| === ChmEarl is now known as PryMar56 | ||
| Takyoji | Any specific way to test if DHCP on a network is functional? | 01:50 |
|---|---|---|
| Takyoji | I've been trying to use dhcping, but get an answer of 'no answer', even on my home network, which has working DHCP | 01:50 |
| Takyoji | otherwise is there any other useful commands? | 01:53 |
| Takyoji | or does dhcping work for anyone else at all? | 01:58 |
| sarnold | heh seems like dhcping requires a lot of command-line fiddling.. | 01:59 |
| sarnold | Takyoji: Got answer from: 192.168.1.1 | 02:01 |
| Takyoji | sarnold: Odd, not sure why it isn't working for me. And I assume there's no difference between Ubuntu Server and Desktop, networking-wise | 02:15 |
| sarnold | Takyoji: well, desktop has NetworkManager and assorted things | 02:16 |
| sarnold | (and I tested on my laptop because it's what's easy :) | 02:16 |
| Takyoji | Because I'm testing from desktop versions | 02:16 |
| sarnold | Takyoji: how did you execute dhcping? | 02:16 |
| Takyoji | Alright, I guess PEBKAC, seems to be working fine now, oddly. I must have just been setting the wrong parameters | 02:18 |
| sarnold | or it was a temporary problem, which might not help you much :) heh | 02:20 |
| === arrrghhh is now known as arrrghhhAWAY | ||
| MrPPS | Hey guys, having some issues with SSL. I have a ubuntu 12.04.5 server that has been connecting fine to a server on an SSL secured port, however, recently, I've been having handshake failure issues. I run "openssl s_client -showcerts -connect server:port", and sometimes it returns the cert, other times it returns: "140010279691936:error:140790E5:SSL routines:SSL23_WRITE:ssl handshake | 06:03 |
| MrPPS | failure:s23_lib.c:177" | 06:03 |
| MrPPS | followed by "no peer certificate available" | 06:03 |
| MrPPS | Yet any of my other servers (including 12.04.5) have no issue doing this | 06:03 |
| MrPPS | Any clues? | 06:03 |
| Takyoji | Perhaps the time is misconfigured, as a very random guess | 06:07 |
| MrPPS | appreciate the guess, but time has been verified as being correct :) | 06:08 |
| Takyoji | or otherwise the port isn't actually SSL, it may be plaintext | 06:08 |
| MrPPS | very occasionally, that same command returns the SSL cert | 06:08 |
| MrPPS | i.e., 1/7 times | 06:08 |
| MrPPS | however, the same command run from other servers returns it every time | 06:08 |
| MrPPS | and also, that same command to any other ssl-secured server works fine | 06:10 |
| Takyoji | are the others newer? | 06:10 |
| MrPPS | yeah, they were set up more recently | 06:10 |
| MrPPS | but they're also part of a different network | 06:10 |
| MrPPS | other servers, SSL secured in the same network, give the same issue | 06:11 |
| MrPPS | but only seen from this one client :P | 06:11 |
| Takyoji | what type of service is it? HTTP, POP/IMAP/SMTP, or? | 06:12 |
| MrPPS | IRC | 06:12 |
| MrPPS | connecting to my own IRC network is fine, but connecting to this one is not :( | 06:12 |
| MrPPS | and prior to today, it'd been no issue, and they haven't changed the certs on their servers | 06:12 |
| MrPPS | so just trying to determine what could be causing it :) | 06:12 |
| Takyoji | able to provide a full paste of what the openssl client command returns (if there's any more content)? | 06:16 |
| MrPPS | yeah, one sec | 06:16 |
| MrPPS | Takyoji: an example of a failure: http://pastie.org/9540973 | 06:18 |
| Takyoji | Is it not just outright dropping the connection? | 06:19 |
| Takyoji | given "has read 0 bytes and written 213 bytes" | 06:19 |
| MrPPS | seems like it theoretically could be, but then I try again a second or two later, and it works fine | 06:20 |
| MrPPS | oh, it is important to note | 06:20 |
| MrPPS | existing connections to it are staying connected | 06:20 |
| MrPPS | i.e., I have ZNC on there that connected a week ago, and that's still fine | 06:20 |
| MrPPS | but trying to set up a new connection, or manually with openssl, is giving that error | 06:20 |
| Takyoji | is there any sort of port forwarding or things going on? | 06:23 |
| MrPPS | none at all | 06:24 |
| MrPPS | these are all servers in DC's | 06:24 |
| MrPPS | with public v4's | 06:24 |
| Takyoji | Tried asking about your situation with a support channel of your IRCd? | 06:25 |
| Takyoji | http://serverfault.com/questions/389197/ssl-routinesssl23-writessl-handshake-failure | 06:26 |
| Takyoji | MrPPS, ^ | 06:27 |
| MrPPS | Takyoji: cheers, having a look at the serrver port article | 06:28 |
| MrPPS | haha, -tls1 flag isn't accepted for my client :/ | 06:30 |
| Arth | i'm trying to make it so that whern you go to the domain, instead of trying to connect to www.domain.com:80, it goes to another service(port) by default | 06:31 |
| Arth | lets say i had a service running at port 555, instead of going to the domain and it points to domain.com:80 i want it to go to domain.com:555 without me having to add :555 to the url | 06:31 |
| Arth | is that possible? without using javascript/meta redirects, doesn't seem to work for me | 06:32 |
| MrPPS | oh, never mind, it is picky about the flags order | 06:32 |
| Takyoji | Arth, that idea really only applies to SRV records, and in the context of HTTP, browsers don't check for SRV records, they only do A/AAAA records | 06:32 |
| MrPPS | If I was lazy, I'd just get socat to do port translation | 06:33 |
| MrPPS | :P | 06:33 |
| Arth | Takyoji: are you saying it's not possible | 06:33 |
| Takyoji | Pretty much. In DNS, an A or AAAA record just tells your computer which IP there is for the domain. | 06:34 |
| Takyoji | Whereas with SRV, it'll tell you which IP and port, for the service your requesting. And HTTP doesn't implement that idea. | 06:34 |
| Arth | thanks, so MrPPS how exactly would socat help me | 06:34 |
| Takyoji | I mean, not that it's HTTP itself, it's just browser vendors and such | 06:34 |
| MrPPS | Arth: for example; sudo socat TCP-LISTEN:555,fork TCP:localhost:80 | 06:35 |
| Arth | and that wouldn't actually affect the funcionality of the webserver itself would it? | 06:36 |
| MrPPS | actually, sorry, I did that the wrong way around | 06:36 |
| MrPPS | but it would essentially take control of that port | 06:37 |
| MrPPS | i.e. anything on port 80 would go to port 555 | 06:37 |
| MrPPS | I'm not entirely sure if that's what you're asking | 06:37 |
| MrPPS | but if you wanted *all* port 80 traffic going to port 555 transparently | 06:37 |
| MrPPS | that'd do it | 06:37 |
| Arth | leme see if i can clarify my question before i look into socat | 06:37 |
| MrPPS | no worries | 06:37 |
| Arth | So normally when you go to www.domain.com in your browser, "technically" it goes to www.domain.com:80 | 06:38 |
| MrPPS | yep | 06:38 |
| Arth | let's say i have service X running on port 555, how do i make it so that when i point my domain to www.domain.com, it goes to www.domain.com:555 without me appending :555 to the end | 06:38 |
| MrPPS | is that the only service that would be used? | 06:38 |
| Arth | don't really understand your question, kind of a long day, excuse me | 06:39 |
| Takyoji | Port 80 is assumed by default, unless specified otherwise | 06:39 |
| Arth | used where? | 06:39 |
| Arth | and yeah I understand it's assumed by default, is there no index workaround or something I can use? | 06:39 |
| MrPPS | Arth: I guess what I mean to say is, are you running a web server as well | 06:40 |
| MrPPS | or is the only intended purpose of this server to provide that service on port 555? | 06:40 |
| Arth | I know it seems like a trivial thing to type in :555 at the end but i'd just like to see if its possible | 06:40 |
| Takyoji | In the way that things are implemented, it's not quite possible. | 06:40 |
| Arth | well, from my knowlege, wouldn't you need a web server up to even be able to serve a page off a server when connecting from an external client? | 06:41 |
| Arth | yes, the only intended purpose of the server is to provide service on port 555 | 06:41 |
| Takyoji | because your ISP blocks port 80, or? | 06:41 |
| Arth | oh no not at all, this is hosted on a vps | 06:42 |
| Arth | im not hosting it, not a port blocking issue | 06:42 |
| Takyoji | So you're stuck to what port it's configured to? | 06:42 |
| MrPPS | Arth: if the only intended purpose is to serve what's on port 555, and port 555 is a HTTP compatible service (which is sounds like it is), then the command: sudo socat TCP-LISTEN:80,fork TCP:localhost:555 would work fine | 06:42 |
| Takyoji | since you don't have control to configure that? | 06:42 |
| MrPPS | but you'd need *nothing* else listening on port 80 | 06:43 |
| MrPPS | oh, wait, you're not hosting this? | 06:43 |
| Arth | not necessarily, i can configure it to whatever port I want, but wouldn't setting it to port 80 conflict with apache? | 06:43 |
| Arth | MrPPS: i'd need the apache server listening on port 80 to even serve me up the page right? so i don't think that will work | 06:44 |
| MrPPS | well, if the application itself (on port 555) is its own web server, you wouldn't need apache at all | 06:44 |
| MrPPS | if it runs on port 555 through apache | 06:44 |
| MrPPS | then you could just reconfigure apache to server that on port 80 | 06:44 |
| Arth | what's the command to turn apache off? sudo apache2ctl stop ? | 06:45 |
| Arth | slipping me | 06:45 |
| Arth | i'm gonna test if doing ip:555 with apache turned off will work | 06:46 |
| Takyoji | Because the thing is, you can have Apache do a reverse-proxy to the service at port 555, I believe | 06:46 |
| Takyoji | service apache2 stop | 06:46 |
| Takyoji | or apache2ctl may work | 06:46 |
| Arth | :o | 06:48 |
| Arth | doing serverip:555 with apache turned off brings up the page | 06:49 |
| Arth | so that means socat should work | 06:49 |
| Arth | i don't need apache | 06:49 |
| Arth | the sole purpose of the server is the service hosted at port 555 | 06:49 |
| Arth | MrPPS: gonna try socat and i'll let you know if it works | 06:50 |
| MrPPS | awesome :) | 06:51 |
| Arth | terminal hangs after sudo socat TCP-LISTEN:80,fork TCP:localhost:555 | 06:52 |
| Arth | *starts biting nails* | 06:53 |
| Arth | well i changed the port to the actual port of the service that was it | 06:53 |
| MrPPS | yeah, it just hangs, sitting there | 06:54 |
| MrPPS | it doesn't display anything :) | 06:54 |
| MrPPS | also, you may have to change TCP:localhost:555 to TCP:<public_IP>:555 | 06:54 |
| Arth | http://puu.sh/bt4dC/05ade2f72b.png | 06:54 |
| Arth | yeah by hangs i mean i'd need to do ctrl^C to be able to type in another command | 06:55 |
| Arth | well looks like ill need to do that and fix the public Ip thing | 06:55 |
| MrPPS | yeah, it doesn't fork into the background | 06:55 |
| MrPPS | you'd have to run it in screen/tmux | 06:55 |
| MrPPS | Arth: if that makes sense? | 06:56 |
| MrPPS | Takyoji: also, had a chat with devs | 06:58 |
| Takyoji | Wouldn't iptables be capable of just changing the destination port? | 06:58 |
| MrPPS | turns out | 06:58 |
| MrPPS | inspircd is crappy half the time, when you exceed maximum connections for the network | 06:58 |
| MrPPS | which is why I was getting the varied results | 06:58 |
| Takyoji | alrighty | 06:59 |
| MrPPS | so thanks anyway :) | 06:59 |
| MrPPS | and yeah, iptables can probably do it, but I've never had to | 06:59 |
| MrPPS | so I've just gotten used to socat for quick stuff | 06:59 |
| Takyoji | because I just figure in the long term, rather than having to maintain to stay running | 07:00 |
| MrPPS | yeah | 07:00 |
| MrPPS | better approach | 07:00 |
| Takyoji | maintain it to stay running* | 07:00 |
| micw | hi | 07:01 |
| micw | i have an ubuntu server that hangs during boot. on console 1 it shows that disks are mounted. on console 7 it shows that some services where starting/stopping | 07:01 |
| micw | but i cannot get a shell, sshd is not yet started | 07:01 |
| micw | i have no idea how to debug/recover | 07:01 |
| Takyoji | Restart in single-user mode (selected at GRUB)? | 07:02 |
| micw | "recovery mode"? | 07:02 |
| Takyoji | yes | 07:02 |
| micw | it boots to recovery menu | 07:03 |
| micw | when i select "resume" it's the same issue | 07:03 |
| Takyoji | and there should be an option to drop to a root shell, and not do a full boot | 07:03 |
| Takyoji | and then pick through any log files in /var/log | 07:03 |
| micw | it did not write logs yet | 07:04 |
| micw | seems that it did not boot so far | 07:04 |
| Takyoji | perhaps this may be of help for troubleshooting: https://wiki.ubuntu.com/DebuggingKernelBoot | 07:06 |
| micw | there was an error with rpcbind (the only error during boot process) - so i remove it and try again (don't need nfs on the system) | 07:06 |
| micw | the kernel boots fine | 07:07 |
| micw | init starts as well | 07:07 |
| micw | event network is set up (i can ping the machine) | 07:07 |
| micw | event -> even | 07:07 |
| === kickinz1|afk is now known as kickinz1 | ||
| micw | http://picpaste.com/Bildschirmfoto-VM_101_-_Mozilla_Firefox-lpm8e57r.png | 07:09 |
| micw | thats what console 7 shows | 07:09 |
| Takyoji | I suppose I don't have any specific ideas | 07:13 |
| micw | same to me. normaly i get such boot issues quickly fixed | 07:14 |
| lordievader | Good morning. | 07:31 |
| === kickinz1 is now known as kickinz1|afk | ||
| === kickinz1|afk is now known as kickinz1 | ||
| HackeMate | hello | 07:44 |
| HackeMate | if i send an email using smtp with TLS, and the destination server also uses TLS, is the email traveling unencripted? | 07:44 |
| === kickinz1 is now known as kickinz1|afk | ||
| === kickinz1|afk is now known as kickinz1 | ||
| === kickinz1 is now known as kickinz1|afk | ||
| === kickinz1|afk is now known as kickinz1 | ||
| Guest22304 | hi all | 10:03 |
| Guest22304 | i boot up 14.04 server livecd | 10:04 |
| Guest22304 | cannot get fdisk command to use, why? | 10:04 |
| lordievader | Guest22304: Should be there, according to the manifest util-linux is installed. | 10:08 |
| lordievader | Guest22304: Could you pastebin the output of "apt-cache policy util-linux"? | 10:08 |
| lordievader | !paste | 10:08 |
| ubottu | For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. | 10:08 |
| Guest22304 | I have just rebooted the livecd, but this time it is entering rescue mode ! | 10:17 |
| Guest22304 | under its shell, it did has the fdisk command | 10:17 |
| lordievader | Since when does the livecd has a rescue mode? Is it not the rescue mode of your physical install? | 10:18 |
| Guest22304 | that is true,! | 10:19 |
| lordievader | Guest22304: Ok, so what seems to be the problem? | 10:19 |
| Guest22304 | I am boot the live cd in normal mode | 10:19 |
| Guest22304 | fdisk command is not exist | 10:20 |
| lordievader | Guest22304: What is the problem you are facing and what steps have you taken to fix it? | 10:21 |
| * lordievader is getting confused | 10:21 | |
| Guest22304 | I boot the live cd and is going to install it, | 10:23 |
| Guest22304 | I don't like the menu method to create partitions | 10:24 |
| Guest22304 | I like command line to do so | 10:24 |
| lordievader | 10-12:08 < lordievader> Guest22304: Could you pastebin the output of "apt-cache policy util-linux"? | 10:24 |
| Guest22304 | so I need fdisk command to that work there | 10:24 |
| Guest22304 | the result is: /bin/sh: apt-cache: not found | 10:26 |
| lordievader | Guest22304: Well ain't that lovely. Guess the easiest is to just use the menu, or make an live-usb/cd of the full desktop iso. | 10:27 |
| Guest22304 | ok, I got it | 10:28 |
| === kickinz1 is now known as kickinz1|afk | ||
| Kunzem1987 | Good day people. I posted yesterday about a backup that i have whith 20 files all *.tar.bz2.aa to *.tar.bz2.at and one small *.par2 file which i could not open with windoze using multipar or 7zip. I realised today that i could tar jxfv *.tar.bz2.aa but not the following *.tar.bz2.ab file. I did some searching and found that i needed to cat the files together to make one big tar.bz2 file to extract. I'm sure most | 10:43 |
| Kunzem1987 | of you know this but i'm newbie with Ubuntu server. Just letting know how i got my problem fixed. | 10:43 |
| === kickinz1|afk is now known as kickinz1 | ||
| ciroingiro | hello everyone. Ubuntu server 12.04 LTS turns off after a time imprecise operation. Sometimes it turns off after 1h. and the other after 4 days. | 13:11 |
| ikonia | turns off as in shuts down ? | 13:11 |
| cfhowlett | ciroingiro, cron jobs? | 13:13 |
| ciroingiro | Excuse me for Enter Key, i have compare the log with normal shut down log and i have notate some partial log file. | 13:13 |
| ciroingiro | i have check crontab directly in /var | 13:14 |
| ikonia | can you just clarify it actually shutsdown/powers off ? | 13:15 |
| ciroingiro | i have substitute the action shut down to power button pressed with a only log and now i have a partial log that say the power button is pressed and immediately the server turns off. | 13:19 |
| ciroingiro | it turns off as a black out | 13:19 |
| ikonia | so you're saying the screen goes black until you press the power button ? | 13:19 |
| eutheria | aha | 13:21 |
| eutheria | ok not the same problem | 13:21 |
| ciroingiro | yes the screen is black, the server is poweroff | 13:22 |
| ciroingiro | is off | 13:22 |
| ikonia | are you sure the server is power off ? | 13:23 |
| ikonia | as if it's powered off why do you then press the power button to turn it off ? | 13:23 |
| ciroingiro | yes but i don't have idea of power off command | 13:23 |
| ikonia | you're not powering it off - so why do you care about the power off command | 13:23 |
| ikonia | the machine as I understand it is going to a black screen - you are then powering it off by pressing the power button | 13:24 |
| ikonia | or am I wrong ? | 13:24 |
| eutheria | wow grub failed while upgrading from 12.04 to 14.04 | 13:26 |
| ciroingiro | is a ibm server closed in a rack, a uman can not touch it bu i can operate from remote. When the connection is closed i see the server turned off, not in standby o screen standby. | 13:26 |
| ikonia | ciroingiro: I'm really sorry, I don't understand what you are saying | 13:26 |
| ikonia | ciroingiro: what is your native language ? | 13:26 |
| ciroingiro | excuse me my language is italian and not speek english very well. | 13:27 |
| ikonia | thats ok | 13:27 |
| ikonia | ciroingiro: if you join me in #ubuntu-it I'm sure someone can help translate | 13:28 |
| ciroingiro | ikonia ok i try #ubuntu-it. thank you very much . | 13:29 |
| ikonia | no problem | 13:29 |
| teward | rbasak: FYI: nginx-naxsi removed before jessie freeze - https://mail.google.com/mail/u/0/#inbox/145a57304822c977 | 14:51 |
| teward | oopsies | 14:51 |
| teward | rbasak: relevant bug and discussion - see response from Christos - https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=746199 | 14:52 |
| uvirtbot | Debian bug 746199 in src:nginx "Outdated naxsi version, incorrect learning tools included in packages" [Important,Open] | 14:52 |
| jamespage | zul, hallyn: afternoon chaps | 15:09 |
| hallyn | \o | 15:09 |
| jamespage | having a few issues with libvirt in utopic | 15:09 |
| zul | jamespage: its hallyn's fault :) | 15:10 |
| jamespage | hallyn, I've seen a crash - https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1367703 | 15:10 |
| jamespage | zul, ^^ | 15:10 |
| hallyn | impossible | 15:10 |
| jamespage | and lockups as well - virsh can't get a response and nova is locked out as well | 15:10 |
| jamespage | kinda limits instance management | 15:10 |
| jamespage | .... | 15:10 |
| zul | hallyn: looks like cgmanager :) | 15:10 |
| hallyn | jamespage: is cgmanager running? | 15:11 |
| jamespage | hallyn, yes | 15:11 |
| jamespage | hallyn, I have one hanging now if that's helpful | 15:13 |
| jamespage | rather than crashing | 15:13 |
| hallyn | jamespage: i has a guess that apparmor is preventing libvirt from talking to cgmanager. does 'cgm apiversion' work? | 15:13 |
| jamespage | hallyn, yes | 15:13 |
| jamespage | hallyn, nothing in the kern.log about DENIED | 15:14 |
| hallyn | jamespage: does 'unix' show up in /etc/apparmor.d/usr.sbin.libvirtd ? | 15:15 |
| * jdstrand notes we don't have an enforcing kernel for 'unix' | 15:16 | |
| hallyn | you're on the right version of libvirt, | 15:16 |
| jdstrand | (that will land soon) | 15:16 |
| jamespage | hallyn, it does | 15:16 |
| hallyn | hm | 15:16 |
| hallyn | jamespage: well the bug report says all your libvirt configs are updated. please attach those. ps -ef | grep libvirt, and cat /proc/pid/attr/current and /proc/pid/cgroups for libvirtd | 15:17 |
| hallyn | jamespage: so this is happening on >1 hosts? | 15:21 |
| jamespage | hallyn, so I've seen that crash once and the lockup a few times on multiple different hosts | 15:22 |
| hallyn | all stock utopic? | 15:24 |
| hallyn | now mind you we shouldn't be crashing when we can't connect to cgmanager | 15:26 |
| hallyn | so i'm not handling a DBusError or something somewhere | 15:26 |
| jamespage | hallyn, yes | 15:30 |
| sturmflut-work | Anybody here who runs Ubuntu on ARM64/ARMv8-A hardware? Not just the emulator, a real machine. | 15:38 |
| TurkerTunali1 | Hi guys, I've created a user when installing ubuntu server 14.04 (namely:ubuntu) . I want to ask you that does this user has root privileges? Because I have some strange behaviours. I have a script which installs Postgresql 9.3 along with our ERP and it works greate on AWS but my local machines can't connect Postgresql. | 15:47 |
| rbasak | TurkerTunali1: cloud images allow the ubuntu user to sudo with no password. | 15:48 |
| CarlFK | TurkerTunali1: the user created by the installer has sudo privs | 15:48 |
| rbasak | The default user on Ubuntu desktop (or traditional server) can also sudo, but only with a password by default. | 15:48 |
| rbasak | See /etc/sudoers.d/ in the different environments. | 15:48 |
| TurkerTunali1 | I am checking, I've edited (sudo visudo) to give my user sudo capabilities without password | 15:49 |
| CarlFK | hmm, prolly better to figure out why you can't connect to Postgresql | 15:49 |
| TurkerTunali1 | sudoers.d folder is empty | 15:51 |
| TurkerTunali1 | on my local machine | 15:51 |
| rbasak | Indeed. | 15:51 |
| CarlFK | same here (well, it includes a README) | 15:52 |
| TurkerTunali1 | AWS has 90-cloud-init-users file | 15:52 |
| TurkerTunali1 | AWS specified ubuntu user as NOPASSWD:ALL | 15:54 |
| TurkerTunali1 | I've given this permission with sudo visudo | 15:54 |
| TurkerTunali1 | on my local machine, so there must be something else | 15:54 |
| === kickinz1 is now known as kickinz1|afk | ||
| TurkerTunali1 | does ubuntu server 14.04 repos and AWS Ubuntu server repos are the same? | 16:00 |
| TurkerTunali1 | maybe they are downloading from different postgresql repos | 16:01 |
| CarlFK | TurkerTunali1: how is postgresql server being installed? (I am guessing a script that installs the whole erp system?) | 16:04 |
| RoyK | I have a wierd memory usage issue here on a vm that's running a java process. it's not ubuntu, but I don't think it's distro specific - ok to ask? | 16:09 |
| patdk-wk | easy solution | 16:12 |
| patdk-wk | uninstall java :) | 16:12 |
| CarlFK | lol | 16:12 |
| === Lcawte|Away is now known as Lcawte | ||
| hallyn | jamespage: on a fresh utopic install, i follow https://wiki.ubuntu.com/SergeHallyn_libvirtnest and cdboot starts fine, no cgmanager errors. | 16:13 |
| RoyK | patdk-wk: not my choice - confluence runs on it :P | 16:16 |
| RoyK | Anyone that knows what on earth is using memory here? http://pastebin.centos.org/12351/ (munin graph here http://karlsbakk.net/bilder/memuse.png) | 16:16 |
| RoyK | it's really confusing | 16:16 |
| patdk-wk | royk, likely your anonhugepages problem | 16:20 |
| patdk-wk | sometimes java likes it, and sometimes not | 16:21 |
| patdk-wk | it depends on the java app | 16:21 |
| RoyK | what are they? | 16:21 |
| patdk-wk | oracle had a nice writeup on that somewhere | 16:21 |
| patdk-wk | transparent huge pages? | 16:21 |
| patdk-wk | it's converting those 4k memory blocks to 2megs | 16:21 |
| RoyK | I just don't get it - the java app just uses 1GB, as it's told to | 16:21 |
| patdk-wk | likely how the app works, it keeps shuffling things around, and conflicts | 16:21 |
| patdk-wk | ya, but if it is getting allocated 2meg pages, but not using all 2megs | 16:22 |
| patdk-wk | cause it keeps randomly freeing parts of it | 16:22 |
| RoyK | As you can see on the graph, some time back, the VM just had 2GB memory. I gave it another 2GB, and it just grabbed it all | 16:22 |
| patdk-wk | cause of how it's memory allocation works | 16:22 |
| patdk-wk | jemalloc has this problem | 16:22 |
| RoyK | (without it showing up in the ps list etc) | 16:22 |
| RoyK | ok | 16:23 |
| patdk-wk | tricks to help malloc, cause other issues :) | 16:23 |
| patdk-wk | can't cross optimize :) | 16:23 |
| RoyK | do you know any resources about this? | 16:24 |
| RoyK | should be possible to fix it without giving each java-running vm 64 gigs of RAM :P | 16:24 |
| patdk-wk | https://blogs.oracle.com/linux/entry/performance_issues_with_transparent_huge | 16:26 |
| patdk-wk | well that is talking about performance | 16:26 |
| patdk-wk | man, ran across it like a month ago | 16:27 |
| Alina-malina | when i connect with ftp, why i get owner groul as www-data www-data? | 17:00 |
| RoyK | Alina-malina: the question is, why do you use ftp? ;) | 17:01 |
| RoyK | !ftp | 17:01 |
| ubottu | FTP clients: Nautilus (Places -> Connect to server), gFTP, FileZilla (for !GNOME); Konqueror, Kasablanca, KFTPGrabber (for !KDE); FireFTP (for Firefox); ftp, lftp (for !cli) - See also !FTPd | 17:01 |
| Alina-malina | RoyK, wel it was example | 17:01 |
| RoyK | ubottu: also, FTP is an outdated protocol, don't use it! Use sftp instead. | 17:01 |
| Alina-malina | RoyK, it is sftp | 17:03 |
| Alina-malina | so what you think? | 17:03 |
| Alina-malina | why it shows www-data www-data? | 17:03 |
| RoyK | Alina-malina: it should be the user/group of the connecting user/group | 17:05 |
| Alina-malina | yes is it ok? | 17:05 |
| RoyK | is it www-data connecting and creating files? | 17:05 |
| Alina-malina | yes | 17:05 |
| Alina-malina | probably | 17:05 |
| RoyK | then it's correct | 17:05 |
| RoyK | and there should be no 'probably's to sysadmining ;) | 17:06 |
| Alina-malina | because it is website, and from time to time it gives an error failed to open stream: Permission denied | 17:06 |
| Alina-malina | so what is wrong with that? | 17:06 |
| FrEaKmAn_ | hi all | 17:13 |
| FrEaKmAn_ | I have a VPS with root access to ubuntu 12.04. I have defined firewall rules with iptables | 17:13 |
| FrEaKmAn_ | for some reason, after few days there are no rules anymore | 17:13 |
| FrEaKmAn_ | my question is: when I define iptables rules, do I need to "flush" them to store? | 17:14 |
| FrEaKmAn_ | I don't think they reset after restart | 17:14 |
| RoyK | !ufw | FrEaKmAn_ | 17:19 |
| ubottu | FrEaKmAn_: Ubuntu, like any other Linux distribution, has built-in firewall capabilities. The firewall is managed using the 'ufw' command - see https://help.ubuntu.com/community/UFW | GUI frontends such as gufw and ufw-kde also exist. | An alternative to ufw is the 'iptables' command - See https://help.ubuntu.com/community/IptablesHowTo | 17:19 |
| FrEaKmAn_ | RoyK: thanks | 17:28 |
| === xar_ is now known as xar- | ||
| === phuh is now known as phuh_____ | ||
| === phuh_____ is now known as phuh | ||
| === cmagina_ is now known as cmagina | ||
| === Lcawte is now known as Lcawte|Away | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
