=== sz0 is now known as sz0` [01:16] Hi all, I am a major ubuntu server newbie (using it for home server). During upgrade to release 14.04, I took a screen shot of a message about PostgreSQL. Is there a way to "paste bin" it here? [01:16] I hate to ask a question that I flat out know nothing about but trying to figure out what it's telling me is lke greek right now. [01:19] bbl === Stuxnet is now known as Stuxnet[A] [01:19] * Stuxnet[A] is now away - Reason : [01:20] Stuxnet[A]: turn off the away announce [01:21] Stuxnet[A]: you can upload it to imgur or somewhere === zz_DenBeiren is now known as DenBeiren === Lcawte is now known as Lcawte|Away === esde is now known as Guest96003 === Guest96003 is now known as esde === zz_DenBeiren is now known as DenBeiren [06:03] anyone here have any experiance with zfs? [06:03] my usb pool isn't mounting on reboot [06:04] jnollette: any errors? #zfsonlinux might be helpful.. [06:05] beatface, basically they're not mounting, so they're unavailable [06:05] having this issue with all usb zfs devices [06:05] sata devices, no problem [06:07] 'zpool import -d /dev/disk/by-id poolname' work? [06:07] it says pool already created [06:07] created/imported [06:07] 'zpool status' [06:08] http://hastebin.com/buzefijewi.vhdl [06:09] bermuda being the usb devices pool? [06:09] yeah [06:09] and the cache for nevis [06:09] and in the zpool import command, did you replace poolname with bermuda? [06:09] yep! [06:10] 'zpool import -d /dev/disk/by-id/wwn-0x5000c5004f2c431f bermuda' [06:10] minus quote [06:10] just /dev/disk/by-id not the wwn [06:11] can i destroy the device then reimport [06:11] i have the data on another drive [06:12] if not, possibly a 'zpool export bermuda' followed by 'zpool import -d /dev/disk/by-id bermuda' again not the wwn at the end [06:12] yes, you can, might not be nessecerry though [06:13] now its online! [06:13] hmmm [06:14] and probably with either ata-* or scsi-* names under zpool status? i've found it more reliable than wwn names [06:15] how could i force umount [06:16] of a zfs dataset? [06:16] or a whole pool? [06:16] trying to export and import the nevis pool to bring the cache back online [06:18] if it is failing at export saying it's being used. you have to stop whatever is using it.. nfs-kernel-server being a common one if you have exports on the pool... cache are non important though and can easliy be removed and re-added [06:18] non-important as the data cached are already stored on the raidz1 === sz0` is now known as sz0 === sz0 is now known as sz0` [06:30] bearface, thats for the help, i got an rc.local blob to fix me up on load [06:30] *thanks [06:31] :) [06:33] zfs is relatively painless other than this one issue... its really fast once you strip down a few of the extras [06:33] im getting like 200mb/s off 5 drives [06:33] if the proper paths /dev/disk/by-id is in the /etc/zfs/zpool.cache the mountall bundled with the unbutu-zfs cache should take care of importing things at boot [06:34] ubuntu-zfs package* [06:37] and aye, i quite like zfs myself too :) [06:53] DonRichie, hey dude [06:54] i got this all hammered out [06:54] working it like a boss!! [07:07] Nice to hear grendal_prime, I also improved my knowledge a bit since yesterday [07:08] ya my problem from what i can triage out of our session and the results was that simple client config portion. [07:08] would you like to see the tool im using to manage this now? [07:09] maybe later, I will soon need to go to work [07:10] ewwww avoid that if you can [07:10] I tell all young men i meet... "Mary rich!" [07:11] it makes everything so much easyer...just remember ugly rich chicks need loving to! [07:11] I got lucky...my wife is beauty full and rich. [07:11] heheh [07:12] I unfortunaly never met such a woman. Maybe some time in the future [07:13] allright well i got that entire setup process down to about 60 seconds now. Mostly just a few packckage intstall and the config process is all done with a gui (addin the servers and their ips that is) [07:14] and i got to tell ya its made everything on my internal network much faster [07:14] peace out im studdying [07:15] Nice to hear that. Have fun! === elliotd123_ is now known as elliotd123 [08:51] Patrickdk: found the real issue === abcd7z is now known as linocisco [08:57] anybody used OCSInventory-NG on ubuntu server? [08:57] with windows 7 clients agent? [11:53] I've got a dumb bash problem I'm sure someone here would be able to help me out with... [11:54] I'm trying to rsync stuff but failing when passing certain variables to my bash script... [11:54] rsync -arvz --timeout=0 --partial --progress 'username@someserver.com:/some/path/$1' /some/other/path/ 2>&1 [11:56] Provided $1 is a valid file / folder / ... shouldn't the above work? [11:56] "" quotes and '' quotes mean something different. [11:56] The difference is pretty important here [11:56] not even sure if the quotes are needed in this case, i don't believe they are. [11:57] Well if for instance $1 contained spaces, rsync would fail. [11:57] then use double quotes [11:57] Yeah, I think maxb has it. [11:58] Thanks guys :) === Lcawte|Away is now known as Lcawte === bilde2910|away is now known as bilde2910 === bilde2910 is now known as bilde2910|away === bilde2910|away is now known as bilde2910 === bilde2910 is now known as bilde2910|away === FreezingAlt is now known as FreezingCold [16:17] does kdump work on ec2 instances? I'm getting a hard hang when I test, the systems don't reboot. [16:21] according to: http://xenbits.xen.org/docs/4.3-testing/misc/kexec_and_kdump.txt [16:21] no, it won't work on ec2 [16:22] patdk-wk, looks like I'm switching to openstack, thanks [16:57] Hi folks. I've been trying to track down an issue with a 14.04 install for the past week. The problem is mixed; sometimes it's a kernel panic, sometimes "recursive error detected", most often the network silently crashes when copying data. [16:58] I had been running it on a RAID 5 array; someone in #ubuntu suggested trying it without RAID, so I wiped and reformatted one of the drives and installed 14.04 solo to try that. [16:58] It just lost the network again doing a large transfer. [16:58] Any idea what to try next? [16:58] Aside from the hard drives, the rest of the system is ~9 years old and had no issues of this kind under 12.04 previously. [17:01] in the past I had a very similar issue with writing data over the network to a raid array, in my case the problem was interrupt sharing, I had to force msi-x on my network adapter and this solved my stability problems. [17:01] Hm. I don't know what msi-x is. It also is happening even without the RAID now. [17:04] Crell, just a form a transporting an interrupt request, usually more efficient [17:05] the solution rberg said, had nothing to do with raid [17:05] but just that the disks and network used the same irq [17:05] I cant remember what I noticed in dmesg that lead me to look at /proc/interrupts [17:05] and one of those two drivers had a bug :) and msix worked around it [17:05] Would that also sometimes cause kernel panics and such rather than just network fails? [17:06] The last kernel panic I saw DID say something about scsi and irqs in the backtrace. [17:06] starving interrupts can cause all sorts of fun stuff, so yeah [17:06] for me, my wifi card caused kernel panics [17:06] I had to disable my wifi, install, upgrade, then I could use the wifi [17:07] Crell, you can see it pretty easily in htop, the cpus being pounding on get sys % much higher than the rest [17:07] I didn't fully understand it, but it involved those [17:07] for me I didnt have the issue until I added a large array because more sata controllers need more interrupts [17:08] hm. OK, so how do I set this network setting? [17:09] (Presumably if this works, then I would try going back to the RAID array next and see if it works there?) [17:10] I would start by taking a look at /proc/interrupts to see if you have much sharing going on [17:10] Crell, well you want to confirm where the problem is first before you start changing stuff. watch -d 'cat /proc/interrupts' [17:10] while you load the network and disk and see how the interrupts are distributed. === markthomas|away is now known as markthomas [17:11] OK, so reboot to reset the network, continue copying gobs of data while that's running, and see what happens? [17:12] yup [17:12] * ppetraki lunches, bbiab [17:12] OK, stand by. It could take a while to crash again. [17:12] Thanks. [17:14] hello everyone. I'm trying to install Ubuntu Server 14.04 LTS and configure 2 of the drives in my system as a RAID 1 array, but every time I try to use the installer to do it (using the Configure software RAID option) it freezes [17:18] OK, so IRQ 23 has eth0, as well as usb5. It's going nuts with over 700k counts. (Whatever the CPU0 column means.) [17:19] I don't know what port the hard drive is on... there's several usb irq ports but none that say scsi. (It's a SATA drive, which usually lists as SCSI, IIRC?) [17:39] OK, there's the broken pipe message again from rsync. [17:40] ppetraki rberg-: I have proc/interupts up on watch. What am I looking for on it? [18:05] Crell, the counters associated with the driver name of the device, the label is all the way to the right like enpFOO for network device [18:06] OK... [18:07] IRQ 23 is the most active, and is labeled IO-APIC-fasteoi uhci_hcd:usb5, eth0 [18:07] Which I think means it's being shared by usb5 and eth0. [18:08] Which, if usb5 actually means "SATA drives", would explain why it's well above 7 million after rsyncing across the network. [18:08] Crell, can you identify which one is your storage controller? [18:08] None stand out. [18:08] There are two pata_via lines, but they're almost empty which makes sense as I have no PATA drives installed. [18:09] usb4 is also ahci, [18:09] so what does your RAID run off of? the onboard sata or like a LSI HW RAID? [18:09] usb1 is also uhci_hcd:usb3 [18:09] RIght now there's no RAID, just one SATA hard drive. [18:09] When I had it configured for RAID it was a 3 drive software RAID setup, created by the Ubuntu installer. [18:09] OK [18:10] so AHCI would be the storage controller [18:10] So that's on interrupt 21, whcih has a counter of about 6k. [18:11] IRQ 23 is the network, and the one that's super high. [18:11] * Crell hasn't thought about IRQs since the DOS days... [18:12] well, that is cause mostly, irq sharing was solved [18:12] but that doesn't mean drivers do it correctly, yet [18:12] Apparently. [18:12] What's weird is that I never had this issue in older kernels, only in 14.04. [18:13] But... my laptop (where I'm rsyncing the data from) is also 14.04 (Kubuntu), and it hasn't fallen over once. [18:13] there's something to be said for not upgrading :) [18:13] what kind of network card? [18:13] on-board VIA something or other. I forget the model. [18:14] VT6102, Rhine II (according to lspci) [18:16] So does this sound like the msi-x thing you mentioned would be applicable, if they're not sharing irqs? [18:17] doubt it [18:17] drat [18:17] What would be? [18:18] Is there even a theory as to the root issue? [18:18] msix doesn't fix the irq sharing, it just makes irqs happen much less [18:18] Ah. [18:18] install and configure kdump, collect a dump, post backtrace to lp bug. http://blog.zedroot.org/linux-kernel-debuging-using-kdump-and-crash/ [18:19] Oy. [18:19] That's the "punt to the developers" option then? [18:19] or install the 12.04 kernel, might be able to get away with it using pinning, not sure if it'll work though [18:19] hello, is there a way to install a package without launching its service? [18:19] Running a 2 year old kernel on a current distro sounds like something well out of my expertise area. [18:20] mostly, sometimes we get lucky and can find that same backtrace on interwebs and viola patch! [18:20] Ie, I fear even more "weird things" happening. [18:20] * Crell tries to make sense of this blog post to get a dump. [18:21] ppetraki: Would kdump work if I'm not getting a panic most of the time, just a network disconnect? [18:22] Crell, it only covers half of your cases, correct, but if you deliberately stress it you'll hopefully run into it sooner than later [18:22] Lovely. [18:22] So install this, and then beat on it until I can cause a kernel panic? [18:22] yup [18:23] Sigh. Not how I had hoped to spend my day. [18:23] there's option C, use this as excuse to buy a new all Intel MB [18:23] I have had that suggested to me. [18:23] Why intel MB specifically? [18:24] well, intel southbridge and nic :) [18:24] That would essentially mean whole new computer, as everything else is a 9 year old AMD Athlon-based system. :-) [18:24] it's more like, why AMD? [18:24] not to mean anything that AMD is substandard or anything it's just for production stuff Intel is the most well traveled [18:24] But overall it sounds like a Linux kernel incompatibility with something or other on the motherboard? [18:25] Oh you mean Intel CPU compatible, not specifically non-Asus/Gigabyte/whoever else makes motherboards these days. [18:25] yup, that's right, which leads to the decision point of 1) downgrade 2) instrument and hope someone can help you 3) upgrade [18:25] well, most of those use intel chipsets [18:26] the ones that use intel cpu but via/nvidia/... chipsets, can still be flaky [18:26] Blargh. [18:26] where 3) is upgrade HW [18:26] it might be, just installing a different nic, will get around the issue [18:26] if the issue is in the nic driver code [18:26] good idea [18:26] hm. [18:27] Let me see if I have a spare NIC lying around. If I do it's probably as old as the rest of the computer. :-) === Noskcaj_ is now known as Noskcaj === huttan_ is now known as huttan [18:27] then it will likely be compatable with that system :) [18:29] I have several modems and multiple sound cards, but no wired NIC. I DO, however, have some old wireless PCI cards. :-) [18:29] Worth a shot, I guess. [18:29] last month, I finally trashed all my old cards [18:30] I no longer have any pcix, pci, isa nics [18:30] I have a closet full of hardware I was going to audit and then recycle when I finished the server. [18:30] That was a week ago... [18:31] I used to be a hardware geek, a decade-plus ago. Then I decided being a software geek was cheaper. :-) [18:31] garage+storage unit [18:31] finally removed the old cisco 5500's [18:31] had like 180 100mbit ports on those things [18:31] When I moved from my apartment to a house a few years ago, I donated 7 computers to Free Geek. [18:32] That was only half my closet. [18:32] oh, around 440ports [18:32] I have a Pentium II CPU in my drawer; 2 Pentium III systems. [18:33] ya, I finally got ride of mine a year ago of those [18:33] A box that I THINK has a Pentium II 400 MHz in it... :-) [18:33] I did still have some of the p3 cpu blanks [18:34] just have no reason to run pcix anymore, everything I have is pcie now, and the old pcix stuff is too slow or has limited 64bit support making it annoying [18:38] I couldn't actually tell you which of my cards was which... [18:54] And I fail at getting a wireless card to work in LInux. === bilde2910|away is now known as bilde2910 === markthomas is now known as markthomas|away === bilde2910 is now known as bilde2910|away === bilde2910|away is now known as bilde2910 === bilde2910 is now known as bilde2910|away === bilde2910|away is now known as bilde2910 [21:33] hello [21:33] is there anyone? [21:33] I need help with ubuntu 12 [21:34] and nginx [21:34] every vhost returns 500/502 [21:38] lnxmen: error logs - look at them [21:38] /var/log/nginx/error.log by default === NegativeFlare_ is now known as NegativeFlare [21:38] read that, check what the errors are, see if you can figure it out, if not, copy-paste the error logs into a pastebin and give us the link [21:39] (note I need to head home for the end of the day so I will be slower to respond) [21:39] teward: I did that. [21:39] I will paste it somewhere. [21:40] Generally, php-fpm says that connection was refused. [21:40] http://pastebin.com/ifDRuyWF === hxm is now known as Guest87023 === mikal_ is now known as mikal [22:01] is launchpad down? === markthomas|away is now known as markthomas [22:06] no === bilde2910 is now known as bilde2910|away [22:12] Good evening guys, so I'm working on setting up a webserver and I keep running into problems (on my 5th re-install). First step is to get a static ip going and it's here that I'm running into trouble. I know I should edit /etc/network/interfaces and put iface eth0 inet dhcp to iface eth0 inet static and add address, netmask, network, broadcast, gateway & dns-nameservers. My problem is [22:12] that I'm not 100% sure as I'm behind a normal router in a home network. I've asked my ISP to set up a static IP adress on their end so if anyone could help me understand this it would really mean a lot to me [22:15] Tobbe-82|ServerI: what is your ultimate goal? [22:16] my ultimate goal is to be able to run a few personal website project on my own computer in my own home network. I have been following a tutorial called The Perfect Server - Ubuntu 14.04 but ran into issues that I think comes from not having configed a static ip properly in the beginning [22:16] (I rather like the idea to being able to use a web interface to create databases etc) [22:17] Tobbe-82|ServerI: who will access the website(s)? [22:17] (clients on private/home network or on the public internet) [22:18] At start myself and my brother for a more private information archive based website. I also have a community/forum idea that I'd like to start off in my own server enviroment until it grows [22:18] (I also would like to learn how to set up a proper server in the process). [22:18] so not available from the internet then right? [22:19] available from the web so it needs to be public [22:20] Tobbe-82|ServerI: need to ask again. do you want people from the internet accessing your website? [22:20] No problem pmatulis, I appreciate the help. Yes people from the internet need to be able to access my websites [22:21] ok [22:21] so you'll need to come up with a DNS strategy so people can access your website with a name [22:23] right now I'm just trying to get this working on my local network. but end goal is that yes. (Still waiting for my static ip from my ISP, then i'll need to config my domain name to somehome use nameservers to link this all together). [22:23] Have no clue on how to do that but figure one step and problem at a time ;) [22:24] there are 2 modes to do this. ① get a static IP address from your ISP and set up a static name:address mapping in DNS or ② get a dynamic IP address from your ISP and change the mapping whenever that address changes (using a utility often called a 'dyndns client') [22:25] the IP address here is 99.9% not for your ubuntu server. it is for your router, usually bundled with your modem these days [22:26] I understand that my public ip adress towards the web is different from my local area network one. [22:27] so in order to set up a proper ubuntu server static ip I would need to have the static ip from my ISP first? [22:27] when traffic hits your router/modem you redirect it to your server (called 'port forwarding') [22:27] ahh yes, it what I had in mind, just don't know the practicalities of it. I figured i'd work on getting the server set up and then put that on or is that thinking backwards? [22:28] yes, backwards [22:29] set up your website. test it locally using it's private IP address. you can set up DNS after when you want to provide access to the internet [22:29] s/it's/its [22:29] allright, so first wait to get the static IP from ISP then configure my server with port forwarding. Wouldn't I need to have set a static IP in ubuntu in order to have a static ip to do port forwarding to? [22:29] ahhh good [22:30] so I guess then my question becomes how do I set up a "local" static IP for ubuntu that will just work for my local network? [22:30] yes, port forwarding should be sending to a static address. but with dhcp (dynamic) you can reserve an address for your server's hardware (MAC address); it's effectively static [22:31] so there is a way to use the MAC adress as a static adress and then automatically compensate for the dhcp given ip adress? [22:33] every network card's port has a hardware address (in hexadecimal format). when a dhcp client asks for an IP address it broadcasts it using it's hardware/MAC address. you can instruct a dhcp server to provide a specific IP address to that specific MAC address [22:34] interesting, I did not know that. Would setting that dhcp setting be in my router or a dhcp server I create in my ubuntu server setup? [22:35] these days every router/modem has a DHCP server built in. use it [22:35] it will have a web interface that you will allow you to configure these things [22:36] Oh yes, I just need to figure it out in regards to my router. Thank you so much pmatulis, I'll start here and will most likely frequent this channel daily from here on :) [22:36] that's also where you configure port forwarding [22:36] Learning quite a lot in the process :) [22:37] Tobbe-82|ServerI: np, good luck [22:37] thanks