[00:07] i am thinking of using 10.04 for my SMB server. I have several questions that I would like to get answer [00:09] Oops, -- to. first off , why does ubuntu use exim as its mta, in the tasksel setup. Is there some advantage to exim over postfix? [00:11] alt, could somebody point me to a paper or other discussion as to why? [00:29] how do I make a networking device come up by default on boot? [00:29] I've added it to /etc/network/interfaces, and ifup eth0 works... it just doesn't happen when the machine boots. [00:31] The http://tinyurl.com/ubuntuserv link in the topic is broken. [00:31] tinyurl: Error: Unable to find site's URL to redirect to. [01:22] anyone here run an nfs file server over a gigabit network? [01:37] sorry, only 100Mbit here so far [01:38] how do I password protect a pdf file.. i need to send it to someone over the email .. i will call the person and give the password over the phone .. [01:39] on a server..? [01:39] might be offtopic for this channel.. will ask #ubuntu or #ubuntu-offtopic [01:40] you could encrypt the entire email instead.. :) [01:46] I have a 'server' (really an old desktop) that now says http://sprunge.us/CaIV about NX... but I don't think the processor supports NX - certainly I don't see a setting for it in BIOS. How likely is it that check-bios-nx is wrong here? [01:47] Zider: i am using gmail... i will just encrypt using openssl and give him the password on separate email [01:47] axisys: gmail webmail or via mail application? [01:47] gmail webmail [01:47] Zider: ^ [01:48] `check-bios-ns --verbose` says http://sprunge.us/MiXI [01:48] i just need to encrypt the attachment .. and we both are in US [01:51] axisys: yeah, you can encrypt with any tool and just mail it then [01:52] Zider: yep.. looking for the syntax with openssl [01:52] or rar, it can encrypt too [01:52] openssl enc -ciphername [-in filename] [-out filename] [-pass arg] <-- looks like this will do [01:53] hi all [01:53] "Doc and resources: http://tinyurl.com/ubuntuserv" -> it's broken ;) [01:54] hey, I dist-upgraded to lucid which removed postgres-8.3 so now I cant access any of my databases? - and it seems I can't reinstall 8.3 as it's no longer available. What do I do? [02:03] hackeron: check out pg_upgradecluster [02:05] nick58b: thanks [02:09] Zider: like rar.. thanks [02:09] Zider: tested .. works pretty good [02:25] I am trying to load 10.4LTS on my server but it fails even though the cd was verified.... Fails at every step once I get to "select and install software" step.. If I continue it will fail with Grub no mater if I install grub 1 or 2 and also fails if I try LILO... Am I missing something? I am installing the 64 bit system on a 64 bit intel system. [02:35] hi folks, I wonder why ubuntu 10.04 take more than 60 minutes to install using the alternate cd installer. [02:36] New bug: #578587 in samba (main) "umount.cifs does not work on logout with libpam-mount" [Undecided,New] https://launchpad.net/bugs/578587 [02:49] anyone know why the images are not loading on my site? http://ec2-204-236-201-28.compute-1.amazonaws.com/ [02:50] it's weird [02:51] RickyWh: The webpage at http://graffitihousewars.com/sliver.png might be temporarily down or it may have moved permanently to a new web address. [02:51] oh wait, those are linking to the domain, right [02:51] ok thx :) [02:51] np [02:52] i was worried there for a few lol [02:52] Also, I'm not sure if url(...) is valid CSS. You might want url('...') [04:05] I'm upgrading my server through ssh from 9.10 to 10.04. My VPS is based on Xen. Should I upgrade to GRUB 2 or there are problems with GRUB 2 and Xen? === mikelifeguard is now known as Mike||gone === clintb is now known as SpamapS [05:40] hi [05:40] can some one help me with the usage of mcelog ? [05:43] I have got this http://fpaste.org/z9XT/ [05:45] I am running ubuntu 8.04 server [05:46] i did /usr/sbin/mcelog --ascii b200000000070f0f it did not worked for me [06:44] Any ideas what would be causing this? sshd[32308]: [06:45] checking in again for my query ? === swift_ is now known as twister [06:58] sorry to ask here, but is there a free vnc server to put on a windows vista system ? [06:59] I have tried RealVNC but can't connect to it from ubuntu [06:59] although the port is openned [06:59] maxagaz: I don't know. Try asking ##windows. [07:00] IIRC on NT5.0 and NT5.1 I used Tight VNC. [07:00] I expect your VNC server is running, but Vista is firewalling it somehow. [07:01] (Obviously, using Windows' native RDP server would be VASTLY preferable to VNC.) [07:04] yeah there's a native RDP client called rdesktop for X [07:04] works quite well most of the time [07:04] And from NT5.1 onwards, even single-seat Windows SKUs ship with the RDP server pre-installed (but off by default). [07:05] From memory you hit "properties > advanced" on the "My Computer" icon and turn on RDP, then go into the MMC and add your user to the appropriate "remote access" group. [07:10] time on my machine is an hour off. how can i fix it? [07:11] timboy: ntpdate [07:11] 11 May 00:11:18 ntpdate[2107]: no servers can be used, exiting [07:12] timboy: what release? Lucid? [07:13] twb, how can I tell from command line? [07:14] lsb_release -a [07:15] twb, 8.04 [07:15] ntpdate-debian, then. [07:16] twb, that updated the time properly. should I put that in my cron jobs? [07:16] No. [07:16] ntpdate-debian should automatically be run at boot time, and ntpd should be invoked at boot time. [07:16] As at 8.04, ntpdate[-debian] is responsible for large jumps, and ntpd performs incremental resyncs. [07:17] twb, this system never reboots it's been slowly losing time daily about 2 mins a day or so [07:17] (In later releases, both roles are performed by ntpd.) [07:17] timboy: that suggests that ntpd is either not installed or not being started at boot time. [07:18] twb, ntpd doesn't appear to be installed should I install ntp or openntpd? [07:20] ntp. [07:21] ok thx twb [07:38] How do I auto mount a usb drive at boot? [07:38] dominicdinada, add it to /etc/fstab [07:39] as in the drive name ? [07:39] !fstab [07:39] The /etc/fstab file indicates how drive partitions are to be used or otherwise integrated into the file system. See https://help.ubuntu.com/community/Fstab and http://www.tuxfiles.org/linuxhelp/fstab.html and !Partitions [07:43] dominicdinada: no, the UUID or LABEL [07:44] ok and if the drive isnt there will it just skip it? === dendro-afk is now known as dendrobates [08:21] New bug: #578665 in dbconfig-common (main) "package dbconfig-common 1.8.44ubuntu1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 10" [Undecided,New] https://launchpad.net/bugs/578665 [08:27] New bug: #578668 in awstats (main) "Please sync awstats from Debian unstable" [Undecided,New] https://launchpad.net/bugs/578668 [08:44] i would like to create a folder in which any user can create a file, yet only root can modify or delete the files once they're created. [08:44] i would appreciate any help in that matter [08:45] Can't be done with the POSIX DAC. [08:45] if the SUID bit worked on directories, this would be quite easy [08:45] really!? this is a totally impossible task? [08:45] High bits have different meanings on directories (compared to binaries). [08:46] l3iggs: yes, totally impossible. [08:46] well darn [08:47] The closest you can get is the permissions on /tmp, which (broadly) allow anyone to create a file, but only the owner to subequently modify it. [08:47] Depending on your wider goal, you might want a set-uid program that allows students to submit files, but not to delete them. [08:47] yeah, that's a sticking point, i don't want the owner to be able to modify or delete it [08:48] Another approach would be to simply have students submit their assignment via email, and have a procmail script place it in a world-readable, owner-writable directory. [08:49] let me elaborate, [08:49] i'm running a web server in which users upload files through a web portal [08:49] the file appears on my hard drive with www-data as the owner [08:50] i don't want the next user to be able to delete this file through the web page [08:50] or modify it for that matter [08:50] That's up to your crappy web app, then [08:50] that's something that would normally be handled in the web application [08:51] yeah, i'll have to modify the web app [08:51] i was hoping to avoid that [08:52] i appreciate your help [08:54] If you REALLY didn't want to edit the web app, you could do something totally horrble [08:54] go on... [08:54] e.g. write a FUSE module along the lines of fuse-rofs that only allows OPEN write syscalls, and not much else. [08:55] And returns ENOPERM or so for everything else. [08:55] Possibly a MAC like apparmor or selinux could express the same thing, but I doubt they're set up for that level of granularity [08:57] the web app is open source and fairly heavily developed, i think i'm better off looking for support from its devs [08:58] * twb grumbles about web apps [08:58] i thought there might be some quick permissions or ACL trick that i don't know that could solve the problem easily [08:58] but i guess not [08:58] l3iggs: not trivially, no. [08:59] well this points me in the right direction. i really appreciate the help [09:07] hreyt [09:07] hey! [09:08] so, does 10 come with nicely colored terminal colors [09:09] KMS? [09:09] jpds: define "nicely colored terminal colors" === dendrobates is now known as dendro-afk [09:10] jnss: ^--. [09:11] doing ls -la shows files colored by their type [09:15] the new terminal is slightly transparent, which is kinda neat [09:15] by the way, [09:15] if anyone was reading on about the issue i brought up earlier, [09:15] i have solved it [09:15] with an .htaccess file that contains: [09:16] [09:16] Require user root [09:16] [09:16] Oh, nice idea [09:16] :-) thankyou [09:16] Of course it assumes your httpd will honour .htaccess files. [09:16] allowoverride authconfig :þ [09:17] RoyK: well, I normally run busybox httpd, not apache httpd. [09:21] New bug: #578681 in apache2 (main) "apache2 crashes on svn query with kerberos + ssl" [Undecided,New] https://launchpad.net/bugs/578681 [09:35] i got this error during update E: Sub-process /usr/bin/dpkg returned an error code (1) [09:37] http://pastebin.com/YfGXvftM [10:30] Does anyone know how you would configure squidguard to block all except specified domains? I'm using 'pass !allowed none' but it's still blocking everything. [10:40] Hey People! Im trying to figure out, where an old partition started. The partition before ends at cylinder 6879 according to fdisk. I tried 6880 for the next partitions, but the fs (reiser) cannot be mounted. Can it be, that the next partition starts at 6879? Can one partition start in the cylinder in wich the one before ends? [10:46] iclebyte-work: pastebin your squid.conf so I don't have to go find mine [10:46] one second! [10:52] Holy wavering attention span, Batman! [10:53] twb, http://pastebin.com/ZpQNzjUS [10:54] iclebyte-work: that's the whole file? [10:54] how to use reverse proxy with ubuntu using apache [10:56] sporedi: do you want to use apache as the reverse proxy, or the backend HTTP server? [10:58] i want my intenal web services to be acess over internet [10:58] Hmm, is mod_proxy full of sexy HTTP/1.1 pipelining goodness? [11:00] twb, has that pastebin given you any clues? [11:01] Hi, i have read that in order to gonfig vncserver you have to edit /etc/sysconfig/vncservers file, i have look for it but that file is not present at my Karmic .... how can i make my second vnc4server desktop to run as a specific user? [11:02] iclebyte-work: answer the question, please [11:04] twb, yea that's the whole file. It's a simple rule I'm trying to implement. [11:05] twb, i've had it working in the other direction i.e. allowing everything except the contents of the allowed domain file by using 'pass !allowed all' [11:05] but I need to block everything EXCEPT the contents of that file and for some reason google is not turning up much, nor is the squidguard documentation.. [11:06] I thought more like "acl whitelist domainlist allowed/domains.txt" and then "http_access allow whitelist" [11:07] acl defines a pattern; http_access applies it [11:10] That's pretty much what I have, except url_regex instead of domainlist [11:18] twb, is that squid.conf your talking about or squidGuard.conf? [11:18] squid.conf [11:18] Which is what I asked you to paste. [11:18] (I don't use squidguard; IMO it doesn't add much value.) [11:25] right i shall try with squid alone then =) [11:29] can anyone tell me how to set up subdomain in ec2 cloud [11:34] ravionrails: EC2/AWS does not provide DNS service. You are responsible for providing this yourself. [11:34] ravionrails: or are you looking for something else? [11:35] erichammond: actually i want to know how to set up in apache configuration [11:37] ravionrails: ah. EC2 does not support multiple IP addresses per instance, but you can support it with standard Apache virtual hosts. [11:39] erichammond: can u point me to the settings in apache configuration file to setup subdomain [11:41] ravionrails: Sure, Google "apache virtual host" [11:41] New bug: #578729 in augeas (main) "Sync augeas 0.7.1-2 (main) from Debian unstable (main)" [Wishlist,New] https://launchpad.net/bugs/578729 === Mike||gone is now known as mikelifeguard [13:11] I have a server here and only a HDMI cable, my monitor supports both HDMI and other type, but I don't have such a cable. Small chance I will be able to install VMWare ESXi on the server without such a cable, right? === mikelifeguard is now known as Mike||gone === Mike||gone is now known as mikelifeguard === mikelifeguard is now known as [[Mike]] [13:47] How can I renumber the ethernet ports? Mine start at eth2 now I have disabled the on board stuff [13:47] a_ok: /etc/udev/rules.d/70-persistent-net [13:47] or something like that [13:47] Jeeves_: thank you [13:49] yw! [13:54] Jeeves_: can I simply remove that file? [13:56] New bug: #578773 in dhcp3 (main) "dhclient waits a random 0-4 seconds before trying to acquire a lease" [Undecided,New] https://launchpad.net/bugs/578773 [13:58] a_ok: Yes. If you remove it, Udev will recreate it the next boot [14:12] hiho === oubiwann is now known as oubiwann_ === oubiwann_ is now known as oubiwann [14:29] I set up a NFS mount based on https://help.ubuntu.com/community/SettingUpNFSHowTo but even with the rw option, I cannot write to the mounted drive - read works fine. [14:30] anything else able to cause it not to be able to be written to? [14:33] tesseracter: have you looked at your log files on your server? [14:35] zul, it was even easier - the folders were set to 755, i just did chmod 777 and it was fixed. === oubiwann is now known as oubiwann_ [15:07] ping! === oubiwann_ is now known as oubiwann [15:26] kirkland, ping? [15:26] RoAkSoAx: pong [15:26] kirkland, where you at? I just fixed the bug I had and I think its ready for merging if you wanna review it [15:26] (testdrive) [15:27] RoAkSoAx: in a session in cocobolo-3 [15:27] kirkland, ok be there in a bit [15:30] do all printers get added to lp group? [15:36] bug #471615 [15:36] Launchpad bug 471615 in lxc "Linux containers not working on EC2" [Undecided,New] https://launchpad.net/bugs/471615 [15:38] how do I install sun java ? [15:38] instead of openjdk [15:39] a_ok: In Lucid, sun's java is in the partner repository. [15:39] !partner [15:39] Canonical's partner repositories provide packages a location for software vendors to publish applications. The repo itself can be added by running this in a !terminal: « sudo add-apt-repository "deb http://archive.canonical.com/ RELEASE partner" » where RELEASE is lucid/karmic/etc [15:40] Pici: is sun-java6-jre the right package? [15:40] a_ok: If you want the runtime environment, you'd need the sun-java6-jdk package for the jdk [15:41] ok [15:48] Hi. I just recently upgraded from 8.10 to 10.04, and now the server wont boot. The server is a machine running on a vmware host. [15:49] A fresh install into a separate guest machine works perfectly. I get a (initramfs) prompt and it complains about not finding the rootfs (which is a LVM2) [15:50] Any ideas to where to start debugging this? If I use the install CD and rescue, I find all partitions of the LVM perfectly intact [15:59] sveinse: Look into bug 360378 and bug 577353. Not sure if the situation is exactly the same. [15:59] Launchpad bug 360378 in linux "Gave up waiting for root device after upgrade then busybox console" [High,Confirmed] https://launchpad.net/bugs/360378 [15:59] Launchpad bug 577353 in linux "System can not find root device with kernel 2.6.32-22" [Undecided,Incomplete] https://launchpad.net/bugs/577353 [16:00] I still have to restest that last bug on the same machine without LVM. reinstall as soon as I have time. [16:01] XDMCP is not working anymore since upgrade to lucid. (configuration unchanged). I can see the kdm on the server, and type login and password, but then I get thrown out. [16:01] Where can I get logs of XDMCP sessions? [16:01] There's only Xorg.0.log [16:07] alvin: Yes, my server misbehaves the same way as the bug reports indicate. Problem now is to find which remedy works best. If course this happened on a production server! [16:07] sveinse: In the second case, I boot the previous kernel. The other bug. well, that one was introduced with karmic. Keep rebooting until it works is my current workaround... [16:08] The other workarounds in the report (kernel options) do not work for me === oubiwann is now known as oubiwann_ [16:09] alvin: No, kernel options is not it for me either, since root=/dev/mapper/xxx. However /etc/fstab uses uuid, so I'm trying to alter that now === oubiwann_ is now known as oubiwann [16:09] sveinse: Yes, do that (see bug 563117) [16:09] Launchpad bug 563117 in ubuntu "Release upgrade converts /dev/mapper entries in /etc/fstab to UUID" [Undecided,Confirmed] https://launchpad.net/bugs/563117 [16:10] This doesn't happen to everyone. Happened to me on some machines. On most, it stayed /dev/mapper [16:10] Strange thing is that the test upgrade machine is now running lucid perfectly in the same settings as this one [16:11] So I cannot say the error is occurring randomly [16:15] Is fstab updated to initramfs? (I.e. do I need to run update-initramfs after updating fstab?) [16:16] dpkg-reconfigure says "you chose not to install grub to any devices - are you sure you want to continue?" - i did not do this - why does it think i did? [16:20] alvin: What does it mean that it asks for password during booting? [16:21] IIRC there is no root pwd set per default [16:23] Hmm. Backup seems to be extremely important right now. Again this being a production server. [16:23] Which live CD can I use which will recognize the LVM partitions from Ubuntu (Intrepid)? [16:49] sveinse: if I remember correctly every (newer) live CD will be able to handle LVM [16:54] It seems reinstall is inevitable for me. Is it at all safe to use Lucid for this? I mean, due to the above bug reports and all [16:57] lunaphyte_: Yes, install grub2 afterwards (it's called grub-pc) [16:58] sveinse: Oh, I didn't log that bug yet. It asked my (non-existing) root password also [16:58] alvin: it's installed. [16:58] this appears to be related to the upgrade from 9.10 to 10.04 i just did. [16:58] lunaphyte_: then run upgrade-from-grub-legacy [16:59] Yes, most of my upgrades told me that I had chosen not to install grub [16:59] (I didn't) [16:59] You can set it right afterwards [16:59] Are this issues related to server upgrades or lucid in total regardless of upgrade/fresh install? [17:00] alvin: thanks. it seems to be fine - i just would like to understand why ubuntu thinks this. [17:01] is there a good reason to upgrade to grub2? i mean, grub1 works on my upgraded systems [17:01] I don't think there is (yet) [17:02] apart from grub2 being a pita to configure [17:03] hey all, I just put up an ubuntu server with samba and whatnot, strange thing is that I lose internet every once in a while and I must reset my router to regain internet, can anyone point in the right direction of how to fix this? [17:03] i like that grub 2 supports lvm. does grub 1? [17:03] i removed bind9... thinking it was DNS messing up [17:03] *bump* Sorry for being persistent: Can I trust lucid on a production server (yet)? [17:03] lunaphyte_: grub 1 supports lvm [17:03] i see. [17:03] if internet works when you reset your router, its the router's fault [17:04] DrNick_ the router is just crappy out and it's not my settings 100%? [17:04] crapping* [17:04] p1und3r: if it works fine after you reset the router, how can it be your settings? [17:04] DrNick_ i don't know, before I implemented the server it never did this [17:05] DrNick_ i didn't think it was my settings but that's why I'm here [17:05] DrNick_ i [17:05] does the server have anything to do with your internet connection? i.e. do the stations on your LAN go through your server as a router, get DHCP addresses from it, or use it as a DNS server? [17:06] DrNick_ no they use a windows server on the network [17:07] it can't be much to do with the linux box then. if you want to convince yourself tho, unplug the linux server for a bit and see if it does it. [17:07] sveinse: I'd love to chat about that (I'm not sure either), but I have to go home now. Bye [17:07] yea it sux cause it only happens once in a blue moon [17:07] alvin: Thanks anyways. [17:07] and the fact that my office is using the shares on the linux server :| [17:08] i very much doubt its anything to do with the server. is the server itself displaying odd behaviour during the problems? have you noticed any other things in common when the problems happen? [17:09] NrNick_ the only symptom is all the internet goes out thats it [17:10] i wouldn't say the problem is the server then [17:10] <[[Mike]]> sveinse: You can trust lucid if you test it and find it to be trustworthy :P [17:11] DrNick_ well thanks for info im gunna look in to it more [17:11] DrNick_ the windows server on the network is a mess.. [17:11] DrNick_ and it does the Ips :D [17:11] that might have something to do with it. [17:12] DrNick_ maybe it doesn't like linux :p [17:12] DrNick_ prior to addin the supplement everything was groovey except for exchange server [17:12] lol [17:13] i admit its odd if the problems actually did start after you bought the linux machine online, but i could just be chance that your router started playing up at the same time. next time it happens i'd pay more attention to whats actually happening on your LAN - use wireshark or something [17:18] so if i add an external drive in fstab will the system error and lock if the drive is not present? [17:20] dominicdinada: no it'll just fail to mount it [17:20] ok :) [17:25] not that i expect anyone will know, but out of interest, anyone got an idea of how close sambe4 is to completion? the wiki seems to have been the same for ages [17:26] er, samba4 * [17:28] DrNick_: last I heard it's still in development - try #samba [17:28] cheers [17:30] would be pretty nice to run a full Windows Domain on linux servers :) [17:32] it's possible, but currently only for winnt4-type domains [17:33] of course [17:33] but i don't know many people who'd want that kind of domain anymore. [17:35] DrNick_: well, it works, mostly [17:35] hi guys i got i little and noob question [17:36] DrNick_: you can probably try and run the alpha [17:36] RoyK: might be interesting to play with on a virtual network or something yeah [17:36] just use a VM [17:36] no need for a vnet [17:36] before install ebox on my server... ubuntu manual says: to access ebox go to https://yourserver/ebox but is does not work [17:37] enav: _before_ installing? [17:37] sorry after install [17:37] you may have to configure apache [17:37] dunno - I don't use webstuff for administration :) [17:38] RoyK: yes thats what i was refering to. a virtual network inside a hypervisor. i can have other VM's as test clients and member servers then. [17:38] royk humm that sound cool... i mean you say console is better [17:38] enav: by far [17:39] thanks mate [17:39] takes a little while to learn it, but when you know it, it is way faster [17:40] enav: "a little while" is something between some weeks to several years, depending on the level of knowledge and how fast you learn [17:41] in my case, it took a year or so of dabling 'part time' so to speak to be able to do most routine stuff. its so powerful though that people who've been using UNIX for 40 odd years are still learning [17:42] i learn fast [17:42] I installed my first linux box in 1994, and I'm still learning [17:42] hey, I have a system running ubuntu-desktop 10.04, but my question is about mdraid. is it a bad idea to create an mddraid array from two block devices (eg /dev/sdb and /dev/sdc) rather than from partitions? [17:42] enav: go for it :) [17:42] it's very powerful [17:42] mikedep333: thats exactly what i do assuming i want the whole disks in the array and i never have problems [17:43] mikedep333: it doesn't really matter [17:43] mikedep333: linux treats all block devices the same [17:43] DrNick_, thanks, now I need to figure out why I can't mount it [17:43] mikedep333: all it means is that instead of creating a partition table on the disks, the RAID metadata is on there instead, which might mean other OS's will recognise them as "new" drives so be careful there, but as far as linux is concerned it won't really matter [17:44] it says /dev/md0p1 already mounted or /media/backups busy [17:44] as opposed to opensolaris that turns off disk cache on partitions/slices by default "just in cse" [17:44] heheh [17:44] I did blkid and stuff, along with mdadm -D, it looks fine [17:45] mikedep333: seems automount mounted it already [17:45] /etc/mtab doesn't show it mounted [17:45] try to umount it [17:45] can't I check the mount point? [17:45] is it reported when you issues the 'mount' command with no arguments? [17:46] "mount" does not show it [17:46] umount /dev/md0p1: not mounted [17:47] i can only assume you have something else mounted under /media/backups [17:47] cat /proc/mounts [17:48] I tried mounting it on /backups, no luck [17:48] note that when I try mount without -t ext3, it says "unknown filesystem linux_raid_member [17:48] add it to /etc/fstab and try to reboot [17:48] it really shouldn't mount unless you have asked it to [17:49] I did [17:49] let me make sure my fstab options are ok [17:49] have you put an fs on it? [17:49] you know what [17:49] previously I had /dev/md0 in fstab, rather than /dev/md0p1 [17:50] let me see if that fixes it [17:50] hang on.... [17:50] have you done makefs on it? i'm assuming you didn't create a partition on it with type "linux RAID"? [17:50] I created it a long time ago [17:50] with ubuntu 9.04 [17:51] partition type? [17:51] I had it mounted before I rebooted [17:51] ext3 [17:51] I just restarted [17:51] it's checking an fs [17:51] that's fs type [17:51] not partition type [17:51] won't say which one [17:51] oh [17:51] fdisk says [17:52] Hello, where i do find the config for the update manager? [17:52] what's the output of 'fdisk -l /dev/md0'? [17:52] it's fsck-ing [17:52] o [17:56] When i do do-release-upgrade and i have hardy lts shouldn't it say me that there is a new lts available? [17:57] that depends on the settings in /etc/update-manager/release-updates [17:57] its set to lts [17:57] then it should, but I'm not sure if ubuntu-server is flagged for updates yet [17:57] DrNick_, the fsck didn't help. ID: 83 System: Linux [17:58] ok ty [17:58] mikedep333: wtf? did fsck finish without warnings? [17:58] RoyK, yes [17:58] which release is this? [17:59] RoyK, at least I think so. Which log does fsck output to? [17:59] 10.04 [17:59] check /var/log/messages [18:01] hope it's not a bug, although I haven't upgraded my 8.04 boxes yet, it'd be nice for 10.04 to be somehow stable [18:01] cat /var/log/messages | grep fsck just says checktime reached [18:01] I think I had this problem with 9.04 [18:01] something to do with a process already accessing the array [18:02] does mount show something? [18:02] and what does /proc/mdstat say about it? [18:02] RoyK, no, mount says nothing, I'll check mdstat [18:03] oh [18:03] I think I see the problem [18:03] md_d0 exists [18:03] and it has a few fewer blocks than md0 [18:04] is it raiding the partitions or something? [18:04] never did that on itself while I was using md [18:04] and that was for fairly large volumes [18:04] what sort of array is this? [18:04] no idea what you've done there... basically when you create your raid device from the block devices, you should end up with /dev/md0 (assuming your first array), then thats the device you can put your filesystem on [18:05] md_d0: inactive md0p1[0](S) [18:05] raid1 [18:05] <[[Mike]]> mikedep333: Did you know that `grep fsck /var/log/messages` is equivalent to `cat /var/log/messages | grep fsck`? I didn't for the longest time :) [18:05] * [[Mike]] 's mind was blown [18:05] lol, thanks [18:06] mikedep333: is /dev/md0p1 the first partition on md0? [18:06] * RoyK never partitions md [18:06] (alternatively a nice idea is to make your raid device a physical volume for LVM, but thats another story) [18:06] md0: active raid1 /dev/sdc[1] /dev/sdb[0] [18:06] RoyK, yes, /dev/md0p1 is the sole partition on md0 [18:07] wouldn't it better to partition the drives and then put a mirror on those partitions? [18:07] s/a mirror/mirrors/ [18:07] DrNick_ said he mirrored drives (block devices) without problems, and that's how linux works [18:08] hmmm, http://ubuntuforums.org/showthread.php?t=1168360 [18:09] oh [18:09] it may be something with the UUIDs [18:09] well, you specify the name of the raid array when you create it [18:09] DrNick_, yeah md0 [18:09]