sniffit [16:26] ivoks: and even when i did my demo yesterday, i made my demo window 80x24 [16:26] kirkland: i know... i'm just reminding... they exist :) [16:26] ivoks: but that was out of convenience and not necessity [16:26] ivoks: fair enough, but I don't think that should be a blocker [16:26] i agree [16:27] ivoks: we should bug any programs that institute such an ancient requirement [16:27] ivoks: and fix those in main [16:27] ivoks: and try to fix those in universe [16:27] ivoks: you're welcome to note that on the whiteboard ;-)\ [16:33] p6pt I just installed the new ubuntu 9.04 "dovecot-postfix" package on my slice--how do I figure out what and how it defines email accounts in this turnkey setup? [16:36] mib_z3mjzb6b, i just installed the same metapackage last night [16:36] mib_z3mjzb6b, and its ready to go out of the box, its uses your system users for email accounts [16:37] mib_z3mjzb6b, i just had to setup my MX records with my domain provider and i was ready to go i accepted emails out of the box with no config, except for in my case my ISP blocks port 25 so i can't send email out through my mailserver [16:37] billybigrigger, hmm... I'm using a user I created but it is rejecting me on the password [16:38] odd.. [16:38] I usually use this user over ssh without using a password, wonder if that matters? [16:38] you use ssh without a user password? [16:38] using a key [16:39] it is good to get a point of reference from you [16:39] knowing that it should work out of the box helps--now I can look at the user and try making a new one to see [17:23] ivoks, any other package that you have in mind? https://bugs.launchpad.net/~ubuntu-ha/+packagebugs [17:24] RoAkSoAx: ocfs [17:25] RoAkSoAx: that's ocfs2-tools [17:25] ivoks, just added it [17:26] RoAkSoAx: so i had an idea of replacing redhat cluster suite with linuxha [17:26] RoAkSoAx: since no one has good knowledge of rhcs and it's hard to support it [17:26] ivoks, that would differentiate us if we manage our *own* cluster stack [17:26] RoAkSoAx: correct; having ubuntu cluster suite would be great [17:26] ivoks, indeed, i think the maintainer was unable to keep with them right? [17:27] RoAkSoAx: well, maintainer was hiered by redhat :D [17:27] ivoks, yeah it would be a nice idea, keep in mind that the CRM for Heartbeat is now migrated to another project called Pacemaker [17:27] RoAkSoAx: well, i'm a big rusty as far as linuxha goes [17:28] it's been a while since i used it last time [17:28] ivoks, haha ok, well.. last heartbeat version that includes the cluster resource manager is 2.1.4, the latest i think it is 2.99.0 and it does not include the CRM... and that's the new project, Pacemaker [17:29] ivoks, http://clusterlabs.org/wiki/Main_Page [17:30] hm... [17:30] openais [17:30] rhcs uses openais [17:30] ivoks, well i'm currently working with Heartbeat 21.4... but we'll have to start thinking on having heartbeat 2.1.4 and heartbeat 2.99.0 or up with Pacemaker [17:30] ivoks, yeah, now the CRM supports both heartbeat and openais [17:31] pacemaker looks promissing [17:31] indeed [17:33] ivoks, maybe we could provide a ubuntu cluster stack with both heartbeat or openais [17:33] * jmedina is interested with HA topic.. [17:33] RoAkSoAx: we should decide on one tool [17:34] otherwise, we will have hard time to support it [17:34] i don't have strong feelings about openais [17:34] ivoks, indeed... but actually it will depend on what type of cluster we'll like to implement [17:34] failover, load balancing [17:35] gfs can't work without rhcs, right? [17:35] ivoks, in case of loadbalancing, we could use keepalived instead of heartbeat if we are implementing lvs based clusters, since it failovers way faster than heartbeat [17:35] that leaves us with ocfs as only clustered filesystem [17:35] ivoks, we could also think about glusterfs [17:35] is that thing alive? [17:35] Hello there [17:36] ivoks, yeah, there's a new version coming soon, version 2. It looks promising [17:36] I have Ubuntu Server 7.10, What's the best way to upgrade to 8.04? [17:36] henriquelm: do-release-upgrade [17:36] it allows us to do active / active without using ocfs2 since it's already a cluster filesystem [17:36] RoAkSoAx: glusterfs isn't in the kernel [17:37] RoAkSoAx: so, ocfs would be easier approach [17:37] ivoks, Do I get to choose which version will I upgrade to? [17:37] ivoks, no, it's just userspace [17:37] henriquelm: it upgrades to the next one - that's 8.04 [17:37] ivoks, ok [17:37] henriquelm: you can't skip releases [17:37] henriquelm: so it always upgrade to the next release [17:37] ivoks, well at work we've been discussing the utilization of glusterfs vs drbd, and they've seen that glusterfs since it's easuy to maintain, upgrade.. because it works on userspace [17:38] ivoks, I already downloaded the cd of ubuntu server 8.04, can I upgrade from the .iso file? [17:39] henriquelm: you can, but you have to do lots of dirty hacks... if you have network link, just do it over the network [17:39] userspace = slow, kernel = fast [17:39] :) [17:39] ivoks, yeah that's true too.. i'm a fan of DRBD aswell :) [17:40] RoAkSoAx: we have to check all options [17:40] yes [17:40] i'm a fan of linux ha clusters [17:41] it's good to know there's more people that use ubuntu in clustered env [17:41] i have couple of rhcs+drbd clusters [17:41] on ubuntu [17:42] ivoks, yeah!! when i was thinking what thesis could I do... I was amazed with HA... so i decided to do my thesis about it... and ended up using Heartbeat, LVS, DRBD to design my architecture [17:44] RoAkSoAx: ok... so for start [17:44] RoAkSoAx: for karmic, we should define cluster set [17:44] RoAkSoAx: and for karmic+1 have a production usable cluster system [17:45] we can't continue having rhcs in the state it is now in next LTS [17:47] ivoks, i see.. well my personal opinion is to use linux-ha, though i've not used rhcs... i've been working with linux-ha for more than a year know and i do not have complaints at all [17:48] RoAkSoAx: i'm in favour of replacing rhcs with linux-ha [17:48] btw... i had this idea: http://brainstorm.ubuntu.com/idea/13315/ with it, I intended to have an easy way to install and configure clusters based in linux-ha and LVS for loadbalancing [17:50] RoAkSoAx: let's build fondations first, and then create management console [17:50] ivoks, haha yeah, i had that idea for hardy UDS [17:50] anyways, yeah! i go for linux-ha rather than rhcs [17:52] ok :) [17:53] it would be great to have rhcs->linuxha migration tool [17:53] :) [17:53] * jmedina downloading RoAkSoAx thesis [17:54] ivoks, indeed [17:54] too many plans = no realization [17:54] ivoks, for example, this bug https://bugs.launchpad.net/ubuntu/+source/heartbeat/+bug/332043 addresses the issue of the separation of heartbeat with the CRM [17:54] Launchpad bug 332043 in heartbeat "provide pacemaker on hardy 8.04 as heartbeat-3" [Undecided,New] [17:54] so, let's create basic foundation for karmic ;) [17:55] and for example, instead of creating a package heartbeat-3 we could start having like: ubuntu-cluster-stack or something like that [17:55] which will be a metapackage for latest heartbeat (2.99.0 and Up) + pacemaker [17:59] jmedina: link? :) [18:00] http://roaksoax.files.wordpress.com/2009/03/thesis.pdf [18:00] spanish? [18:00] it is in spanish [18:00] yeap [18:00] ivoks, yes [18:00] ok, let's try :D [18:00] ivoks, here's a summary: http://www.roaksoax.com/2008/07/ubuntu-in-my-thesis-part-2 [18:07] RoAkSoAx: looks really good your thesis, im researching for a outgoing project, so that will help me a lot [18:07] heh, spanish is easy :) [18:08] jmedina, awesome.. i'll soon update part of the implementation with Jaunty. since I think I used Gutsy for it [18:09] RoAkSoAx: good, im suscribing to you rss :D [18:10] ivoks, the upgrade over the internet is way to slow, can I stop in the middle of the download? [18:10] ivoks, my thesis should help us decide a cluster stack and as a sample implementation [18:10] henriquelm: yes, you can [18:11] ivoks, where does ubuntu keeps these downloaded files? [18:11] henriquelm: /var/cache/apt/archives/ [18:34] ivoks, should be announce the creation of the ubuntu-ha team and accept members to it that would like to contribute or something? [18:35] RoAkSoAx: sure [18:35] ivoks, ok i'll do a blogpost later today [18:42] ivoks: are you using any vacation/autoresponder program for your postfix/dovecot/ldap setup? [18:43] jmedina: i'm looking at gnarwl [18:46] ok, I think it is the only one :D [18:48] :) [18:51] I used maildrop with mailbot, but that requires manual creation of .maildrprc files [18:52] using gnarwl it is better, now we need a web interface so users can change their vacation info [19:27] does Ubuntu DVD contain all the server-stuff too? [19:28] I think yes some os the repo's [19:28] what does that mean? :) [19:28] It includes some of the packages that are in the synaptic [19:29] right. I know that, I hope it includes serverstuff too. [19:29] Remember that using packages from the DVD or CD is not using some of the latest realeses [19:29] I seem to remember that it does. [19:29] So maybe you can isntall them but rember to do and upgrade later on [19:29] it doesn't include updates, I know. There aren't any big differences though. [19:30] Security Updates must be a consern [19:30] yes, but what has that got to do with anything? [19:30] I'd have to install updates if I installed from the net too. [19:32] The thing having the packages in the DVD are that you don't have to connect to the net because, maybe you are out of the net [19:32] but still when you connect you need to upgrade [19:32] exactly. Or just to save some gigabytes of bandwidth. [19:32] yesyesyes :) [19:32] of course if the computer is not connected to any networks you don't have a problem [19:32] yes [19:34] hi, does ubutnu 9.04 runs fine on quad core xeon xpu? [19:34] I have never installed it before on quad core [19:35] Izantal - yes [19:37] lzantal, the number of cores is mostly irrelevant. [19:38] great, the server comes with 3 sata II raid 5. On 8.04 sometimes I had issue with sata raid. Is 9.04 would recognize the hdd without much hassle? [19:38] I'd think so. Alot of people prefer to use linux' raid functionalities, but I don't know much about it. [19:39] I usually end up using software raid on them. [19:39] It would be nice if it would just work. [19:40] Thank you for all the pointers. I guess I just have to get the server and try it :D [19:40] Good luck. [19:40] the chance I'll say something stupid is directly proportional to the number of messages I send regarding this topic. :) [19:41] however, I think jaunty server has enhanced its LVM setup options, if that's something you can benefit from. [19:44] XiXaQ: thank you i will look into that. [19:46] May be somewhat off topic question:: Last week I did a mass local network install of 9.04 desktop on old icompaq desktops without cd-rom drive [19:47] I'm going to install a 64bit desktop server at home tonight. I'm very excited about it. I haven't used 64bit desktop in a couple of years. I guess the server stuff should be the same as in 32bit, but I'm curious to see how well the desktop stuff will work. [20:54] can Oracle Database XE and MySQL run in parallel ? [20:55] orudie, I see no reason why not. If they use the same ports, then you'd have to configure one of them to use a different port, but other than that, there should be no conflicts. I also doubt that they would use the same ports. [20:57] Morning [20:58] i require a new version of ImageMagick than is currently avaialble in the default repos. Is there a untested or beta repo i can use to obtain the latest? [21:00] E: Couldn't find package oracle-xe :( [21:08] ok, I am running 8.04 LTS. can i use IM from the jaunty repo without actually moving to jaunty? [21:10] slicslak: You could add the deb-src of a later dist and then apt-get source and build it === jorge_ is now known as jcastro [21:24] it seems i'm don't have the right packages to compile things with yet. is there a meta package for this? or should i just isntall gcc, etc. [21:24] nvm, found https://help.ubuntu.com/community/CompilingSoftware [21:31] just installed oracle-ex , no problems during installation, trying to access oracle by going to http://my.ip.adr:8080/apex geting page load error [21:31] hi folks, I have a virsh xml question for you guys: when I specify a , , , etc. the paths given are relative to what exactly? [21:31] orudie try another browser [21:32] check that's the correct port [21:38] My port is plugged [21:38] need drano [21:41] New bug: #368851 in bacula (universe) "package bacula-fd 2.4.4-1ubuntu5 failed to install/upgrade: subprocess post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/368851 [21:49] mathiaz: when you use your iso testing scripts, are you able to virsh -c qemu:///system "console [id]"? I'm getting: No console available for domain [22:03] anyone around? [22:04] i'm getting random kernel panics in 7.10 lts server running on vmware... [22:04] I-Am-Sam [22:04] :) [22:04] lol [22:05] err, 8.04 [22:05] hardy [22:05] anywho, it likes to randomly panic [22:06] randomly panic/ [22:06] ? [22:06] yeah [22:06] like... poof, panic [22:06] it'll work fine for a few hours [22:06] what do you mean by panic? [22:06] resources? [22:07] kernel panic [22:07] and halt [22:07] the usual 'killing interrupt handler' bit [22:07] halt and catch fire, etc :) [22:11] Am-I-Sam, you should look into /var/log/syslog /var/log/messages, also what other software is installed of importance (mysql, apache, ect) [22:11] its just serving NFS [22:11] and is an ldap client [22:12] Sam-I-Am, did you recently pull updates? [22:12] Sam-I-Am, also, do you like green eggs and ham? [22:12] yeah... the box was installed recently too [22:12] hahaha [22:12] depends [22:12] Sam-I-Am, which kernel are you using? [22:12] uname -r [22:12] lemmie reboot it from its panic, one moment... [22:14] 2.6.24-23-server [22:14] hey guys I am having problems logging into squirrelmail I have used this setup :http://flurdy.com/docs/postfix/ [22:14] TimReichhart, I would go ask the postfix guys #postfix they helped my fix mine sometime ago [22:15] in squirrelmail there more issues regarding to IMAP server than SMTP (postfix) [22:16] Sam-I-Am, sudo apt-get update && sudo apt-cache policy linux-image-server [22:16] jmedina, correct but the postfix guys sure know allot about both [22:17] foxbuntu: of course :D [22:18] foxbuntu: 2.6.24.23.25 ... installed and candidate [22:19] well I can send out just fine but i cant login to squirrelmail [22:20] i'm seeing some sunrpc oops throughout the day before it panicks... [22:20] probably me restarting nfs parts [22:20] Sam-I-Am, you on a sun server? [22:20] but those still shouldnt happen [22:20] Sam-I-Am, what hardware are you on? [22:20] nope... this is a 64-bit VM inside vmware esx [22:21] ah [22:21] yeah that shouldnt be an issue [22:21] i assume you have vmware tools installed? [22:21] yup [22:21] build 153875 [22:21] precompiled modules worked ok [22:21] and in vmware you have ubuntu selected as the os [22:21] yup [22:21] k [22:22] well back to the logs then [22:22] all of my other ubuntu servers are in VMs too... 8.10 and 9.04... no problems yet, although none of them have served NFS either [22:22] just have a poke around in all the logs and see if anything sticks out [22:22] the actual panic data is not in any logs... so it dies pretty hard. [22:22] you could try shutting NFS down for awhile to see if it panics without that running [22:23] that'd probably work... just need to stop my clients for a bit. [22:23] yea [22:23] thats prob the easiest thing to try [22:24] mmmkay... [22:25] should i just make sure nothings mounting it or stop the NFS daemons too? [22:25] stop the daemons [22:26] I have my doubts that its going to be the problem, but its an easy test [22:27] sure [22:27] well gl...I have to run for awhile [22:28] foxbuntu: thanks [22:59] cr3: I'm using virt-viewer to connect to the console [23:00] anyone used ConVirt (formerly xenman) ? [23:10] mathiaz: I need to jet but I found vncviewer to work quite well too, cheerio [23:51] Bhooooo!