[00:00] only shows 2011-03-20T23:59:41+0000 [00:00] which ami should i use here: http://uec-images.ubuntu.com/maverick/current/ [00:01] We are using ami-22423c70 on a huge machine quite happily. [00:01] i need a 64 bit ebs, not sure if it really matters which one i choose [00:01] nice [00:02] Client.InvalidAMIID.NotFound: The AMI ID 'ami-22423c70' does not exist [00:04] OK... that's odd, I got that ID straight off my EC2 panel. Ahh.. its in ap-southeast-1b that's probably why, are you using US? [00:04] pnunn: i will use any region, whatever works... i 'm in the US [00:05] OK.. well, that ami is in the southeast zone. [00:05] We are running it in an m2.4xlarge [00:05] ec2-describe-availability-zones on ly shows us-east-1 so i guess i cant put anything there [00:05] nice, so it is possible... hmm [00:06] Change your region to Asia Pacific (Singap.). [00:06] But... there should be workable instances in US regions too. [00:07] i'd prefer the US [00:07] oh southeast = southeast asia, thought you meant southeast US [00:07] When I was setting this one up, I just looked for EC2 instances that were pretty recent. [00:09] just tried the ami i-9ce227f3 from http://uec-images.ubuntu.com/releases/10.04/release/ ... terminated and no console output. [00:09] lemme try from the gui [00:11] pnunn, u using ebs? [00:12] Theravadan: Yep. [00:12] pnunn, how is that going? reddit.com seems to be down because of ebs [00:13] i am just going to run m2.4xlarge and put the db in memory so it doesnt use I/O. [00:14] Theravadan, pretty much what were doing. The reason were on Huge is so we can put our db into a ramdisk. [00:14] pnunn, do any load balancing? Not sure if a need a static id for each of the servers in rotation, hope not [00:14] err static ip [00:16] ok i created the machine via the gui and that worked [00:17] Good stuff... [00:19] pnunn, gonna create a 500gb ebs and mount it, never done this, any tips? [00:20] i'll create it via the gui first, gui seems to be better than cmd line [00:24] Theravadan, I've only ever used the gui to this point.. it seems to work well. [00:24] pnunn, last time i used ec2 was 2 years ago and there was no gui so i'm used to the cmd line but.. gui seems to be easier to use [00:25] Theravadan, yeh, its come a long way. [00:33] anyone have success with the new cluster instances? seems to be kinda beta from the feedback i see via google searches [02:15] downloaded & burned the iso... booted it up & i get "1. \n 2. \n Select CD-ROM Boot Type : _" ... uh, ok? [02:30] pteague: do you have a question? [02:35] was wondering which to do... apparently 2 skips the cd/dvd so must be an odd bios issue even though i've set up the bios to boot correctly off the cd/dvd [04:20] if I create an ami from an existing ebs instance via the gui, it takes a snapshot, so when i create the new instance from the created ami, will it also have a totally new ebs volume? [04:41] hey guys any way to get the numeric permission value of a file? such as 440 or 755? [04:43] sako: stat [04:43] stat -c %a foo [05:07] thanks! [05:16] i want to make a good ubuntu desktop/server anyone build a computer recently can give me some advice? :) [05:17] was looking at either i5 or i7 [05:17] not sure if i _really_ need an i7... [05:21] sako: those are just marketing names [05:21] i5 is marketed at home/desktop, i7 at business/server [05:22] But you should be looking at the actual specs [05:22] yea [05:22] Most businesses won't care about paying an extra $80 for an i7-flavoured system [05:22] Check with your management [05:37] Ugh, aptitude is so slow on embedded hw === michael is now known as Guest54956 [07:27] <\sh> moins [09:22] good morning [09:22] Mooooorning [09:55] Hi folks .. What's the root password for lxc container [09:58] kim0: How did you set it up? [10:02] soren: sudo lxc-create -n lxc1 -f lxc.conf -t natty ? [10:02] lxc.conf is just networking stuff [10:05] kim0: "root", it seems. === _human_blip_ is now known as human_blip [10:16] soren: thanks :) [10:30] kim0: Sure :) === ogra is now known as Guest34611 === Guest34611 is now known as ogra_ [11:17] New bug: #739297 in nmap (main) "Please update nmap: stable upstream version is 5.51" [Undecided,New] https://launchpad.net/bugs/739297 [11:52] New bug: #738968 in samba (main) "Can't See Parts of DFS CIFS Share" [Undecided,New] https://launchpad.net/bugs/738968 === jjohansen is now known as jj-afk [12:05] anyone have problems with tftpd-hpa on lucid after recent kernel upgrade? [12:15] pmatulis, nope, but I haven't used a tftp client yet [12:15] I can't imagine the kernel upgrade could have any effect at all on it === chuck_ is now known as zul [12:33] morning [13:05] cat /proc/sys/fs/file-nr < --why is file descriptors alwayz zero ? is that okay ? [13:07] dunno, kind of hard for me to see that file on your system from over here === dendro-afk is now known as dendrobates [13:08] morning all [13:17] Hello there. I have a general question - I'm running 8.04 on a rented vps. I've asked about doing an upgrade to 10.04 which appears dooable in the community guide, however they say they can only do 9.04, and require a complete recreation of the vm. Could you think of why they would go against doing an inplace update? [13:21] because it's a vps [13:21] and I wouldn't use 9.04, it hasn't been supported for awhile, and will get no security updates [13:21] updates for 9.10 will end next month also [13:21] so options are 8.04 or 10.04 [13:22] <_ruben> or 6.06 ;) [13:22] doubt he wants to downgrade :) [13:22] _ruben, heh i noticed that in the list too (6.06) [13:22] 6.06 runs out in june [13:23] I know it was a joke, but I don't want there to be *any* confusion :) [13:23] patdk-wk, I understand. I didn't understand why I could only be offered 9.04 though... It seems the guide isn't relevant for a vps? [13:23] nothing is relevent for a vps [13:23] o.O [13:23] aw :( [13:24] Sure it is. It just depends whether you're planning on using the OS images provided by your provider or upgrading yourself. [13:24] I would imagine it would be pretty easy to break the vps doing an upgrade [13:24] pici - preferbly I was preferring to do an inplace upgrade if possible [13:25] SuperRoach: I've had no issues upgrading this VPS from 8.04 to 10.10, one release at a time as they came out. Then again, this isn't being used for anything critical, just my personal stuff. [13:25] (linode) [13:25] I only have one vps, and hate it, they are doing 10.04 ontop of centos5 [13:25] The lack of grub made me weary of attempting it for starters though [13:25] kernel is so broken :( [13:25] pici - yeah just personal stuff here too [13:26] You've mentioned that you done one release at a time. I take it you mean 8.04 -> 9.04 -> 10.04 on your vps? [13:26] that is two releases [13:26] 8.04 -> 8.10 -> 9.04 -> 9.10 -> 10.04 .... [13:27] ahh, so not just the lts, got it. [13:27] 9.04 is not lts [13:27] "not just the lts" [13:27] pici - yeah, the .10's ? [13:28] Or maybe it was 8.10 first. Anyway, I started paying for it in late 2008, and as each release has come out I've upgraded. === ogra_ is now known as ogra [13:29] Ok. And you had no problems in your experience with your vps? Did you need to modify your boot parameters with each inplace update [13:29] SuperRoach: No. I didn't touch them. [13:29] There have been some configuration file changes across upgrades, but nothing unexpected. [13:30] That's good to know [13:32] SpamapS: you up yet? [13:32] patdk-wk, you mentioned nothing is relevant for a vps before - is that because of a vendor thing, or because of vps - actual installation difference? [13:33] well, it all depends on the vps [13:33] there are atleast 3 different versions they like to use [13:33] and each one works differently [13:35] I atleast have never seen a ubuntu vps upgrade manual, they are all for bare hardware [13:35] so they aren't really relevent [13:35] I see. Would knowing the vps version help? I'm aware it is Xen. [13:35] are you sure it's xen? generally vps's don't use xen [13:35] normally openvz [13:36] if it's xen, it can be considered a vm [13:36] and upgrades are ok [13:36] Just got an email confirming... yeah I thought it was openvz too... Xen, and thats the reason for 9.04 it seems... they don't support 10.04. [13:36] the only issue you will have is, when using 10.04 or higher, you need to use an ec2 kernel [13:37] and you will loose console output doing that [13:37] Eek. [13:37] I have two I run like that [13:37] don't really need console [13:37] if you don't have console, how do you admin? [13:37] ssh [13:37] and hope it doesn't break :) [13:38] lol [13:39] was there a change that requires the need of ec2 [13:39] ya, xen isn't supported :) [13:39] losing console sounds like one of the scariest things possibnle [13:39] ec2 is based on xen though [13:39] so instead of installed kernel-xen, you install kernel-ec2 [13:40] and if you don't have grub [13:40] you probably don't have control over the kernel or ramdisk images [13:40] and therefor can't upgrade them anyways [13:40] ah, darn. [14:02] while ubuntu server installation I get the following red screen "unable to set password for the mysql root user" , any idea ? [14:17] can someone please test http://transport.nilu.no/products/fukushima and see if it works ok? the news hit .jp/.us and the server wasn't too happy about that... === dendrobates is now known as dendro-afk [14:17] the page loads, are you after testing anything specific? [14:20] hmm, that is an old page [14:26] New bug: #739408 in mysql-dfsg-5.1 (main) "package libmysqlclient16 (not installed) failed to install/upgrade: trying to overwrite '/usr/lib/libmysqlclient.so.16.0.0', which is also in package mysql-cluster-client-5.1 0:7.0.9-1ubuntu7" [Undecided,New] https://launchpad.net/bugs/739408 === nkv_ is now known as nkv [15:10] mysql question: should I have a file somewhere in /etc/mysql/conf.d/ with my local settings, which then will outrank my.cnf, or is it ok to edit my.cnf? [15:14] Caribou: o/ [15:14] Hello pmatulis [15:18] hmm. looks like grub2 will be a good friend of mine... [15:26] smoser, good morning, we need another hardy ec2 from proposed... [15:26] hggdh, remember you don't so much need it [15:26] because you can apt-get install kernel.. and reboot [15:26] if you've launched with pv-grub aki [15:27] smoser, huh. [15:27] * hggdh blushes, and hides [15:28] hi! is it possible to make my samba-server like a domain controller? on my client computers on my lan i want to logon to a domain wich has all users and passwords etc centralized instead of using the local users accounts on the client computers.. is this the right channel for that question? im using ubuntu server 10.10 [15:31] New bug: #739447 in puppet (main) "puppet-common: home directory `/var/lib/puppet' permissions issue" [Undecided,New] https://launchpad.net/bugs/739447 === gtaylor2 is now known as gtaylor_laptop === jj-afk is now known as jjohansen [16:03] hallyn: fyi, spice question on ubuntu-server@ mailing list [16:03] hallyn: figured you'd be best equipped to answer that one [16:04] kirkland: actually kim0 might be as i've packaged it but he's used it :) but thanks i'll take a look [16:04] just replied [16:04] kirkland: this is the week that dev-zero should be updating his packages and taking some of my stuff into his, and then hopefully pushing into universe [16:04] kim0: rock [16:05] zul, can I use cempedak? [16:05] hallyn: can I post another answer that it should be landing in universe soonish :) [16:05] didn't know that [16:07] kim0: i don't have your response yet, but sure [16:08] hggdh: go ahead [16:09] zul, thank you, destroying cempedak now [16:09] nooooo :) [16:10] heh [16:25] Hi all [16:25] How I configure in ssh for some users can't to do login? [16:28] plm: any idea why they can't login ? [16:28] kim0: i thought he was asking for a way to disable SSH for some users [16:28] :s [16:29] ah maybe indeed [16:30] plm: check out AllowUsers and AllowGroups in sshd_config [16:30] or change their shell to something like /bin/false :P [16:34] kim0: thanks [16:36] Japje: that way they can't login locally either :) [17:02] what's the bwhat's the best way to create an AMI from a running instance? I am about to create a bunch of instances, should I make them EBS instances so I can make an AMI from the first one easily and clone the rest?e [17:03] New bug: #739495 in openssh (main) "include in .ssh/config" [Undecided,New] https://launchpad.net/bugs/739495 === dendro-afk is now known as dendrobates === chuck_ is now known as zul [18:00] Theravadan: you might want to consider using configuration management rather than image management to make your systems repeatable. === chuck_ is now known as zul === chuck_ is now known as zul [18:41] got some problem with PAM... [18:41] my errors looks like this: http://pastebin.com/6rAFEBAr [18:44] Orfeous: erm ... that shouldn't happen. Take down the system in single and fsck the root [18:44] s/root/root dev/ [18:47] i cant understand why this has happened [18:47] Orfeous: really, it shouldn't [18:47] maybe there was an upgrade of a package that did this [18:47] doubt it [18:47] using "natty" [18:47] I guess a filesystem corruption [18:48] well, playing with alphas is like gasoline and matches... :) [18:48] what filesystem? [18:48] ext4? [18:49] ext4 yes [18:49] the harddrive of my test ubuntu server is failing, how can I copy the system to a new harddrive? what's the best way? [18:50] RoyK: this pam issues is causing my imap-server not working propertly when i try to login.. :( [18:50] Orfeous: obviously [18:51] Orfeous: but I don't think it's a pam issue - those files really _should_ be there unless they have changed something major on natty [18:51] RoyK: the files isnt in the specified directory.. [18:51] ive checked [18:52] i reboot and i will see :) [18:52] soon back === chuck_ is now known as zul [18:58] the reboot did it :) [18:58] PAM working again [18:58] but the files isnt there anyway :D [18:59] they're there on my natty VM [19:00] amd-64? [19:00] yep [19:01] lemme update to the latest... [19:03] lemme? [19:03] let me === mquin_ is now known as mquin [19:27] Orfeous: heh - just upgraded my natty install and that seems like it broke it [19:29] yes, you see.. [19:30] forgot to make a snapshot before it also - oh well - a test VM isn't much of a loss :P [19:30] is it a bug or just a change? because pam is working without the files [19:30] probably a change [19:30] when i rebooted [19:30] or somewhere in the middle of two changes :P [19:31] hehe [19:31] * RoyK wouldn't use an alpha for anything but testing [19:44] I am trying to use trickle to rate-limit a file transfer over ssh (as in, tar ... | trickle -s -d 5 -u 10 ssh tar ...). without trickle it works great. with trickle the file transfer fails, and all I see is "Read from socket failed: Resource temporarily unavailable". Any ideas where I should be looking? [19:44] this is on ... urgh, hardy, apparently. [19:45] jkg: rsync --bwlimit [19:46] rsync uses ssh for transport by default (since five years back or so) and does a better job than most [19:46] I'm trying to remember why I ditched rsync in the first place [19:46] standard ops (that I use) are rsync -avP --bwlimit /path/to/somedata somehost:/new/path [19:47] also ... you're telling me for *5 years* I've been typing "-e ssh" needlessly? [19:47] oh well :) [19:47] yeah, waste of calories :) [19:47] rsync had some nasty bugs before v3 [19:47] v3 works well [19:48] part of me says "switch to this other tool" is the wrong answer when "why doesn't this tool work" is the question ... however "ditch the tool you've never encountered and go back to one you use daily" is, so I will :-) [19:48] * RoAkSoAx starving... off to lunch === jgjones__ is now known as jgjones [19:48] rsync 3.0.4 is in Hardy's backports, but 2.6.9 is in its normal repos. [19:48] i'm not sure if this is helpful but if you use scp there is a -l switch [19:48] for rate limiting [19:49] Pici: rsync 3.0.7 is in lucid [19:49] i'd never heard of trickle :) [19:49] RoyK: jkg stated he was using hardy. [19:50] k [19:50] ah [19:50] ok [19:50] sorry [19:50] :) [19:50] then I'd use a backport [19:50] yeah, for my sins. legacy machine, and all that. [19:50] rsync Hardy is still supported. [19:51] perhaps they've backported the fixes, then [19:51] rsync | 2.6.9-6ubuntu2 | hardy | source, amd64, i386 VS rsync | 3.0.4-3ubuntu1~hardy1 | hardy-backports | source, amd64, i386 [19:51] ok [19:52] we've moved (almost) all our servers to lucid [19:52] so, backports ... it's been a while, and it was probably on Debian Etch, so, any pointers? :) [19:52] sorry that's a dumb question, I'll STFW. [19:53] one 8.04 box still on - can't get g77 for lucid :P [19:53] jkg: If you're just looking to upgrade rsync from backports, I'd uncomment the line in /etc/apt/sources.list , sudo apt-get update then sudo apt-get install rsync (replace with aptitude if you prefer that). [19:53] jkg: You *may* want to disable the repo after that though, as there were other packages backported that you may not necessarily want. [19:54] ok, thanks [19:54] * RoyK should go pack for the trip to the north - a bit sad to leave the spring now.... :P [19:55] "rsync version 3.0.4 protocol version 30" -- well that was less painful than I remember, and I didn't see the word "pinning" once ;) [19:55] jkg: scp -l should work, and it uses ssh for its transport [19:56] I suppose there are other ways to enable backports safely using pinning, but that is complicated. [19:56] can scp preserve timestamps/permissions though? (the original reason to use tar | ssh tar was to keep those) [19:58] Pici: yeah, I think if the machine was ever going to go back into external-facing production I'd look deeper into the "right" solution, but for now, that worked great [19:58] man page says it can do timestamps and 'modes' with the -p option. I've never used that though [19:58] a7ndrew: any reason to use scp when rsync exists? [19:59] a7ndrew: oh! I was looking for -a, as per cp/rsync. doh :-) but rsync -az will suffice [20:00] jkg: add a --partial (or -P if you want --progress too) so that if you're rsync'ing large files, they can be resumed if interrupted [20:00] well, i could be wrong about this, but i find rsync sometimes spins its wheels a lot before doing anything, and uses a bunch of memory. I guess I only uses rsync for those cases where some of what i want to copy is on the other side. [20:00] typing that out, those seem like sketchy reasons :P [20:01] RoyK: this is for an unattended process run out of cron, so I'll keep it quiet, but thanks for -P (combining those options is what I usually want and don't bother to do) [20:01] a7ndrew: rsync may use rather large amounts of memory (or at least in earlier versions) for the initial index process - no idea if that still is a case for v3 - don't think so [20:02] jkg: for an unattended process, you surely don't want -P :P [20:02] no reason to get emails that size :þ [20:04] indeedy [20:04] given this process is going to run every minute (but look for, create and destroy a lock file, so not really every minute), I *really* don't want those emails [20:04] I used to rsync stuff to and from a ARM-powered NAS box at home, very slow process. In fairness though it was probably the processor struggling with the encryption as much as the memory issue. [20:06] I'm on the mailing list for backuppc and there are regular complaints there about rsync for the memory issue, but that's moving whole disks around [20:11] woo, all working. thanks folks. [20:13] oh, spoke too soon. it failed on one file :-/ [20:13] http://paste.ubuntu.com/583487/ [20:14] (a typical file is /srv/data/documents/queue/test.add/$timestampTIFF001.XSM/blah.tif) [20:16] Hi everyone. I am getting ready to setup a new webserver, Does anyone know if ill have trouble with ubuntu on a poweredge 1650 / raid 5 [20:18] emindmedia: I think you shouldn't. Of course no one can say for sure in advance if you will. [20:19] ScottK: I was just preparing to follow the guide step by step. I figured I would check, thank you. [20:19] emindmedia: it might be worth checking the list on http://www.ubuntu.com/certification/make/Dell/servers -- if you want a 'sure thing', pick from that list. [20:20] that said I wouldn't expect a problem with any kit that wasn't really quite weird :-) [20:21] JKG: Thank you, unfortunately I'm stuck setting this up on servers we own before I can justify the CAPEX to buy certified equipment. [20:22] jkg: Thanks, I'm sure it will go smoothly. Or hopefully. :-) [20:24] * jkg finds and fixes the stupid bug in his rsync invocation [20:28] hrm, no. rsync is failing right at the end of transferring each file, when it comes to rename the file [20:29] (I fixed the path-doubling-up thing, but it still breaks in the same way) === jgjones_ is now known as jgjones [20:37] New bug: #739694 in cloud-init (main) "KeyError: 'handler'" [Undecided,New] https://launchpad.net/bugs/739694 [20:45] ok, throwing away rsync and picking up scp. wonder hwo long until I'm back trying to get trickle working ;) === dendrobates is now known as dendro-afk [20:46] jkg: why did you ditch rsync? [20:47] at the end of transferring any file of significant size, the rename it does was failing. I tried with/without --delayed-updates, got the same errors. === dendro-afk is now known as dendrobates [20:50] (glad I tested it with real files, I had just been creating a directory structure with mkdir/touch!) [20:51] oddly, scp has worked perfectly, except that it failed to set the timestamp on one particular file, claiming it didn't exist === ogra is now known as Guest27445 [21:12] Anyone else notice that Ubuntu EC2 instances take -forever- to restart? Or get stuck and don't restart until you restart them a second time through the EC2 management console? [21:13] gaibdube [21:13] GAHA [21:13] wrong terminal [21:14] uh, nobody saw that, right, I don't need to change any passwords... [21:14] jkg: of course not [21:14] (also: when will GNOME implement focus-follows-eyes?) [21:14] and the logs wil just show ***** [21:15] the blessing is that I only use it in 3 places, for relatively unimportant accounts... [21:15] ... make that used. whew. [21:39] does anyone here use the open-vm-tools? The community documentation doesn't say anything about the kernel modules in open-vm-dkms. Are they actually required on the host for the vm-tools in the guest to work? (i'd assumed so) [21:44] SpamapS: ping [21:47] i want to add swap to my ec2 ebs instance, i'd prefer to put it on local disk, is that an option? using ebs for swap seems wrong [21:47] hallyn: I don't believe they're needed on the host at all, no. [21:52] RoAkSoAx: pong, sup? [21:54] SpamapS: quick question about mysql. I'm doing this: mysql_install_db --datadir=/mnt/export [21:54] SpamapS: that installs a mysql dir under /mnt/export (/mnt/export/mysql) [21:54] soren: interesting, thanks. so they're for the guest? [21:54] SpamapS: now, when running the daemon, should I sue /mnt/export or /mnt/export/mysql as datadir? [21:55] RoAkSoAx: don't sue your system - sorry, couldn't resist :-)) [21:55] RoAkSoAx: /mnt/export [21:56] hallyn: Yes. [21:56] hallyn: They're their paravirt drivers. [21:56] hallyn: ...and such. [21:57] good then i don't need to worry as much about ltp crashing with them installed [21:57] hallyn: There's a special driver for the emulated nic, there's a driver for a host-guest shared fileystem... err... and some toher stuff. [21:57] hallyn: I forget. It's been a number of years by now :) [21:57] one custom-designed to screw up xfs? :) [21:57] soren: thanks. [21:58] SpamapS: I get this though: [21:58] Mar 21 21:57:36 natty1 mysqld_safe: Starting mysqld daemon with databases from /mnt/nfsexport [21:58] hallyn: Sure thing. [21:58] Mar 21 21:57:41 natty1 mysqld_safe: mysqld from pid file /var/run/mysqld/mysql1.pid ended [21:58] RoAkSoAx: looking at /var/log/mysql/mysql.err? [22:10] SpamapS: for some reason the port is still open even though mysql is stopped [22:13] *blink* [22:14] RoAkSoAx: lsof -n -p | grep 3306 [22:14] err [22:15] netstat -tnlp | grep 3306 [22:15] mixed two commands there [22:17] SpamapS: will ping you later. Split brain in my cluster :) [22:30] Is it possible to use a PERC 5 as an HBA instead of as a hardware controller? [22:39] I've got 45 disks in a single 4u, how do I determine the identity of a bad disk from an error in the syslog. [ 8670.445436] ata11.03: exception Emask 0x100 SAct 0x0 SErr 0x0 action 0x6 frozen [22:40] I need to figure out which disk that is in /dev/ so I find identify the right backplane [22:41] Hi .. How do I update the tgz file that lxc clones to create guests? [22:41] for whatever reason this sytem does not boot with the same disk order every time. [22:42] pyrophelia: I'm not sure .. but /dev/disk/by-path/ would probably show reboot consistent device paths [22:42] SpamapS: ok this is what I keep getting: http://pastebin.ubuntu.com/583558/ [22:43] I do but it reports PCI addresses not, ATA :/ [22:43] RoAkSoAx: netstat -tnlp | grep 3307 should show anything listening on 3307 [22:43] SpamapS: doesn't [22:43] s/I do/it does [22:43] RoAkSoAx: whats bind-address set to? [22:44] SpamapS: 127.0.0.1 [22:44] SpamapS:daahh I think that's the problem [22:45] I don't get it. Why does the syslog report errors in ATA addresses but I can't translate that to an actual address. [22:45] SpamapS: apparently not [22:49] SpamapS: i think it is apparmor [22:50] RoAkSoAx: is there anything in dmesg? [22:52] pyrophelia: There's a lot of symlinks in /sys that might help you. [22:53] sbeattie: it is not an error with apparmor, but rather it is enforcing mysql whne I don't really want it to [22:53] sbeattie: at least that's what I'm thinking might be going on [22:55] RoAkSoAx: I'm only picking up the tail of the discussion, and am not sure what you're trying to do, but you can put the mysql profile in non-enforcing mode via aa-complain or aa-disable (the latter is natty only) [22:56] ooo, i didn't know about aa-disable... [22:57] sbeattie: I'm trying to get MySQl in an HA Cluster, and actually launching mysql from another datadir which causes apparmor to not allow MySQl to do so [22:59] well, edit apparmor [22:59] patdk-lap: that';s what I already did [22:59] :) [23:07] soren, nothing uselful. thanks anyway [23:07] ok maybe i'm going about this wrong [23:07] when you have a ton of disks in a single system, how do manage them? [23:07] there has to be something better than syslog errors [23:07] ata11.03 tells me nothing if I can't resolve it [23:44] SpamapS: ok, this is the mysql command I'm using: /usr/bin/mysqld_safe --defaults-file=/etc/mysql/my.cnf --pid-file=/var/run/mysql1.pid --socket=/var/run/mysql1.sock --datadir=/mnt/nfsexport --user=mysql --bind-address=192.168.122.254 and the conf is: http://paste.ubuntu.com/583581/ [23:44] SpamapS: the error it shows is that it cannot bind the port, ideas? [23:45] just wondering, if I put 2 NICs (in the same box) on 2 diff subnets, then have the router with static NAT routers for the "web only" subnet, what other configs do I need to do? [23:46] jeeves_moss: what are you trying to achieve? [23:47] twb, I'm trying to keep ETH0 as our "web interface" so all web traffice coming into/out of this box stays on that one interface,and the gig interface is ONLY for inter.... nm, answered my own question. dual gateways [23:47] RoAkSoAx: does that IP actually exist? [23:47] SpamapS: yes [23:48] RoAkSoAx: strace it and figure out the errno that its catching [23:48] SpamapS: ok ;) [23:48] RoAkSoAx: that really doesn't make any sense [23:49] twb, can you quickly have a look @ the config for me please? http://pastebin.com/nQ1CmK3j [23:53] jeeves_moss: you have two upstream interfaces? That is, two ISPs? [23:53] Hm, except they're both using /24s in 192.168/16, so that doesn't make any sense [23:54] twb, they dual into a single router with dual NATs to the same external interface [23:54] SpamapS: weird, it is like if this bug: http://bugs.mysql.com/bug.php?id=28516 was still present [23:54] twb, with the current config, I can see it externally. but I've only got 2 ports open, so I think I've got it licked. I don't think it's a Ubuntu issue [23:54] I think the problem is you not knowing what you're doing [23:57] twb, that's a distinct possibilty