=== dendro-afk is now known as dendrobates === samferry is now known as ircd === ircd is now known as samferry === samferry is now known as yerrfmas === yerrfmas is now known as clangor [03:12] <[1]kaffien> can someone tell me what would cause this line in the fstab to not mount? UUID=8095a8a2-80f0-4942-9a61-4738f3566209 /media/lsmnas2 xfs 0 1 [03:12] that is just a line for a secondary hdd for storing backups [03:13] i looked up the block id via blkid [04:11] Can anyone assist me in getting indexhibit on my ubuntu server? It has some sort of php semi-auto install file, but I'm not really sure how to use it. [04:35] Hi there [04:36] I've just purchased a Western Digital HDD 2TB, then plug it into my server running Ubuntu server 10.04 (Upgraded from 9.10) but seen it not detected my new HDD [04:37] I have not see it in list of `lspci`, `fdisk -l` or `lshw -C disk` commands [04:37] Please help me! [04:38] Is anyone troubled with this HDD? [04:39] tiger2wander, counting the new drive, how many drives on the system? [04:40] ChmEarl, I have 2 Seagate drivers 320GB running as RAID 1 before [04:41] ChmEarl, may RAID mode prevent kernel to detect new driver? [04:41] my kernel version is: Linux [hostname] 2.6.32-23-server #37-Ubuntu SMP Fri Jun 11 09:11:11 UTC 2010 x86_64 GNU/Linux [04:42] tiger2wander, so now you have three? did you use the same MB header array as the seagates? [04:42] ChmEarl, I have not configured it yet because of I don't see any message from `dmesg` about new hardware detection === MTeck-ricer is now known as MTecknology [04:43] tiger2wander, the seagates are on a separate card or directly to MB? and what is the chipset? [04:43] ChmEarl, I have not a plan to use new driver as RAID array, I'll use it as standalone driver to store data [04:44] tiger2wander, the new drive should be /dev/sdc [04:46] tiger2wander, if detection failed then try to attach the Western Digi to another SATA chipset. Perhaps the chipset used for it is disabled in the bios [04:47] tiger2wander, many MB's have 2 different SATA chipsets [04:47] tiger2wander, my MB has intel ich10 and jmicron [04:49] ChmEarl, there is only /dev/sda and /dev/sdb only [04:50] and RAID array is /dev/md0, /dev/md1 and /dev/md2 === KenjiP0p|away is now known as KenjiP0p [04:56] ChmEarl, my driver specs is here: http://www.newegg.com/Product/Product.aspx?Item=N82E16822136344 [04:56] tiger2wander, you need to go into MB bios and activate SATA or ahci mode on all applicable headers [04:57] ChmEarl, and 2 old drivers specs is here: http://www.newegg.com/product/product.aspx?item=n82e16822148140 [04:57] anybody using postfix with subdomains? [04:57] ChmEarl, both old and new drivers using Serial ATA [04:57] i need to install a new mail server [04:57] tiger2wander, then maybe you will need to run initramfs for your current kernel [04:58] ChmEarl, run `update-initramfs` ? [05:00] tiger2wander, yes after you check your MB bios and turn on the headers [05:03] Incidentally, what is up with some motherboards taking several seconds to load the AHCI controller on every boot? :-/ [05:03] ChmEarl, yep! thank you so much, I'll do that later, because of my time zone is in business hours and can not turn off server now :) [05:04] tiger2wander, run dmidecode and see if all headers show-up [05:05] tiger2wander, no that won't help- didn't help on my system [05:07] ChmEarl, I'm looking for something in output of `dmidecode` now [05:07] tiger2wander, grep dmesg for the chipset names - that works on my system [05:10] tiger2wander, I search dmesg for jmicron and ich10 http://paste.ubuntu.com/462345/ [05:11] ChmEarl, in my case there is no jmicron found in dmesg [05:12] tiger2wander, those are the chips that control my Sata/ahci headers. Your MB might be different [05:17] ChmEarl, yes, and there is my dmesg: http://pastebin.com/CWzcnXP5 === Jordan_U_ is now known as Jordan_U [05:33] tiger2wander, your ahci reports that RAID mode is turned on... no good [05:33] tiger2wander, my ahci reports SATA mode [05:33] ChmEarl, yes I'm using RAID mode for 2 old drivers running as RAID 1 [05:34] tiger2wander, you are using linux software raid (mdx) [05:34] ChmEarl, yes [05:34] tiger2wander, the fakeraid is turned on in the bios [05:35] its turned on but your don't use it [05:36] ah, may I forgot to turn it off in bios seen install new Ubuntu server on that server last year [05:37] before I install Ubuntu on that server, it has run on Debian Lenny and using RAID mode too, may it has use fakeraid from bios [05:38] tiger2wander, when you go to data center you need to carefully look at the AHCI and RAID bios menus [05:39] ChmEarl, to avoid any conflict settings in bios? [05:39] your new drive will showup and most likely your bios wants to use it in RAID and you have to OK this [05:39] tiger2wander, of course thats not what you want, so turn RAID off and use AHCI or SATA [05:40] ChmEarl, ah! understand, btw once time bios is configured in fakeraid or something like that it will always add new driver to RAID array? [05:40] tiger2wander, my guess is that there is no auto config, you have to manually do it === KenjiP0p is now known as KenjiP0p|away [05:41] ChmEarl, yes, I think so [05:43] # [05:43] [ 1.039904] ahci 0000:00:1f.2: AHCI 0001.0100 32 slots 4 ports 3 Gbps 0xf impl RAID mode [05:43] # [05:44] tiger2wander, in that line there are 3 possible modes: AHCI, SATA, and RAID (fakeraid) [05:45] for your new drive AHCI is best [05:45] ChmEarl, Why don't use SATA mode for all? [05:46] ChmEarl, I think if we have difference drivers with difference specs implement so we will need AHCI [05:47] ChmEarl, and if all drivers has SATA compatible then choose SATA, it it right? [05:50] tiger2wander, sure SATA is fine (its legacy).. only RAID is bad since you need to manually be at console for too many things === KenjiP0p|away is now known as KenjiP0p === Jordan_U_ is now known as Jordan_U === EvilTrek is now known as Mithos === Mithos is now known as EvilTrek [08:09] hi there, somebody know where i can get a doc to configurate DNAT with ubuntu server, inlcuding the definition in interfaces === KenjiP0p is now known as KenjiP0p|away === KenjiP0p|away is now known as KenjiP0p [09:03] hi there, sombody know any doc to install DNAT in ubuntu? and how to sett it in the intercaes? === KenjiP0p is now known as KenjiP0p|away === KenjiP0p|away is now known as KenjiP0p [09:38] so, my file server hard locked again after almost exactly the same amount of uptime, with no log information [09:38] I'm just looking for opinions as to whether that's an indicator of a hardware issue or a software issue [09:43] qman__: do you have a local screen/keyboard or serial console connected to it? [09:48] I have a local screen [09:48] Do you get a kernel panic on it? [09:48] but it's completely hard locked, even sysrq is no use [09:48] no, nothing on screen [09:48] OK, that was all [10:21] qman__: Anything in /var/crash? (also, take a look at /etc/default/apport and enable apport) [10:30] I would like to back up the contents of my server to a USB hard disk automatically, what is the easiest way to do it? [10:33] RudyValencia: rsnapshot for example (which is based on rsync) [10:33] RudyValencia: depending on what and how you want to backup your data there are better solutions [10:37] I want to backup both the 80GB / and 250GB /srv partitions on my server. [10:38] hi there, sombody know any doc to install DNAT in ubuntu? and how to sett it in the intercaes? [10:38] RudyValencia: then go with rsnapshot [10:40] Roxyhart0: standard netfilter/iptables thing. see http://www.frozentux.net/iptables-tutorial/iptables-tutorial.html#DNATTARGET for example [10:41] thanks a lot joschi! [10:41] what about cpio? [10:44] cpio is pretty historical [10:44] I'd be using rsnapshot or squashfs [10:44] ah [10:44] The point being they do a better job of incremental backups [10:45] can this 'rsnapshot' thing store to a USB drive? [10:45] RudyValencia: it stores on any filesystem you tell it to... [10:45] rsnapshot just works on a directory, so you'd make a normal filesystem and mount it somewhere [10:45] Although you'd want to use a Unix filesystem, since rsnapshot requires hard links. NTFS would also work, but probably not as well. [10:46] My USB drive is formatted ext3 [10:46] Hi all. I'm sharing sysadmin duties on some boxes with some other sysadmins. Is there a recommended way under ubuntu to set up a version control system for configuration files in /etc ? [10:47] RCS seems the best tool for the job, but it's outdated and am not sure how it would work with ubuntu's sudo infrastructure. [10:47] perhaps soemthing that integrates with bzr [10:47] Kream: etckeeper [10:47] etckeeper ? hmmm [10:47] hands down [10:48] Apart from anything else, it will tell you 1) who installed a package (if they used sudo); and 2) it'll make daily autocommits, so you have 24-hour minimum granularity of "who caused that fuckup?" [10:48] oh yeah baby [10:48] that sounds hot [10:48] etckeeper defaults to git, except on recent versions of Ubuntu where canonical decided they liked bzr better :-/ [10:48] will solve so many issues [10:49] Even though bzr can't handle being run with sudo properly... [10:49] Jordan_U: I wouldn't know; I've never used etckeeper with bzr [10:50] Come to think of it, I haven't used bzr much AT ALL since etckeeper was created... [10:50] " By default etckeeper is configured to use bzr for version control. The repository is automatically initialized (and committed for the first time) during package installation" [10:51] Kream: yes, but if you do something like "apt-get install git etckeeper", it will fail to find bzr and so abort the post-inst init. [10:51] twb: I tried it, and using sudo with etckeeper with bazaar leaves you with root owned files in ~/.bazaar/. So it's fine for all of those Ubuntu users that log in as root instead of using sudo... [10:52] uh.... ok now i'm confused. I've not used any vcs system apart from the most basic SVN co stuff. [10:53] Jordan_U: that's a bug; report it. [10:53] Jordan_U: in fact, it's fixed in 0.47 [10:54] Lucid only has 0.41, which is a pain. [10:54] I have 2 people other than me who will be logging in as their usernames and using sudo to do admin tasks. [10:54] no one gets to log in as root. [10:54] twb: IMHO it should have been fixed before bzr was made the default for etckeeper, which pretty much requires you to use sudo. [10:54] few people get to log in at all. [10:54] which should I use? default bzr ? [10:55] or git or something else? [10:55] Kream: I'd personally go with git, but that's what I already use for my own projects. [10:56] Kream: whichever VCS you know best and have best support for [10:56] what are the drawbacks to using bzr? [10:56] i know none :) [10:57] * Kream is a clean slate, VCS-wise [10:57] The downside of bzr is that only Canonical and their fanboys use it [10:57] And GNU projects. [10:57] Ha! [10:57] That decision was a fucking crock [10:58] "We'll use bzr because, technically, it has GNU in its name" [10:58] Kream: lots of people use git, so even though its UI is confusing and counter-intuitive, you can at least get help. [10:59] Kream: And I personally don't find git's UI confusing and counter-intuitive. [11:00] It reminds me of gnus [11:00] Kream: Also git is *fast*, which means that etckeeper won't be slowing down upgrades as much if you use git. [11:00] Jordan_U: slowness isn't really an issue unless you're doing upgrades one package at a time [11:00] etckeeper invokes the VCS twice -- once pre-apt, once post-apt [11:01] twb: I like to follow development releases :) [11:01] Jordan_U: well, even so, I only apply updates once a day... [11:02] However on my git ones, I did find that etckeeper was conusuming 8MB for every 1MB of working tree in /etc -- I think joeyh fixed that by tweaking the git auto-gc settings [11:03] http://bugs.debian.org/483804 hmm, maybe not [11:04] eh, system run at 0 load most of the time so i'm not too worried about raw speed. [11:05] runs* [11:05] a bug in ubuntu-bug, classy [11:05] consuming? as in, diskspace? [11:06] twb: ? [11:08] Kream: read the ticket [11:09] oh ok [11:15] soren: downsides for bzr is that it's way slower than git, right? [11:16] sorry, hit tab after "lol" [11:16] so [11:16] apart from slowness is there any other downside to using bzr ? [11:18] Kream: In general or specifically for etckeeper? === KenjiP0p is now known as KenjiP0p|away === KenjiP0p|away is now known as KenjiP0p [11:29] etckeeper [11:51] Kream: bzr in lucid still has the sudo bug, which means that you can't use etckeeper with sudo without breaking your user's bzr configuration. [12:03] 17814 clint 20 0 346m 157m 1984 D 5.0 42.8 0:27.98 cc1plus [12:03] ahh c++ .. such an awesome language. :) === mathemonkey is now known as darkk^ === freeflyi1g is now known as freeflying [13:05] alvin, I don't have a /var/crash or /etc/default/apport [13:10] qman__: I have apport on every system. I thought it was part of the base install, but can't find out why (apport-symptoms Recommends apport and apport Recommends apport-symptoms) [13:11] this was an 8.04 install upgraded straight to 10.04 [13:11] after some googling I installed it, it may help [13:11] if I get nothing at all I guess that means hardware [13:12] qman__: Probably, but it's a lot of coincidence if it happens directly after an upgrade. Let us know what you find out. [13:12] will do [13:56] hi [14:00] hi === Bhootam is now known as Kream [15:04] Daviey: can you look at bug #604400? [15:04] Launchpad bug 604400 in eucalyptus "Confusing comment in /etc/eucalyptus/eucalyptus.conf" [Undecided,New] https://launchpad.net/bugs/604400 === dendrobates is now known as dendro-afk === dendro-afk is now known as dendrobates [15:09] dsa or rsa? [15:09] rsa. [15:09] zul: on it [15:10] merci [15:10] thought so couldn't remember [15:11] thanks [15:57] hi, is it possible to upgrade a KVM virtual machine from hardy to lucid with a normal do-release-upgrade? [15:57] (on a hardy host) [16:08] \quit === dendrobates is now known as dendro-afk [16:17] howdy folks , I've setup a media server in our TV room , and I would like to control i from my laptop like a remote . i don't need X or remote screen to run on the laptop. I would just like to be able to access media files and play them on the server which is hdmi connected to the tv . [16:30] <_ruben> BluesKaj: xbmc comes to mind [16:36] Hi! I have a problem with configuring sendmail to use gmail, after restart I get: unknown configuration line "define(`SMART_HOST',`smtp.gmail.com')dnl" and about 8 more such errors (pointing to newly added lines). What's wrong? [16:50] WG1337, the first problem is that you're using sendmail [16:50] it's ancient and a total nightmare to configure [16:51] use postfix or exim instead, as those are the supported mail servers on ubuntu [16:52] Do I need to uninstall sendmail? [16:52] The thing is that I need to make PHP work with gmail, but most scripts use /usr/bin/sendmail [16:52] I would use the purge option, otherwise you're going to have a lot of leftover configuration [16:53] other mail servers create a '/usr/bin/sendmail' for that reason [16:53] it doesn't mean you have to actually use sendmail [16:54] there are a lot of guides for configuring postfix to use a gmail account, and when I did it myself it wasn't that hard [16:54] Hi I need Help with Setting up mail server in ubntu 10.04 [16:58] Room is too silent [16:58] Can any body please help me with Mail Server in ubuntu 10.04 [16:58] I really need it before next 2 hours [17:03] rojoloco47: I don't know much about it, but I would recommend following the steps in the server guide for the components of a mail server that you may need. https://help.ubuntu.com/10.04/serverguide/C/email-services.html [17:04] They tend to be good recipes that work in most cases. [17:04] I am just coming back from this page, This did not helped much :( @ thrain][ [17:05] !ask | rojoloco47 [17:05] rojoloco47: Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [17:05] we can't very well help you if you don't ask about your specific problem [17:06] Ok I want to install a mailserver ,just a simple mailserver with postfix, with which my clients can use Microsoft outlook in windows, [17:07] in other words i need a mailserver which provide smtp address and pop3 address so that my clients can recieve and send mail over internet [17:07] that's quite simple, and I believe is covered in the server guide at the link above [17:07] simply install the mail server and it comes configured that way [17:08] I tried but what I am not getting is, when I install postfix then what to do next? I mean what are my smtp and pop3 addresses for what I configure my mail client softwares ? [17:08] as for receiving mail, that requires you to set up DNS properly for your domain [17:08] you set your email clients to however it is they connect to that server [17:08] if it's on the same local network and you don't have DNS, use the IP [17:08] Yes that everything I can do, what I can't do is simply a smtp and pop3 address to configure clients [17:09] you simply use the name or IP of the server [17:09] lets suppose I have configured postfix using ubuntu official guide for a domain secure.sajidsaif.com now what are my smtp and pop3 address to configure mail client softwares ? [17:09] has anybody installed 10.04 in EC2? [17:09] _ruben, xbmc , ok I'll check it out ,thanks. [17:10] the address would be [user]@secure.sajidsaif.com [17:12] I am sorry if I am not able to make you understand with the situation, like if we configure gmail with some email client , in incomming mail server we put address as pop.gmail.com and outgoing server we say its smtp.gmail,com, in the same way, what addresses will I have to use for incomming and outgoing mailservers ? [17:12] the name or IP of your server [17:12] if you have local DNS, use the name [17:12] otherwise, it's the IP [17:13] IP ? can you make it more clear, [17:13] I will use it same like secure.sajidsaif.com ? [17:13] IP address, whatever you configured [17:13] yes, if that name resolves to the server [17:13] IP address will be DNS , is not it? [17:14] if DNS works, that name will work [17:14] both incomming and outgoing mailserver address will be secure.sajidsaif.com right ? [17:14] yes, if both postfix and dovecot are on the same server (which they are if you followed the guide) [17:15] you may need to enable pop3, as I think only imap is enabled by default [17:16] Ok tell me once again, What I have to do, is simply install dovecot and postfix server nothing else from that guide ? [17:16] it is actually even simpler than that [17:16] sudo tasksel [17:16] and choose mail server [17:17] when asked, choose internet site [17:17] and put in the requested information [17:19] I have put this command and it gave me a list of things [17:19] where mailserver is already selected [17:19] then it is already installed [17:22] anyone here have some experience setting up autofs? [17:22] let me finish guide with dovecot and postfix again, I have uninstalled everything assuming that its not working but let me check that i will update you [17:22] i'm trying to set it up to allow ejecting carts from a powervault RD1000 [17:30] thanks qman__, with exim4 it sends mail :) also I found out that my ISP given mailbox sucks! [17:56] is there a way to extract a seed file from an installed server? [17:58] How do I convert the Desktop into Server? [17:58] sorry, rtfm debconf-get-selection [18:46] New bug: #604717 in ntp (main) "Please convert init script to upstart" [Undecided,New] https://launchpad.net/bugs/604717 [18:48] Hey guys! I cant do: /etc/init.d/samba start <- it says it cant find it [18:48] samba? [18:48] ofc in root@server [18:48] the server tool thingie? :) [18:49] called smbd on my system :) [18:49] using lucid? [18:49] oh, ill try that, yeah I am [18:49] service smbd restart [18:50] oh... its already start [18:50] started* [18:50] anyhow, ive tried putty to connect it, but it wont [18:51] putty connects to samba? [18:51] it wont [18:51] I should hope not [18:51] samba != ssh or telnet [18:51] what to use then? [18:51] anything that supports cifs [18:52] windows, smbclient [18:52] smbmount [18:52] sec [18:52] erhhh? [18:52] dunno what that is [18:52] please enligthen me :) [18:52] those are all samba/cifs clients [18:52] prefering any of those? [18:52] depends on what os your using [18:52] PuTTY is an SSH and telnet client, developed originally by Simon Tatham for the Windows platform. PuTTY is open source software that is available with source code and is developed and supported by a group of volunteers. ? [18:52] on the client/test machine [18:53] ya, putty has nothing at all to do with samba [18:53] to connect to the server? [18:53] putty can connect to the server, using ssh or telnet [18:53] still nothing to do with samba [18:53] dunno why you talk about samba, and say putty won't connect [18:54] I talked about samba, cause it wouldnt start, but ive obvious used the wrong command to start it. anyhow, putty wont connect to my server [18:54] that is what I meant [18:56] could you help me with that? [18:56] probably not [18:56] if ssh won't work, you don't have ssh installed [18:56] or something wrong with your network [18:56] :o [18:57] that wasnt much [18:57] openssh-server is already installed [18:58] it seems [19:04] patdk-wk? :) [19:06] hmm? [19:06] !glassfish [19:06] I'm happy, just got nx working, dunno why, dunno how [19:07] hey everyone, im trying to figure out how to run a java JAR file on ubuntu server, its a gui program, and i can get it to run when i ssh -x using x11 but when i log out it kills the process. could glassfish fix this problem? [19:08] openssh-server is already installed [19:15] patdk-wk - openssh-server is already installed - so that shouldnt be the problem... my server's "ip" is set to 83.92.69.50 [19:16] putty says connection timed out when I try to connect to the server [19:16] can it be because both the server and the client is connected to the same switch, which is connected to my actual router? [19:16] dunno, depends on how you setup your router, server, client [19:16] I really don't want to go over network engineering [19:17] Znow are you behind a NAT router? [19:19] T3CHKOMMIE - dont think so [19:19] T3CHKOMMIE - aint sure [19:19] Znow, tell me how your network is set up in a net shell, i had some problems with ssh at first im sure i could walk you through it if i know how youre set up. [19:20] T3CHKOMMIE - can we go private then? [19:20] sure [19:46] New bug: #604744 in nmap (main) "ncat when connecting trough proxy only sends lines in pairs " [Undecided,New] https://launchpad.net/bugs/604744 [19:51] jjohansen, smoser: Have there been any thoughts about or investigation into bug 575193? I have a client who is also experiencing high reported load on an idle EC2 Lucid instance. [19:51] Launchpad bug 575193 in linux-meta-ec2 "high reported load average on idle EC2 instances (dup-of: 574910)" [Undecided,New] https://launchpad.net/bugs/575193 [19:51] Launchpad bug 574910 in linux-ec2 "High load averages on Lucid while idling" [Undecided,In progress] https://launchpad.net/bugs/574910 [19:53] Thanks uvirtbot :) I didn't even notice that bug was a duplicate. There has been a fair amount of activity on the main bug. [19:55] can you imagine how annoying an audio uvirtbot would be? [19:55] "Hey bob, I was looking at bug 557739[uvirtbot starts talking.. launchpad.. bug..] [19:55] Launchpad bug 557739 in jockey "jockey-gtk crashed with BackendCrashError in convert_dbus_exceptions() (dup-of: 413624)" [Undecided,New] https://launchpad.net/bugs/557739 [19:55] Launchpad bug 413624 in jockey "jockey-gtk crashed with BackendCrashError in convert_dbus_exceptions()" [Undecided,Incomplete] https://launchpad.net/bugs/413624 [19:55] erichammond, i've asked jjohansen before. he's pretty convinced that its just accounting that is wrong. [19:56] but i dont think he's gotten to do much on that. [19:56] smoser: Makes sense to me . There doesn't seem to be anything going on with the boxes. [19:56] is it possible Xen is over-counting jiffies if something else steals its CPU cycles? [19:57] I believe I've seen this type of behavior before in the last couple decades, but I can't remember which Unix/Linux system it was on. [19:57] like, if it spent 100 cycles doing nothing, but it didn't get to count them as doing nothing because the CPU was stolen.. then load would look high [19:58] I've learned to take load average as just an indicator that I should investigate to see if something really is wrong. [19:58] yeah, load is useless w/o context. :) [19:58] SpamapS, yeah, thats generally what we're thinking. is it is representing load more correctly as the load of a full cpu. [19:58] and its not getting a full cpu [19:58] yeah, i've seen kernel patches addressing things like this on power also. [19:59] load seems high in the guest because the guest was getting 0.1 of a cpu [19:59] erichammond, are there things reading load other than humans ? [19:59] ie, is this going to cause some tool to say "oh no, my servers are more loaded than usual, i need to fire up more instances" [20:00] its why response time always trumps artificial gauges and even counters.. if it takes a long time, something is wrong. [20:00] smoser: In the cases that I'm aware of I don't think it is causing other problems. [20:01] ...but it would be good to fix as it is alarming folks upgrading to Lucid on EC2. [20:01] yeah, or at least clearly document "seriously, don't trust that!" [20:01] smoser: Nobody reads the documentation. [20:01] as if its designed in behavior that isn't going to change, changing it is ongoing maintainence that we dont want to incur [20:02] and it man not be something that is easily "fixable" [20:03] I remember Theo Schlossnagle always berrating us for using mod_backhand's ability to read load average for traffic balancing.. he was like "no its useless use busy apache processes" .. ahh.. autoscaling w/ broadcast on a local LAN.. the good old days [20:11] smoser: ill do xinetd if you want to do apache :) [20:11] nah. thanks for the offer though. [20:11] smoser: grumble === jorge is now known as jcastro [20:16] how do I add a timestamp to a log being created by crontab? [20:22] spartan07_: use bitcron [20:24] Jeeves_, never heard of it. is it installed on ubuntu server? [20:25] spartan07_: Nope. But it is on launchpad, just a sec [20:26] is it bicron? [20:26] https://launchpad.net/~f-launchpad-bit-nl/+archive/bitcron [20:27] It enables you to email you if something is broken [20:27] instead of 'when there is output' [20:27] awesome, thanks! [20:27] It also enables you to keep logfiles of your cronscripts [20:27] I'm setting up a KVM host cluster out of 10 blades. I've got a 14 disk array RAID-5'd attached to a separate server, and exported that as an iSCSI target. My question is can I mount that from all the blade KVM hosts without corrupting it? [20:28] spartan07_: You're welcome [20:28] matt_keys: As long as you're not writing to the same lun with all the hosts, i think you can [20:30] Jeeves_ : that's just the thing--it's all one big LUN. From what I'm reading I can't use ext3 because it isn't "cluster aware". [20:30] matt_keys: Indeed. [20:30] Why not create a lun per vm? [20:30] Or export it using nfs, and create an image per vm [20:30] Would rather have one big storage pool for all the hosts to use [20:31] So, use nfs? [20:31] RHEL's best practices show NFS to be too slow after about 5 guests start using it [20:31] what about cLVM? [20:31] matt_keys: So what do they suggest? [20:31] well RHEL uses the GFS/Cluster Suite [20:34] I guess NFS is going to be the ticket probably [20:43] matt_keys: I use opensolaris/zfs for that sort of stuff === unreal_ is now known as unreal [20:51] RoyK: That's the better option, yes :) [20:52] I use linux for most stuff, but for storage, opensolaris (or nexenta) beats most of the other options [20:54] would this "command > outout.log 2 > error.log" work to dump the programs output to output.log and any errors to error.log? [20:55] CppIsWeird: which shell? [20:55] umm, not quite sure how to answer that. w/e the default shell is when you ssh into a ubuntu-server. [20:55] ps $$ [20:56] that'll be bash [20:56] command > out.log 2> err.log [20:56] without a space between 2 and > [20:56] ok. thanks. :-) [20:57] np :) [20:58] man bash for more - but that manpage is rahter gross (4k lines or so) [21:02] cLVM or just using local storage might be the better option... then just use NFS for migrations and ISO storage [21:03] blades have two scsi hotswaps on them, all are raid1 [21:04] Anyone know the kernel parameter to limit your RAM? I have 48G in a box, and I want it to think it has 1G. [21:06] matt_keys: Why limit yourselve? [21:07] matt_keys: IMHO it's smarter to create a lun per box, or search for other fs'es that can do what you want [21:07] bpgoldsb: Hmm [21:08] bpgoldsb: max_addr=[KMG] [KNL,BOOT,ia64] All physical memory greater than or [21:08] equal to this physical address is ignored. [21:08] bpgoldsb: http://www.cyberciti.biz/howto/question/static/linux-kernel-parameters.php [21:09] Jeeves_: mem=1G solved it. [21:09] bpgoldsb: bingo - but why limit? [21:09] * RoyK was about to answer bpgoldsb [21:09] Benchmarking disk i/o, need to overcome ram buffer, don't want to write 50G files. [21:10] ic [21:10] benchmarking i/o or filesystems? [21:10] i/o [21:10] k [21:10] I've got a box with a RAID-6 of 8 15K SAS disks, and a Corsair SSD. [21:11] writing with O_DIRECT will bypass the buffers, btw [21:11] Testing performance of the native host, KVM linux guests using virtio, and KVM Windows guests using whatever qemu gives it for HW [21:11] orly? That might be more useful. If iozone supports it. [21:11] bpgoldsb: erm - have you considered using zfs? with osol or fbsd? [21:12] RoyK: Not really. The FS doesn't seem like a big issue to me. What would I be gaining? [21:12] bpgoldsb: bonnie will let you tell it how much ram to use [21:12] matt_keys: doesn't matter if the OS is buffering, does it? [21:12] uses direct calls if I remember right [21:12] bpgoldsb: read this http://hub.opensolaris.org/bin/download/Community+Group+zfs/docs/zfslast.pdf [21:12] bpgoldsb: zfs rocks [21:13]