[00:32] <brunost> hi! I'm working on setting up a little home server (my first server I may add) and I am trying to figure out what would be the best way to deal with a drive to run the OS on
[00:32] <gdeeble> You only have 1 drive?
[00:32] <gdeeble> or multiple?
[00:33] <brunost> gdeeble: I'll have a raidz of 4 drives for storage
[00:33] <brunost> I am trying to figure out what solition to go for for storing the OS
[00:33] <brunost> optimally a solution like freenas where it loads an image from a pendrive to ram would be the best
[00:34] <brunost> but I don't really have enough knowledge about this myself to figure out if this is a clever solution
[00:34] <gdeeble> brunost: See I have my os on 1 drive then mounted my raid to a mount point.
[00:35] <ironm> duli, I was thinking also about such concepts (just using xen ... but I didn't get the xen live image working as wanted wanted and had to stop due to missing time resources)
 ironm: i'd like to use ubuntu on the flash drive as a server solely for manage virtual machines. The real storage will be built on 3 HDs using ZFS
[00:35] <ironm> but ubuntu-server is definitively the best base for kvm hosts
[00:35] <brunost> ofc I'll keep the OS on a separate drive, just not sure if I should try out loading from usb to ram or use a 2.5" drive of some sort
[00:35] <duli> ironm: i think so algo.
[00:36] <duli> ironm: but I've given up the idea of using ubuntu-server on a usb stick
[00:36] <duli> would require a lot of tweaks to reduce disk writes
[00:36] <gdeeble> brunost: depends on your needs. I have a little hard drive running my os, but USB solutions are def. nice as long as everything wants to play ball.
[00:36] <duli> I'll go with a simple notebook hd (5400 rpm) to t install the sustem
[00:37] <brunost> gdeeble: thats the thing, will I get it to play nicely
[00:37] <ironm> duli, I gave such ideas too ;) ... I run ubuntu-servers as kvm hosts on raid5 or raid6 (with 4 or 8 drives) .. and use additional raid5 or raid6 as storage for VMs
[00:38] <ironm> duli, well .. when you run in live you don't need really to care if your KVM host have enough RAM (ECC)
[00:39] <ironm> gdeeble, I use own debian live images (booting from a USB stick) for my daily work .. it works great :)
[00:39] <duli> ironm: sure, but my concern would be the usb key falling appart in a short period of time
[00:39] <duli> ironm: since you mentioned kvm
[00:39] <ironm> duli, yes .. that can happen
[00:39] <gdeeble> brunost: I never had success myself but I am still a novice at this. I ended up using Zentyal which uses ubuntu with a web management face.
[00:39] <ironm> gdeeble,  eg. ... http://rsync.it-infrastrukturen.org/postgresql/live-wheezy-amd64-awesome-public-databases-20121024e.iso
[00:40] <duli> ironm: is it possible to open a vm machine remotely, using the virt-manager?
[00:40] <ironm> duli, yes
[00:40] <ironm> duli, like: ironm@wheezy:~$ virt-manager -c qemu+ssh://192.168.1.75/system
[00:40] <duli> ironm: ah, that's so great!
[00:41] <duli> ironm: so I can perfom the os install on the vm remotely, right?
[00:41] <gdeeble> ironm: I always ran into something but I think my problem was lack of patience and knowledge when I was trying.
[00:42] <ironm> duli, 192.168.1.75 is a ubuntu-server KVM host (HP DL385g7 ,,, with 32 or 64GB ECC RAM and 2 pieces of P410 RAID controller with 1GB cache each ... and 16 SCSI drives)
[00:42] <gdeeble> but that was a few years ago.
[00:43] <ironm> duli, you have to put .iso images for the installation inside /var/lib/libvirt/images/ an the KVM host, but the installation of VMs happen mostly from remote client (with the command like above)
[00:43] <duli> ironm: ah, ok, that was my doubt
[00:43] <duli> well, fair enough
[00:43] <duli> seems nice
[00:44] <ironm> virt-manager doesn't offer you the possbility to use local .iso images on the virt-manager client for installation on the KVM host
[00:47] <ironm> duli, you need "virtual machine host" tasksel option of ubuntu-server to install the KVM host. You need also the ssh-server tasksel option for virt-manager (remote)
[00:47] <duli> ironm: ok, that's totally fine. I could always do a scp to the kvm host
[00:47] <duli> ironm: ok,. I just complet4ed the install and selected those to options
[00:47] <duli> ironm: I already have ssh access to it
[00:48] <duli> ironm: do you recommend ant specific tutorial on kvm/ubuntu?
[00:49] <ironm> gdeeble, now a day linux live images are very comfortable
[00:50] <ironm> duli, eg. this one : Red_Hat_Enterprise_Linux-6-Virtualization_Host_Configuration_and_Guest_Installation_Guide-en-US.pdf
[00:50] <ironm> I can't find the link
[00:50] <duli> ironm: no prob., I'll go through the redhat manuals
[00:51] <duli> tks a lot
[00:51] <ironm> duli, give me few minutes . I have some links also for ubuntu
[00:51] <duli> ok
[00:52] <ironm> duli, may I pc you? (query)
[00:52] <duli> sure
[02:51] <LeChacal> hello, i am trying to setup the 2nd nic in my server for a local connection to another machine through a swtich using static IP and it isn't working, this is my interface file http://paste.ubuntu.com/1349628/ am I missing something in my interface file or something else i needed to do. If I reboot the interface doesn't automatically come up and if I bring it up manual it doesn't ever list as having the address set in the interface file.
[02:54] <qman__> two things
[02:55] <qman__> you can't have more than one default gateway without a lot of extra config, or it will just choose at random which interface to send traffic out of
[02:55] <qman__> and you're missing auto eth1
[02:55] <qman__> and you can't have IPs from the same subnet on different interfaces
[02:55] <qman__> I lied, three things
[02:56] <LeChacal> ok so do i just not set a gateway on the 2nd nic then?  And I have differnt subnets DHCP gives out 192.168.1.x and the static is 192.168.0.141.
[02:56] <qman__> yes
[02:56] <qman__> you only set a gateway on your internet-facing side
[02:57] <qman__> if you have more than one internet-facing address, you have to do some extra stuff to get it to either prefer one or load balance
[02:57] <qman__> dhcp sets a gateway automatically, so just remove the manual one
[02:59] <LeChacal> ok i have made the changes and rebooting, ill see if that was all my problems shortly
[03:01] <LeChacal> qman_ thank you that seems to fix my problem
[03:09] <uvirtbot`> New bug: #1070322 in quantum (main) "again creation of quantum db" [Undecided,New] https://launchpad.net/bugs/1070322
[07:35] <star314> Hi! How can I (temporary) disable a software raid device, e.g., /dev/md2 from the system such that it will not re-added after a system restart?
[07:36] <star314> I've tried mdadm --stop /dev/md2 and removing it from /dev/mdadm/mdadm.conf
[07:36] <star314> but this didn't work.
[14:12] <bananapie> Is there a command in ubuntu to which I can pipe an email and an IP address and it would send an abuse report to the abuse email for the given IP ?
[14:14] <bananapie> I got hits attacks from 200 different IPs in a 24 hour period, I could never work on this list by hand.
[14:20] <patdk-lap> bananapie, only 200?
[14:20] <bananapie> that's if I count each IP only once.
[14:25] <dassouki> I'm looking for a scheduling service for my employee similar to ms exchange
[14:25] <dassouki> we use google mail
[14:32] <queency> hello all: can someone tells me where is the logs of start-stop-deamon ?
[14:43] <bananapie> queency, what is it you are looking for ?
[14:43] <bananapie> some daemons send information into /var/log/syslog when starting or stopping, but not all do.
[14:44] <queency> My main task is fly away when starting with start-stop-deamon and i don't know why !
[14:45] <bananapie> what do you mean fly away ?
[14:46] <queency> die
[14:48] <queency> btw how can i replay like you did with the starting of "queency," ?
[17:09] <mrrothhcloud__> anyone here used Transcenders?
[17:21] <gucki> does it make any sense to have "R/W multiple sector transfer" enabled for sata disks? i see it's set to 0 for all my hdds...
[17:30] <patdk-lap> that is extreemly odd
[17:31] <patdk-lap> generally you want it to be atleast 8
[18:07] <jacobw> i have a problem using --location with virt-install, the files are downloaded to /var/lib/libvirt/boot/ but disappear immediately after they are downloaded and seabios in the VM hangs on booting from ROM
[21:42] <NomadJim> anyone running an ubuntu webserver that can share some anecdotal evidence of how often they have to patch their server?
[21:46] <ikonia> you patch your servers when ubuntu provides package updates
[21:46] <ikonia> it's that simple
[21:47] <DaveR> iirc there is a setting to have your sever automatically deploy security related patches
[21:49] <ikonia> yes, but that's not wise
[21:49] <ikonia> as you don't want to blindly patch unless you have a test system first
[21:50] <DaveR> that's fair
[21:56] <greppy> NomadJim: I use apticron to get emailed notifications of pending updates.
[21:57] <NomadJim> ikonia:  i'm looking for number of times/year or some other kind of statistic
[21:58] <ikonia> NomadJim: it changes from release to release and version to version, you can't guess how many bugs or security alerts will be found
[21:58] <NomadJim> because it's going to be different based on packages, i'd like to focus on people with web servers
[21:58] <ikonia> NomadJim: it will differ for different people depending on what packages they are using
[21:59] <NomadJim> yup
[22:03] <NomadJim> maybe i'll compile something if I can't find it on google
[22:14] <DaveR>  NomadJim, I make a point on my box to check weekly.  Having said that though, I will say this: my box is for personal use and therfore if it breaks/gets owned its not that big of a deal.
[22:14] <DaveR> ie: I'm not saying that weekly updates are the best or that it's how you should do it
[22:15] <greppy> I still like apticron as a way to get notification of new package updates.
[22:15] <NomadJim> yeah i think automatic notification like greppy said would be best when i'm actively administering a production server
[22:15] <greppy> that way you don't have to remember to login and check, if there is an update, you get an email.
[22:16] <NomadJim> i have a whole bunch of personal use servers
[22:16] <NomadJim> but i usually outsource it to a managed provider if it gets important
[22:17] <NomadJim> so I am not very good at best practices
[22:25] <NomadJim> greppy:  might be kind of cool if someone combined apticron + uptime monitoring
[22:26] <NomadJim> as a service for lazy people
[22:26] <ikonia> or you could just not be lazy
[22:26] <ikonia> and run the server properly
[22:27] <NomadJim> ikonia:  i'm talking about a business idea as opposed to personal administration
[22:27] <ikonia> doesn't seem a good idea to me personally
[22:27] <NomadJim> i bet someone is already doing it though
[22:27] <NomadJim> what's the downside of having apticron in addition to the normal uptime monitoring most services provide
[22:28] <ikonia> that's not what it was meant to do
[22:28] <ikonia> so you're just adding a pointless function to it
[22:28] <ikonia> there are tools for monitoring - this isn't one of them, so adding one random monitoring measure to it doesn't seem good
[22:29] <NomadJim> I'm talking about uptime monitoring services like https://www.pingdom.com/
[22:29] <ikonia> yeah, I'm saying I think it's a bad idea
[22:29] <NomadJim> I think it would be valuable if they added apticron notifications as well to their offering
[22:29] <ikonia> that's not what the tool was meant for
[22:29] <NomadJim> notifying of updates isn't what apticron is for?
[22:29] <ikonia> no, monitoring
[22:30] <ikonia> there are already monitoring plugins for most proper monitoring solutions
[22:30] <ikonia> sorry, I just think it's a bad/pointless idea
[22:30] <NomadJim> apticron doesn't do the monitoring it's an addition
[22:30] <ikonia> what ?
[22:30] <NomadJim> i'm not quite sure what you're arguing
[22:31] <ikonia> that uptime monitoring included as part of aptcron is a bad idea
[22:31] <ikonia> which is what you stated
[22:31] <NomadJim> i'm saying uptime monitors like https://www.pingdom.com/
[22:31] <NomadJim> could add apticron
[22:31] <ikonia> why ?
[22:32] <greppy> NomadJim: xymon can monitor apt status, including apticron.
[22:32] <ikonia> there are already solutions/plugins for proper monitoring solutions
[22:32] <ikonia> tools like pingdom are not monitoring solutions
[22:32] <NomadJim> greppy:  that's cool. currently trying to learn zabbix, don't know if that's equivalent ot xymon
[22:33] <greppy> NomadJim: xymon used to be hobbit which used to be big brother.  I find that I prefer it to most other monitoring solutions.
[22:36] <NomadJim> http://en.wikipedia.org/wiki/Comparison_of_network_monitoring_systems
[22:36] <NomadJim> looks like they are pretty similar
[22:36] <NomadJim> at least in terms of features
[23:00] <bananapie> fail2ban