[00:12] <Ryan_Lane> howdy. it seems 14.04 is shipped with a broken virtualenv package? ImportError: No module named _sysconfigdata_nd
[00:12] <Ryan_Lane> looks like it's been this way since 13.04? any plans on this being fixed?
[00:13] <Ryan_Lane> it sucks to need to upgrade virtualenv from pip to have a working python system
[00:15] <sarnold> Ryan_Lane: on a first guess, poke barry in #ubuntu-devel  ?
[00:15] <Ryan_Lane> thanks
[00:22] <Ryan_Lane> ahhh user-error. so great.
[00:22] <sarnold> :)
[00:23] <mwhudson> you always figure those out shortly after complaining in public
[00:23] <Ryan_Lane> yep ;)
[01:55] <bananapie> My sound is not working. I am using sox play to test the problem. If I add 'AUDIODEV=hw:0' before the play command, the sound is ok. Without this line, I get no sound. I think sound is being sent to my hdmi port instead of built in speakers, what do I do?
[02:18] <patdk-lap> sarnold, heh
[02:19] <patdk-lap> this renaming nics is screwing me totally
[02:19] <patdk-lap> all my 12.04 machines have starting having this issue when I updated them today
[02:19] <patdk-lap> eth0 renamed to rename2, eth1 renamed to eth2
[02:19] <patdk-lap> and then boot hangs, cause network isn't functional
[02:19] <sarnold> patdk-lap: how'd it wind up on your 12.04 systems?
[02:20] <patdk-lap> not the devbios thing
[02:20] <patdk-lap> just the normal 70-persistant-net stuff
[02:20] <sarnold> !!
[02:20] <sarnold> man I've never seen that stuff go wrong
[02:20] <patdk-lap> but the issue is just more noticable with that devbios thing
[02:20] <patdk-lap> well, what seems to have started to cause the issue
[02:21] <patdk-lap> I had eth0, and eth1
[02:21] <patdk-lap> then I added eth2, and removed eth0
[02:21] <patdk-lap> but it's getting confused
[02:21] <patdk-lap> cause now renames need to go eth0 -> eth1 and eth1 -> eth2
[02:21] <sarnold> ugh
[02:21] <patdk-lap> but eth0 -> eth1 can't happen, so it calls it rename2, and dunno, gets lots
[02:21] <patdk-lap> lost
[02:22] <patdk-lap> seems like a concurrency issue to me
[02:22] <patdk-lap> no idea what package to even think about for this issue :)
[02:23] <sarnold> udev owns /lib/udev/write_net_rules -- I suspect that's who'd I'd fle the bug against first..
[06:40] <lordievader> Good morning.
[10:13]  * [gnubie] waves
[10:13] <[gnubie]> how am i going to permanently apply my changes to the /lib/init/fstab file? somehow, after i rebooted my system, it went back to its original configuration, removing my changes.
[10:16] <mardraum> vi /etc/fstab ?
[10:16] <mardraum> man fstab?
[10:18] <[gnubie]> mardraum: i know /etc/fstab but somehow, there are few tmpfs based partitions that are not stated in /etc/fstab and they're at /lib/init/fstab
[10:18] <[gnubie]> and i want to edit some options into the tmpfs based partitions
[10:24] <[gnubie]> brb
[10:44] <k3> i want file server that share file among windows and ubuntu pc and sharing is based on the user how can i achieve this?
[10:47] <k3> can any one solve my problem?
[11:03] <k3> i want file server that share file among windows and ubuntu pc and sharing is based on the user how can i achieve this?
[11:03] <yossarianuk> is there any info about doing live snapshots with KVM/libvirtd and QCOW images ?
[11:46] <yossarianuk> tried doing a live backup - using - virsh snapshot-create-as --domain test snap1 --diskspec vda,file=/var/lib/libvirt/qemu/snapshot/test.snap.disk.qcow2,snapshot=external --memspec file=/var/lib/libvirt/qemu/snapshot/mem.snap.disk.qcow2,snapshot=external --atomic
[11:46] <yossarianuk> but get the following error
[11:47] <yossarianuk> error: internal error: unable to execute QEMU command 'transaction': Could not open '/var/lib/libvirt/images/test.img': Could not open '/var/lib/libvirt/images/test.img': Permission denied: Permission denied
[11:47] <yossarianuk> --> running 14.04 .
[11:47] <yossarianuk> any ideas why ?
[12:13] <med_> jamespage, I'll also test the 0.80~rc1 this weekend. Thanks for pushing it out.
[12:13] <jamespage> med_, np
[13:02] <Voyage> this 96.47.70.253 is my VPS ip, the VPN is something else.  now is see nmap, the VPS has bazzillions of ports open. I think its because of the openvpn client?  If I didnt installed any servers/daemons, why the ports are open? is there a security risK?
[14:06] <BLZbubba> ugh so what is the proper way to report kernel panic problems on 14.04?
[14:08] <patdk-wk> spin around 3 times, and tap your heals together
[14:09] <patdk-wk> file a bug against that package that cause it
[14:09] <patdk-wk> be it the kernel in general or a dkms module
[14:16] <BLZbubba> heh
[14:17] <BLZbubba> kernel, sadly.  nfs-rdma appears to be completely useless
[14:30] <bitbyte> hey guys really silly question
[14:30] <bitbyte> im trying to kill a process on my server and iv’e tried with sudo kill <PID> but its not closing the process
[14:30] <bitbyte> and also tried sudo kill  <pid>
[14:30] <bitbyte> and kill <pid> as its running under my current user
[14:31] <bitbyte> any ideas ?
[14:31] <rbasak> bitbyte: kill effectively asks the process to terminate
[14:31] <rbasak> bitbyte: you can force it with kill -9 <PID>, but then the process can't clean up. Eg. you'd get corruption on a database server.
[14:31] <bitbyte> so if the process is stuck it’ll queue the kill command till its unstuck
[14:31] <rbasak> Yes, or just ignore it.
[14:31] <rbasak> However -9 is the KILL signal, which cannot be trapped by a process.
[14:31] <bitbyte> ah ok
[14:32] <rbasak> "man 7 signal" for details
[14:32] <bitbyte> well the item im trying to kill is www-browser http://localhost8081
[14:32] <bitbyte> so I can’t think too much damage will happen
[14:33] <zul> jamespage/coreycb: ok migrate and sqlalchemy should be playing nice in utopic now
[14:44] <Kully> hey all; I've got a server that won't reboot, won't shutdown, won't logout, and apache can't launch ("cant open logs"). Any idea's where to look? nothing's been updated just started today. I'm afraid to hard power that it won't come back up
[14:46] <BLZbubba> it sure sounds like your hard drive isn't responding.  can you run dmesg?
[14:47] <Kully> yea
[14:47] <BLZbubba> anything interesting toward the end?
[14:49] <Kully> repeats  INFO: task sync:10048 blocked for more than 120 seconds.
[14:49] <Kully> and then a bunch of call traces after some hex codes
[14:52] <Kully> if it's harddrive related what would I see
[14:53] <Kully> i'm sorting through
[14:53] <rberg> things like sync never finishing... programs loaded in ram will still work but reading from disk will hang
[14:54] <Kully> ok; so like exactly what i'm getting... great.
[14:56] <rberg> can you see what smart has to say?
[14:59] <Kully> no
[15:12] <Kully> how to run smart with raid
[15:18] <rberg> software raid?
[15:18] <Yaann> Hi !
[15:18] <Yaann> I'm looking for a way to generate a shutdown initramfs on ubuntu, does anybody know how to do that ? It needs to disconnect properly the nbd root device.
[15:24] <onorua> guys, I need to install Ubuntu server on 100 nodes, on encrypted by luks discks
[15:24] <onorua> do you have any documentation on how to do this?
[15:25] <onorua> Something like kicktool for Gentoo
[15:26] <rberg> fai is a great tool for mass imaging.. but I have never used it to setup luks.. so I cant speak for that
[15:31] <xnox> onorua: use automated d-i pressed installation which allows preseeding full-disk luks & passpharases for encryption.
[15:32] <onorua> xnox, I'll give it a try
[15:33] <xnox> onorua: if you google it there is plenty of official resources and guides and howtos.
[16:28] <Elfuego> hey anyone around who might be able to help with an issue im having - high CPU/ram overheads when using SMB shares on ubuntu server
[17:44] <Voyage> heres my ifconfig http://pastie.org/9134287 I want to setup this server (its a vps) as a vpn client. But I dont want to loose ssh and port 8080 access (which does happens when I conenct to vpn server. it looses direct connectivity).  What should be the ips in the rules ip rule add from IP-HERE table 10        ip route add default via IP-HERE table 10
[17:47] <sarnold> Voyage: check your system's routing tables before and after you start the vpn client. I have a feeling the difference between the two will let you know what needs to be modified
[17:47] <Voyage> hm
[17:47] <Voyage> sarnold,  am. how to check the routing table for my VPS?
[17:49] <Voyage> sarnold,  ip route command tells me : for this : default via 96.47.70.225 dev eth0  metric 100       ,       96.47.70.224/27 dev eth0  proto kernel  scope link  src 96.47.70.253 .       Whats my gateway ip?
[17:50] <sarnold> Voyage: heh can you pastebin that intsead? it's miserable to read in irc :) hehe
[17:50] <Voyage> ok
[17:50] <sarnold> (the pastebinit package is awesome for pastebinning the output of commands)
[17:51] <Voyage> sarnold,  http://pastie.org/9134306
[17:51] <patdk-wk> are you using openvpn?
[17:52] <Voyage> patdk-lap,  yes
[17:52] <sarnold> Voyage: there we go :) 96.47.70.225 is your gateway
[17:52] <Voyage> sarnold,  ok
[17:52] <Voyage> sarnold,  am. how to check the routing table for my VPS?
[17:52] <sarnold> sarnold: same thing
[17:53] <patdk-wk> wait, what is the question here?
[17:53] <patdk-wk> oh, you want default route over the vpn
[17:53] <Voyage> so this is ok ?
[17:53] <Voyage> ip rule add from 96.47.70.253 table 10
[17:53] <Voyage> ip route add default via 96.47.70.225 table 10
[17:54] <patdk-wk> what is the ip source of these ssh and 8080 clients?
[17:54] <Voyage> patdk-wk,  i wanted to check the table routs before doing anything. as sarnold  suggested
[17:54] <patdk-wk> 96.47.70.253?
[17:54] <Voyage> yes
[17:54] <patdk-wk> only? ever?
[17:54] <patdk-wk> what?
[17:54] <patdk-wk> that is your own ip
[17:55] <Voyage> .253 is my vps. that is what I have made a vpnclient.    but as soon as I died. I lost ssh and other ports
[17:55] <patdk-wk> this makes no sense
[17:55] <patdk-wk> yes, you don't understand the issue
[17:55] <Voyage> patdk-lap,  .253 is not my own ip
[17:55] <patdk-wk> it's the ip on the vps
[17:55] <Voyage> patdk-wk,  whats the issue?
[17:55] <Voyage> yes
[17:55] <Voyage> so?
[17:55] <patdk-wk> and your running the vpn client on the vps?
[17:55] <Voyage> yes
[17:55] <patdk-wk> so when it connects, you loose ssh to the vps?
[17:55] <Voyage> I connected to to a vps server
[17:55] <Voyage> yes
[17:56] <Voyage> i loose all. even ping
[17:56] <patdk-wk> yes, cause default route is going over the vpn
[17:56] <patdk-wk> is that what you want?
[17:56] <Voyage> hm
[17:56] <patdk-wk> do you want, by default, all internet traffic to use the vpn, or not
[17:56] <Voyage> I want the VPS to browse/ use internet through vpn server. but I also want the VPS to be available for ssh and 808
[17:56] <Voyage> 8080
[17:57] <Voyage> patdk-wk,  got my point?
[17:58] <patdk-wk> ya, but you need to do it another way
[17:58] <Voyage> what way?
[17:58] <patdk-wk> so yo uhave table 10
[17:58] <Voyage> I dont know whats table 10
[17:58] <Voyage> really
[17:59] <Voyage> heres what i got help from https://forums.openvpn.net/topic7175.html
[17:59] <Voyage> patdk-wk,  but first. How can I see the table routes already present. ? like sarnold  suggested?
[18:00] <Voyage> patdk-wk,  secondly, will doing this solve my problem?
[18:00] <Voyage> ip rule add from 96.47.70.253 table 10
[18:00] <Voyage> ip route add default via 96.47.70.225 table 10
[18:02] <patdk-wk> http://pastie.org/9134328
[18:02] <patdk-wk> I wonder how well from ip works
[18:02] <Voyage> patdk-wk,  sarnold  http://pastie.org/9134331
[18:03] <patdk-wk> I normally do this via mac address matching
[18:06] <Voyage> patdk-wk,  nothing  chanted with what i did in table 10. http://pastie.org/9134338#1,7-9
[18:06] <patdk-wk> it shouldn't
[18:07] <patdk-wk> cause you didn't change the routing table
[18:07] <Voyage> patdk-lap,  the routing tables should have changed. shouldnt it
[18:07] <patdk-wk> ip rule show
[18:07] <patdk-wk> no
[18:07] <patdk-wk> you are viewing the main table
[18:07] <patdk-wk> you edited table 10
[18:08] <Voyage> patdk-wk,  http://pastie.org/9134346#22
[18:08] <Voyage> patdk-wk, ok. how to see table 10? and after that command that changed table 10 . are we good to go?
[18:09] <patdk-wk> ip route show table 10
[18:09] <Voyage> $ sudo ip route show table 10
[18:09] <Voyage> default via 96.47.70.225 dev eth0
[18:09] <patdk-wk> looks ok
[18:09] <Voyage> good to go?
[18:09] <Voyage> ok
[18:09] <Voyage> now I should run the openvpn client. ?
[18:09] <patdk-wk> ya
[18:09] <Voyage> and I wont loose ssh
[18:09] <Voyage> ok
[18:09] <Voyage> great
[18:10] <patdk-wk> no idea
[18:10] <Voyage> let me see
[18:10] <Voyage> am..
[18:10] <Voyage> no idea? why?
[18:10] <patdk-wk> I have never tested with that method
[18:10] <Voyage> https://forums.openvpn.net/topic7175.html
[18:10] <Voyage> oh ok
[18:10] <patdk-wk> in my case, I was concerned about the source mac address of packets, not the source ip
[18:10] <patdk-wk> cuase the source ip, I was sending to many gateways
[18:10] <Voyage> k
[18:10] <Voyage> will get back in a minut after starting
[18:11] <patdk-wk> I was doing symetric multihoming
[18:11] <patdk-wk> asymetric is much simpler
[18:23] <patdk-wk> so ,either it worked
[18:23] <patdk-wk> or he lost his internet
[18:24] <sarnold> lets hope it worked :) i don't know details of any of the vpn things..
[18:24] <Voyage> patdk-wk,  looks working.! thanks! I just dont know how to be sure that openvpm wont stop as close the console (ssh console to vps)
[18:24] <patdk-wk> use nohup
[18:24] <patdk-wk> or better yet
[18:24] <sarnold> Voyage: you could run it in tmux or screen
[18:24] <patdk-wk> just set it up correctly in /etc/openvpn
[18:24] <sarnold> oo
[18:24] <sarnold> better
[18:24] <patdk-wk> and use the /etc/default/openvpn to do it
[18:25] <Voyage> sarnold,  or just on "screen"?
[18:27] <Voyage> hm
[18:28] <Voyage> Now if some one wget / curl some site on the VPS , it will use the ip of VPNserver ? correct
[18:29] <chestwood96> can anyone quickly eplain something about load averages?
[18:35] <Voyage> patdk-wk,  sarnold  it seems that I cant ping google or wget google or do anything from VPS but I can how ever , on my local pc, do ssh -D 8080 moez@theVPS-IP and browse web. guess what. the whatismyip.com tells me that my ip is of the VPN server
[18:37] <patdk-wk> Voyage, that is what you wanted
[18:37] <patdk-wk> you said you wanted all internet traffic to go over the vpn, by default
[18:37] <Voyage> hm
[18:38] <patdk-wk> now, if it isn't working, it's an issue with your vpn
[18:38] <patdk-wk> what changed had nothing to do with that
[18:38] <Voyage> if i do ping google.com form my VPS. it should work. no?
[18:38] <patdk-wk> I don't know
[18:38] <patdk-wk> I didn't setup the vpn, or setup the firewall rules on it
[18:44] <sarnold> chestwood96: load averages aren't as useful for measuring a system's load as their name implies :)
[18:45] <chestwood96> but they show that it is overloaded dont they?
[18:46] <chestwood96> A system load of 8 is not good on a songle core system if i get this right
[18:46] <Voyage> patdk-wk,  see. mylocal pc > -D 8080 Vps (which is VPN client) > vpn server. works fine.       but when I ssh to my vps, ping google or do wget or curl. Resolving google.com (google.com)... failed: Temporary failure in name resolution.
[18:47] <sarnold> chestwood96: it could be that you've got eight tasks set to run at an idle or low priority, it might not be anything to worry about or it might be big trouble -- it's more useful as a measure over time than any specific measurement
[18:47] <patdk-wk> well, there would be your next issue
[18:47] <sarnold> Voyage: does your vps have a route to a dns server?
[18:47] <patdk-wk> your turned up the vpn
[18:47] <patdk-wk> can you stil lcontact your dns servers?
[18:47] <patdk-wk> man
[18:48] <Gallomimia> using ubuntu server 14.04 installer to install a new system (on my desktop, as suggested by them, will add desktop packages later) with encrypted LVM on top of a software raid5. does the raid need a first partition with some kind of unencrypted boot volume, who determines what size it is and installs the various things, and how do i then configure my encrypted volumes afterward, with mountable such and such
[18:48] <chestwood96> well i am running a raid5 mdadm array wich i am triing to share over nfs
[18:49] <chestwood96> and i get reasonable writes locally but less than 10mb/s over nfs
[18:49] <Gallomimia> yeah that i want to do secon
[18:49] <Voyage> sarnold,  I dont know... patdk-wk  how to find out?
[18:49] <Gallomimia> ew slowness
[18:49] <Voyage> patdk-wk,  sarnold  how to contact  dns servers?
[18:49] <sarnold> Voyage: find the ips of your nameserver in /etc/resolv.conf and make sure that you have routes that can cover those resolvers
[18:50] <Voyage> sarnold,  nameserver 96.47.64.10
[18:50] <Voyage> search keyreactions.com
[18:51] <sarnold> Voyage: does your routes let you reach that host? any chance ping to it works?
[18:51] <Voyage> sarnold,  how to create routes for that now?
[18:51] <Voyage> should I ping  96.47.64.10 ?
[18:51] <Gallomimia> diagnose before making changes
[18:51] <sarnold> Gallomimia++ :)
[18:52] <Voyage> sarnold,   ping  96.47.64.10
[18:52] <Voyage> PING 96.47.64.10 (96.47.64.10) 56(84) bytes of data.
[18:52] <Voyage> .. no output
[18:52] <chestwood96> sarnold: the hardware is really weak so it is likely to overload
[18:52] <Voyage> sarnold,  ya. I cant ping the  96.47.64.10. what should I do
[18:53] <sarnold> chestwood96: could be; check ps aux -- and look for 'R' and 'D' tasks; the 'R' tasks are ready to run and the 'D' tasks are blocked on IO; hopefully that'll give you some feeling of which tasks are being starved
[18:54] <sarnold> Voyage: since the kernel didn't complain about a route to the host, you've got a route -- it may not make sense for a local dns resolver if it goes back to your laptop or something -- so if the packets aren't getting responses they've either been firewalled on your vps or elsewhere...
[18:55] <Voyage> sarnold,  everything was working fine untill the changes.
[18:55] <Voyage> sarnold,  the above said things are done on the vps
[18:56] <sarnold> Voyage: I don't know if openvpn decided to firewall off everything or if it decided to install broad routes over its tunnel rather than specific small routes over the tunnel...
[18:57] <chestwood96> so there is a kworker procces on D and some jbd2/md0-8 when writing locally
[18:57] <Gallomimia> i had this problem when i was trying to configure my vpn. i eventually gave up. i wanted local traffic on my box to still take local routes and only traffic to the internet would use the vpn. i couldn't get it working. but your dns packets aren't going where they should. either you have the wrong dns now configured, and it should be something in your vpn's services, not something at home. perhaps its still set as your local 
[18:58] <Gallomimia> anyone know how to install a bootable encrypted raid5?
[18:59] <Gallomimia> im a little mystified halfway thru the config
[19:00] <chestwood96> sarnold: when writing over nfs there are 4 nfs processes on D
[19:00] <sarnold> Gallomimia: hrm. I'm got a feeling I've heard that grub is far happier with mirrors than with other raid types.. throwing in encryption can't make it any easier :)
[19:01] <sarnold> chestwood96: sounds fair. you may be able to tune the nfs processes in some way -- block sizes, tcp window sizes, etc., but I'd go slowly when making changes
[19:02] <chestwood96> sarnold: for some reasons reads are ok over nfs, i get 110mb/s wich is pretty much gigabit
[19:02] <sarnold> chestwood96: oh, try local writes to the same pool?
[19:02] <chestwood96> my clients are 2 xenserver6.2 (centos) machines
[19:03] <chestwood96> local writes are about 130mb/s the mdadm array runns pretty well
[19:04] <sarnold> chestwood96: are those read speeds because the clients are using cached reads rather than going to the server for their reads? :)
[19:04] <Voyage> sarnold,  after this http://pastie.org/9134455      I cannot do ping google.com from the vps. but I can how ever do  mylocal pc > -D 8080 Vps (which is VPN client) > vpn server. works fine.       but when I ssh to my vps, ping google or do wget or curl. Resolving google.com (google.com)... failed: Temporary failure in name resolution.
[19:04]  * Gallomimia imagines this install magically working and presses forward
[19:05] <chestwood96> sarnold: i do not know where that would becahced im triing with 1GB files
[19:07] <Voyage> sarnold,  patdk-wk  any clue?
[19:07] <mesut> is it possible port ubuntu touch htc desire hd ?
[19:08] <sarnold> mesut: try #ubuntu-touch ?
[19:08] <Gallomimia> anything's possible. the question becomes are you going to do it?
[19:08] <sarnold> Voyage: hrm, no, looks like that ought to work :)
[19:08] <mesut> opss sory
[19:08] <Voyage> then why sarnold , the dns is not working?
[19:08] <chestwood96> sarnold: btw iscsi has slightly lower reads but more than tripple writes (35mb/s)
[19:18] <BLZbubba> ok this is weird, i have a mysql test that is cpu bound - using 100% of every core - but load avg is under 1.  how can that happen?
[19:19] <patdk-wk> BLZbubba, exactly what does the one have to do with the other?
[19:19] <patdk-wk> do you know what loadavg is?
[19:19] <sarnold> BLZbubba: sounds like the machine does nothing else
[19:20] <Voyage> sarnold,  for learning . this is very strange. what possible reasons can there be?
[19:20] <patdk-wk> voyage, basic networking 101
[19:21] <sarnold> Voyage: firewalling on your local system, firewalling on your gateway, firewalling before the dns server, firewalling on the dns server, access control lists on the dns server.. yanked wires, broken switches, broken routers, it goes on :)
[19:21] <Voyage> patdk-wk,  ya, but what did I do wrong
[19:22] <Voyage> ..... :(
[19:22] <Voyage> sarnold,  can I be close or with a clue. there would be a solution any way
[19:22] <patdk-wk> Voyage, I didn't setup your vps, your vpn, your ..., I dunno
[19:24] <sarnold> Voyage: I'm sure there's a solution, it's all just software, but I don't know enough about your provider or you system to really offer a good next step. if your vps can directly contact dns servers itself you could run your own recursive resolver such as pdns-recursor, but that's a big step to take if you don't want its other features..
[19:24] <Voyage> patdk-wk,  sarnold  line 72. thats where the vps was setup. normal. http://pastie.org/9134455#69
[19:31] <chestwood96> sarnod: would more cpu power fix my problem?
[19:31] <sarnold> chestwood96: no idea, sorry
[19:31] <chestwood96> sarnold: the machine could just be insufizient
[19:31] <sarnold> chestwood96: could be, or could be some tuning could make it more effective.
[19:32] <Elfuego> side note - do you know if they fixed the issue with loading huge SMB shares causing a large amount of cpuram usage
[19:32] <chestwood96> itl like a dinosaur athlon 64 i blought for 12 bucks including a mainboard
[19:33] <sarnold> chestwood96: hahah fantastic :) performance per dollar is pretty good ;)
[19:34] <chestwood96> it did its initial job pretty well but mow i repurpoused it
[19:34] <chestwood96> it was easily strong enough to let my spare 7970 mine a buch of dogecoin
[19:34] <sarnold> wow much coin!
[19:34] <chestwood96> well i stoped because of power efficiency
[19:36] <mgw> stgraber: yesterday I asked about resolv.conf behavior and you said the secondary is only used for failover
[19:36] <mgw> But I mis-presented my situation
[19:36] <mgw> I actually have dhcp+resolvconf+dnsmasq
[19:36] <mgw> and dhcp is serving up two IPs
[19:37] <mgw> I want to be sure that if I take the secondary (as served by dhcp) offline, everything will continue smoothly
[19:37] <sarnold> take it out of the dhcp configuration beforehand?
[19:37] <mgw> that's probably the safe way
[19:37] <chestwood96> sarnold: I could get a core 2 for 40 bucks, that should do doesnt it ?
[19:38] <mgw> and then, what's a safe way to force renewal
[19:38] <mgw> I have salt installed, but am afraid of losing connectivity
[19:38] <mgw> if I run dhclient -r
[19:38] <sarnold> mgw: easiest is to wait however many seconds you've configured :)
[19:38] <mgw> sarnold: thanks. And second easiest?
[19:38] <sarnold> chestwood96: could be, those core 2 chips sure seemed impressive seven or eight years back :)
[19:39] <sarnold> mgw: dunno :/
[19:39] <mgw> sarnold: thanks
[19:39] <mgw> if I run dhclient -r I will lose network, right?
[19:40] <sarnold> mgw: probably; you could also try ifdown eth0 ; ifup eth0 -- or whatever the nics are ..
[19:40] <chestwood96> sarnold: its apparantly about 3x faster than the 3000+ so that might work
[19:41] <sarnold> chestwood96 :D
[19:43] <chestwood96> but as I googled this seems to be a bug with nfs
[19:45] <sarnold> chestwood96: could be, among my pals it seems more folks prefer to use illumos-based nfs servers than linux nfs servers, but that may say more about my pals than the situation :) hehe
[19:45] <chestwood96> what are illumos-based servers?
[19:45] <sarnold> chestwood96: smartos, omnios, openindiana, nexenta
[19:47] <chestwood96> sarnold: as if that poor old 3000+ could handle ZFS
[19:47] <sarnold> :)
[19:51] <chestwood96> sarnold: Thanks a lot it was apleasure talking to you, i am going to try out ubuntu 10.10 where it apparantly works(according to the bugtracker) cya
[19:52] <sarnold> chestwood96: yikes
[19:52] <sarnold> chestwood96: 10.10 has been eol for ever :)
[19:53] <sarnold> chestwood96: pick from one of these :)  https://wiki.ubuntu.com/Releases
[19:53] <sarnold> chestwood96: 12.04 LTS or 14.04 LTS would be 'best' choices, you won't have to upgrade immediately and they should be pretty stable
[19:55] <chestwood96> its running 14.04 atm
[19:55] <chestwood96> pretty fresh install
[19:55] <sarnold> oh good good
[19:56] <chestwood96> but there seems to be a bug in >10.10 where nfs goes ammok https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1006446
[19:58] <sarnold> chestwood96: comment #29 mentions switching to the deadline scheduler; switching schedulers may do the trick and would be far faster to test :)
[19:59] <chestwood96> im at it just have to fund out how
[19:59] <sarnold> chestwood96: comment #49 also looks easy and simple to try thogh it may not be -your- problem... but again, some tuning may help
[20:00] <chestwood96> allready tried 49 -> reads went down writes stayed the same
[20:01] <sarnold> chestwood96: so, scheduler.. look around in /sys for a block/sda/ directory -- mine has it under /sys/devices/blahblah/ in a really inconvient location :) but go find those..
[20:02] <sarnold> chestwood96: ah, find /sys -name scheduler  ought to find them
[20:02] <sarnold> chestwood96: cat those schedulers, see what they are, and pick a different one to test with :)
[20:04] <chestwood96> cat /sys/devices/virtual/block/md0/queue/scheduler returns "none" am I doing it wrong?
[20:04] <sarnold> chestwood96: how about the block devices underneath the raid?
[20:07] <Voyage> sarnold,  patdk-wk  I see that when the openvpn is closed. I get normal internet connectivity back. Now I can use DNS and ping google.com
[20:07] <patdk-wk> yes, cause your dns servers are foobar
[20:07] <patdk-wk> we know that
[20:07] <patdk-wk> use dns servers provided by your vpn
[20:07] <Voyage> foobar?
[20:07] <patdk-wk> make sure dns works over the vpn
[20:07] <Voyage> ok. how?
[20:07] <patdk-wk> I dunno, I didn't setup the vpn
[20:08] <Voyage> how to use the dns provided by my vpn?
[20:08] <patdk-wk> again, I didn't setup your vpn
[20:08] <Voyage> any idea?
[20:08] <chestwood96> sarnold: i think ill just put the drives into one of the xenserver hosts and share it from there they got plenty of power
[20:09] <sarnold> chestwood96: hehe makes sense
[20:09] <sarnold> give that poor machine a rest :)
[20:09] <chestwood96> sarnold: the actual point of it was to have the storage separated so i can turn of either one without compormising the other
[20:12] <chestwood96> sarnold: tanks again, you were very helpfull. good night whatever timezone youre in
[20:13] <sarnold> chestwood96: thanks, have fun! :)
[22:10] <rostam> Hi I see some of the ubuntu packages have "ocaml", what it means? thx
[22:10] <sarnold> rostam: ocaml is a functional programming language similar to ml or haskell
[22:11] <sarnold> rostam: those would probably be language bindings for whatever library it is into ocaml
[22:12] <rostam> sarnold, thank you.
[22:26] <raydeo> does anyone know why upstart (consistently) gets into unfixable states for me? specifically "start nginx" <hang/Ctrl-C> "status nginx" <start/starting> "stop nginx" <hang/Ctrl-C> "status nginx" <stop/starting> "start nginx" <start/starting> rinse/repeat
[22:27] <raydeo> I have other scripts with this behavior, and I cannot find any workaround to even get out of this state.. if I copy nginx.conf to nginx2.conf and start that, it works fine. I have this problem on 12.04 was hoping 14.04 would solve it no to avail.
[22:27] <raydeo> the only way to fix the nginx state appears to be to reboot
[22:27] <raydeo> but more importantly why does it keep breaking
[22:30] <Elfuego> nginx is pretty hard to get worknig right - but ones u do its pretty solid
[22:30] <raydeo> sorry but this isn't an nginx question specifically
[22:30] <raydeo> I have other services that are breaking in upstart as well
[22:32] <rostam> HI I am getting a linker error and I do not what package I am missing: /usr/lib/x86_64-linux-gnu/libavcodec.so: undefined reference to `schro_decoder_reset'      How could I find the missing package and resolve it? thx
[23:09] <sarnold> rostam: looks kinda like "schroedinger" of some sort: http://ubuntu-codesearch.surgut.co.uk/search?weighted=1&q=schro_decoder_reset
[23:11] <rostam> sarnold, thank you