[00:03] ivoks you're awesome [00:03] i'm back up [00:04] where can i take a course on ubuntu or some trainning to know all that? [00:05] training.canonical.com [00:05] or #ubuntu-training [00:06] that would be a good first step [00:06] awesome, i'll take one soon [00:06] make that [00:06] http://www.ubuntu.com/support/training [00:07] i'm sure work wouldn't be to happy if this had of happened on monday morning instead of saturday [00:08] well, you were the all variables in equation.. so :) [00:08] yuuuup [00:08] one quick question for you before you go? [00:10] do i need to turn on a gui in order to log into kvm for the first time? i tried putting the openssh-server in the --addpkg but it never seemed to work for me unless i messed up the bridged network config [00:13] maybe not so quick lol [00:18] thanks a million ivoks, i owe you a beer or 10 [00:21] nope [00:22] ah, he left [00:22] didn't have a change to mention vnc :) [00:22] chance [00:29] anyone knows how to configure compiz ? [00:30] stupid question to ask here, I know [00:46] im upgrading ubuntu to latest version it now asking if I want to keep localy modified files or install package maintainers version? What should I do? [00:47] depends on what you've done to the local fiels [00:47] ok ill keep the current versions [00:47] I did that and kept my locally modified files so I didn't lose my samba config [00:48] god it goes on this upgrade [00:48] hope the whole thing doesnt come crashing down [00:48] I did that and lost a very customized samba config [00:48] Now if only I could get my NFS working, I'd be happy [00:59] system is rebooting. Fingers crossed [00:59] touching wood [01:13] It works.... well website does an stuff also fixed my previous date issue today [01:13] good night [01:13] Thanks ivoks [02:26] anyone have a working extables-addons-source ? mien gives error messages every time I run m-a a-i xtables-addons-source :( [03:16] does someone know something about monitor authentications ? [03:47] what is the default directory for the imap.conf, and smtpd.conf ? [04:00] I have a pair of Apache2 sites (www, www-other). How do I restrict the BackupPC alias (in conf.d) so that /backuppc only exists in www? === ktk_ is now known as KurtKraut [05:46] Hello [05:46] Anyone know of a twitter CLI app that actually works? Trying to through it into a script for an auto update... [05:46] The ones i have tried give me a 401 unauth error [05:46] the pass is correct with the username [05:47] Starhero: are you sure? 99% of times I have thought the app was broken it turned out I had done something wrong [05:48] recheck the password / username [05:48] I have quite a few times [05:48] What do you use? I'll do more testing on what app you speak of... [05:49] even gotweet isnt working [05:50] Starhero: the app I referred too was a general app not a specifc - as in when one has a problem with a given app it is generally down to user error [05:50] sorry I should have been clearer [05:52] hmm I trhink maybe it is because something called Oauth isn't installed....I have see referances to it ...but i think that is an embeded class or something i am not sure [07:06] I have a raid 5 setup using mdadm. My server has been on for a few days with little activity. All of the sudden I noticed the sound of my hard drives working a lot more then they should. I checked the state of the raid 5 and it says "clean, recovering" and has a rebuild status of 31%. this raid has been built and active for many months, what does this behavior mean? [07:07] listed under "Active Devices : 4, Working Devices : 4, and Failed Devices : 0" [07:09] whats the uptime of your server cppisweird? [07:09] is it possible it has crashed or had a power failure ? [07:28] no, it does not boot back up after a power failure [07:29] uptime is 4 days [08:12] CppIsWeird, check your smart status, there should be something in there. [08:21] anyone familiar with taskset? [08:22] i split two processes across two different cores, but only one is crunching while other one is idle? [09:16] anyone familiar with octopussy (http://www.8pussy.org/)? [09:28] register [09:41] New bug: #685426 in tftp-hpa (main) "package tftpd-hpa 5.0-11ubuntu2 failed to install/upgrade: el subproceso script post-removal instalado devolvió el código de salida de error 127" [Undecided,New] https://launchpad.net/bugs/685426 [09:56] Hello, I have a ubuntu 10.04 64 bit server which is used to server as much as 40Million web requests per day [09:57] the request ranges from all over the world and ofcourse involved millions of different IP addresses [09:57] I am not sure how to best configure the server as far as IP conncetions setting is involved [09:58] sunday is perfect timing for that question :) [09:59] are there any problems with current setup? [09:59] Sparsh: ^ [10:00] there are some problems btu I am not sure if they are because of this issue [10:00] last night we had 504s [10:00] and a lot of connections were refused [10:01] so I guess we reached the saturation and our server wasnt accepting more connections from IPs [10:01] but I dont know where to start investigation and fix [10:01] web server is? [10:02] yea. it got Varnish Cache on it [10:02] only varnish and pound cache on it.. with backend connected with 5 apache servers on a round robin basis [10:02] so... http error messages are above ip [10:03] if you get http error messages, then your ip communication is ok [10:03] possible SYN flooding on port 80. Sending cookies [10:03] this is one of the error we got [10:03] you'll get those all the time [10:03] and also our monitoring scripts reported 504s (Gateway Timeout) [10:03] what does this error mean [10:03] nothing to worry about unless you get it bunch of times and no one can access the web [10:04] let's start all over again [10:04] hmm.. but how to check the IP tables / IP connections related configuration [10:04] sure [10:04] we are talking about varnish or apache machine? [10:04] varnish machine to start with [10:05] so, that machine reports 504 when accessing apache? [10:06] Yes, [10:06] then you have a problem [10:06] http://monitor.wingify.com/munin/visualwebsiteoptimizer.com/varnish.visualwebsiteoptimizer.com.html#Varnish [10:06] this is a munin graph [10:06] there were a lot f backend conn failures [10:08] yes [10:08] and you have a very bad cache hit/miss ratio [10:08] all we know is something went very wrong yesterday.. but what exactly is something I have jst started investigation [10:09] its usually around 70 [10:09] because we ourselves force pass to a few web requests [10:09] our config demands hit for only one specific type of request [10:09] well, have you seen your 'connection through firewall'? [10:09] how exactly to see that [10:10] http://monitor.wingify.com/munin/visualwebsiteoptimizer.com/varnish.visualwebsiteoptimizer.com-fw_conntrack.html [10:10] ohh sorry.. [10:10] the big chunk of data missing is because we switched the website to our backup server cluster [10:11] i have a website with half the traffic you get [10:11] when we realized that something is going bad with it [10:11] but with only 2 apaches [10:11] do you also use some sort of caching system ? [10:11] for us response times is the most critical thing. [10:11] and with bigger cache hit/miss ratio [10:12] i use varnish too [10:12] okay, we have recently moved to varnish and still trying to learn [10:12] 2 days back we moved completely from nginx based load balancing to varnish front [10:12] it takes time [10:12] was there a reason for that? [10:13] was nginx failing? :) [10:13] well nginx was sending request to apache servers [10:13] then we installed varnihs [10:13] and started routing traffic via varnish [10:14] so it was nginx->varnish->apache which is a waste.. so we removed nginx and let varnish face the traffic [10:14] ok [10:14] and suddenly we realized http://monitor.wingify.com/munin/visualwebsiteoptimizer.com/varnish.visualwebsiteoptimizer.com-fw_forwarded_local.html has shot up [10:14] and still increasing [10:15] and I wonder we arent doing something correctly.. maybe ip connections are active for a lot longer then required [10:15] there's no single anwser for this setups [10:15] every setup is customized to your needs [10:16] http://paste.org/pastebin/view/25768 [10:16] this isn't unusal graph [10:16] this is the varnish settings we have at the moment [10:16] on my less frequent web, it's ~18k [10:17] hmm [10:18] so, when you used nginx [10:18] i guess your servers had public ips? [10:19] nginx wasn't doing nating, right? [10:19] yes [10:19] it was again a very basic kinda default setup and there wasnt any NATing [10:20] and now there is NATing? [10:20] varnish has public ip, while apaches have private ips [10:21] that explains ipconntrack [10:21] if you don't do NAT, then there aren't connections to track [10:21] hmm.. sorry but I am not exactly aware of this [10:21] is ipconntrack bad ? [10:21] no [10:21] what exactly is it [10:21] that's a normal thing [10:22] when your machines does NAT, it tracks the connections [10:22] if there is any efficiency issues here [10:23] looking at your graph, you don't have that many connections [10:23] you have 10x less than i do [10:23] well during peak its around 400 req/sec [10:24] i'm talking about ipconntrack [10:24] okay [10:24] your peek is 2k [10:24] which is... low [10:24] okay. [10:24] anything else which is unusual in the graphs which you can spot [10:25] or any other variable / config on the server which might be important to the type of setup we have [10:26] no, everythin looks fine [10:26] you should look at your apache servers [10:26] one thing i'm missing on this graphs [10:27] if varnish server is doing NAT, why do you have only eth0 [10:27] what does it mean [10:27] you know what NAT is? [10:28] I have some idea YES, but ofcourse wont say I know it all [10:28] so, if you have private network on eth0 [10:28] your kernel does NAT to eth1 [10:28] or eth1 to eth0 [10:28] http://en.wikipedia.org/wiki/Network_address_translation [10:29] or you don't do NAT? [10:29] my /etc/network/interfaces [10:29] does your varnish server have a public ip? [10:29] auto lo [10:29] iface lo inet loopback [10:29] auto eth0 eth0:1 [10:29] iface eth0 inet dhcp [10:29] iface eth0:1 inet static [10:29] address 192.168.162.141 [10:29] netmask 255.255.128.0 [10:30] don't paste here [10:30] and what address does it get from dhcp? public or private? [10:30] public [10:31] But I would want to confirm [10:31] how can I do that [10:31] any particular commands I should run on the server [10:31] ip a [10:32] if it's a public, then you have an ok design for small office, but a bad one for serious web sites [10:32] http://paste.org/pastebin/view/25770 [10:33] consider this as a serious website [10:33] then get your self additional ethernet card [10:33] and ? [10:33] connect your private machines to isolated, private network [10:34] don't have them all on the same swith (public and private) [10:34] okay [10:34] right now we have a hosting on Linode [10:34] that's a basic thing to do [10:34] okay [10:34] this all is on linode? [10:34] yes [10:34] around 12 servers on linode [10:34] we are thinking of moving dedicated [10:34] then i must say i'm surprised it works so well :) [10:35] 40 milion visitors per month... [10:35] that's 10x my country has citizens :D [10:35] haha, where are you from [10:35] croatia [10:36] anyway... separate private from public [10:36] well we have struggled to get this setup working but we are struggling even more now [10:36] okay [10:36] linode hosts virtual machines [10:36] yes [10:36] so you have additional overhead there [10:36] hmm.. [10:36] i use linode too, but not for stuff like this [10:36] for stuff like this we use, almost, dedicated DC :) [10:37] hmm [10:37] linode is awesome, but you have to be aware of its limitations [10:38] i doubt it can handle that many requests [10:38] yea i guess now we have started realizing the limitations [10:39] If you odnt mind, can I add you on skype. YOu have been very very helpful to me. I am working with wingify which develops visualwebsiteoptimizer.com and this all setup is for this website [10:39] we have some script which gets added to clients website which this setup sends. [10:40] well, we can stay on IRC and I can help you when i have time [10:40] or... we can go skype, but that will become serious, not pro bono :) [10:40] sure whatever you prefer [10:40] *not pro bono* [10:41] http://en.wikipedia.org/wiki/Pro_bono [10:41] :) [10:41] aah :) [10:49] HI I upgraded ubuntu to latest version last night but now my LTSP doesnt work no more :( [10:50] :) [10:50] hi ivoks you still here [10:50] well its broken my system [10:51] any idea how I can get this fixed? [10:51] yes, for a minute... i'll be out for next 10 [10:51] Website works etc which is a plus, but now nobody in the office can log into the LTSP side of the server [10:51] :( [10:51] do you know what's the problem? [10:52] on a tech level, not on a 'user' level [10:52] no [10:53] all i know is when we turn a Computer on that runs off the LTSP it now says BOOT DISK FAILURE [10:53] which means it isnt booting from the LTSP server no more [10:53] normaly get that when the cable is unplugged [10:53] ok [10:53] that's a good start [10:53] i guess ltsp boots from nfs? [10:53] its like a year since I touched it.... n [10:54] yes i think so, that rings a bell [10:54] check if nfs is running on your server [10:55] how [10:55] ps ax | grep nfs [10:55] 3910 pts/0 S+ 0:00 grep nfs [10:55] so, it's not [10:55] EEeeeek [10:55] going out for a smoke [10:55] we'll solve this when i get back [10:56] ok thanks [10:56] your a saint [10:56] :S [11:08] i don't believe in ferrys :) [11:08] i beleive in fairys [11:08] :) [11:08] or that :) [11:08] nice smoke? [11:08] fairies [11:08] hi folks, im trying to set up a tf2 server, and when doing so ive started and stopped the program a few times, but it appears it hasn't stopped properly, as according to ps it is still running, however when i try to kill it via the PID it doesnt seem to work, can someone help ? [11:09] garymc: ok, let me install nfs, to see how it starts and stops [11:10] i don't have lucid with nfs yet :/ [11:10] ok [11:10] garymc: ok does /etc/init.d/nfs-kernel-server restart helps? [11:11] let me try [11:11] no such file or directory [11:11] /etc/init.d/nfs-kernel-server restart [11:12] no such file or directory [11:12] is there a file /etc/exports [11:13] nope [11:14] could it be anything to do with ethup etc? [11:14] hm... ltsp doesn't use nfs [11:14] ok.... [11:14] it's using nbd [11:15] ok.. [11:17] so, how does the problem look like on your think client? [11:17] you power it on [11:17] it starts booting from network [11:17] get's the ip and then starts booting? [11:17] normaly yes [11:18] but now says BOOT DISK FAILURE [11:18] before even getting to boot from network? [11:18] yes [11:18] or it tries and fails [11:18] im not at the office Im being told this over the phone [11:18] :S [11:19] ok [11:19] let's assume tftpd or dhcp isn't working right [11:19] ok [11:19] ps ax | grep tftp [11:19] stgraber: around? [11:19] ax command not found [11:20] garymc: i said 'ps ax' [11:20] dont think he/she is as I posted this in the LTSP channel no response [11:21] 1359 ? Ss 0:00 /usr/sbin/in.tftpd --listen --user tftp --address 0.0.0.0:69 /srv/tftp 3941 pts/0 S+ 0:00 grep tftp [11:22] ok, so that works [11:22] ok [11:22] ps ax | grep dhcp [11:23] should I pastebin this one [11:23] i best had [11:23] http://pastebin.ca/2011070 [11:23] so, that's running too [11:23] ok [11:24] then you can't get BOOT FAILURE [11:24] im thinking maybe I should have updated through the GNOME and not the Terminal [11:24] *upgraded [11:25] you would get that message on a PC if your boot devices do not exist [11:25] and clearly, pxe boot does [11:25] if there's an error in pxe boot, it would look totally different [11:25] could the netwrok switch be faulty in the office? [11:25] check your /var/log/syslog [11:25] look for DHCP entries [11:25] see if there are errors or odd behavior [11:27] look with me :S http://pastebin.ca/2011075 [11:28] so, it works [11:29] you might want to forward your email for root to the user [11:29] yeah I remeber seeing dovecot failures last night when it was installing [11:30] system is trying to send you an email [11:30] maybe it's related to the problem [11:30] how do I fix that then? [11:30] in /etc/aliases [11:30] add root: your_username [11:31] and then run newaliases [11:31] ok so at command prompt type: [11:31] root : garymc [11:32] newaliases [11:33] none of them commands are working [11:34] i said: [11:34] root: garymc [11:34] it's not a command [11:34] it's something you type in [11:34] in /etc/aliases.db [11:34] in /etc/aliases [11:34] and save [11:35] ok ive added that lin in aliases "root:garymc [11:35] yes [11:35] newaliases [11:35] then run newaliases [11:36] ok done [11:37] so what does that do? [11:37] it will forward email for root to you [11:37] do all the computers have that error BOOT FAILURE [11:37] or just one? [11:38] all of em [11:38] did they try powering it off and back on (computer, not server)? [11:38] yes [11:38] im gonna get them to do it again [11:38] hm [11:39] have you been playing with firewall? [11:39] on server [11:39] i aint done nothing but upgraded last night [11:39] ok [11:40] install tftp-hpa [11:41] apt-get tftp-hpa ? [11:41] apt-get install tftp-hpa [11:41] doing it [11:41] done [11:41] it said selcecting previously deselected package [11:45] should i get the terminals tested again? [11:46] no [11:46] do: [11:46] tftp localhost [11:47] yes [11:47] im there [11:47] status [11:47] does it say Connected to localhost.localdomain [11:47] ? [11:48] Connected to localhost [11:48] ok [11:48] Mode: netascii Verbose: off Tracing: off Literal: off [11:49] Rexmt-interval: 5 seconds, Max-timeout: 25 seconds [11:50] ok, quit [11:50] ok [11:50] go to /var/lib/tftpboot/ltsp [11:50] is there anything in it? [11:51] my guess - not [11:51] yes [11:51] there is another folder [11:52] i386 [11:52] in there there is pxelinux and aload of files [11:52] ok [11:53] mkdir /root/ltsp-backup [11:53] ok [11:53] mv /var/lib/tftpboot/ltsp/* /root/lts-backup/ [11:53] sudo ltsp-build-client [11:54] NOTE: Root directory /opt/ltsp/i386 already exists, this will lead to problems, [11:54] please remove it before trying again. Exiting. [11:54] error: LTSP client installation ended abnormally [11:54] mkdir /root/ltsp-backup-opt [11:54] mv /opt/ltsp/i386 /root/ltsp-backup-opt/ [11:55] sudo ltsp-build-client [11:56] ok its doing something [11:56] yes, that will take some time [11:56] it will build new OS for the clients [11:57] ok [11:57] I hope this works [11:57] will my users still be there with there files etc? [11:58] no clue :) [11:58] :S [11:58] but you backed up your old config, so... [11:58] ok [11:58] did i.....:S [11:58] yeah [11:59] go to /root/ltsp-backup-opt/ [11:59] there should be home subdirectory in it [11:59] and there should be all the files from your users [12:00] are they? [12:00] i cant find it [12:01] well go to /root/ltsp-backup-opt/ [12:01] what do you have there? [12:01] i386, right? [12:02] i cant find root/ltsp-backup-opt/ [12:02] /root/ltsp-backup-opt/ [12:02] you created it [12:02] shit found them [12:02] whoops sorry for swearing [12:02] yes its there [12:02] how many users? [12:03] I cant find the users [12:03] in home [12:03] how many directories? [12:04] 10? 100? [12:04] 1000? [12:04] hold on [12:04] i have to go soon [12:04] none :S [12:04] no users in home [12:04] 12:02 < garymc> shit found them [12:04] did you found them or not? [12:05] ooooh shit, no I found the backup folder [12:05] no users inhome [12:05] we are talking about backup folder [12:05] yes [12:05] no users in the home part [12:05] problem is that you don't have a clue about what you are doing [12:05] go to /root/ltsp-backup-opt/i386/home [12:06] type in ls [12:06] are there any directories? [12:06] no. none :( [12:06] no files [12:06] no directories [12:06] so, what did you find then? [12:07] when you said you found them [12:07] nothing an empty folder [12:07] I found the backup folder. [12:07] ltsp-backup-opt [12:07] ok [12:08] its still rebuilding [12:08] i take it its gonna take a while [12:10] this isnt gonna mess up my Apache web server on here is it? [12:10] no :) [12:10] ok good [12:11] you have that thin client manager, right? [12:11] what you mean? [12:11] http://doc.ubuntu.com/edubuntu/edubuntu/handbook/C/ltsp-tcm.html [12:12] ltsp installation completed [12:13] i don't see where it pulls home from :/ [12:13] yes, but I can only acces that when im at the ofice [12:14] AShh all my users are still there [12:14] its in /home [12:15] ok rebuild is done [12:15] should I ge tit tested at office? [12:15] yes [12:15] will do [12:16] and stay with them at the phone [12:16] try to figure out if the machine is actually looking for network booting [12:17] it should be they are set to check network boot first [12:21] disk boot failure still [12:21] should i reboot server? [12:22] you can try, but that shouldn't solve anything [12:23] i have to go now [12:23] good luck with booting your clients [12:24] ok thanks [12:25] you need to go to the office and see what's really going on [12:26] booting problems is easy to debug [12:26] but you have to know where the problem is happening [12:26] simple line BOOT FAILURE doesn't say much [13:29] can someone tell me if there is a more permenent solution these days when running dhcp3 server on a bridge , from the side of dhcp3 [13:30] atm i'm using a post-up pre-down script on br0 [13:30] but it doesn't seem the way to go tbh [13:39] ivoks : Im gonna go to office verry soon. ill be back on irc when I get there if your about would be great. [13:43] everyone gone? [14:01] "FATAL: introduce_user() loop detected" can someone help me out with this little issue that I'm having ? [14:01] I'm able to get anope connected to my unreal ircd but it is giving me that error. [14:06] New bug: #685505 in bind9 (main) "package bind9 1:9.7.0.dfsg.P1-1ubuntu0.1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/685505 [14:21] New bug: #685514 in net-snmp (main) "snmpd doesn't run as root by default - it should" [Undecided,New] https://launchpad.net/bugs/685514 [14:31] hello to all. i have big problems after the upgrade to maveric with the memcached service sometime this line in php $memcache->set("MyVar", "MyContent", false, 0); stores the value other times it dont store it even the variable dont exist in memcache [15:19] Hello, [15:19] I have a machine here that does not have "vmx" in the flags portion of cpuinfo. Is it still possible to run kvm even if it's slow, I want to test out kvm before installing it on a proper server. [15:28] hello to all. i am using this script here on a ubuntu maveric server. normaly this script should output the success message but it does not. what is wrong with memcached in ubuntu ? before the upgrade it worked everything well [15:28] http://paste-bin.com/view/raw/6b16c45d [15:29] does this script works okay on your side ? [15:44] garymc: ok :) [15:46] Is there a single virtual machine image that I can load on my desktop to test out UEC locally? Trying to set it up on a single image so far has been a pain. [15:46] Hi Ivoks [15:46] I managed to fix it by my self :) [15:47] with the help of google [15:47] I changed a setting in a file rebooted server an HEY PRESTO [15:48] nice [15:48] finally learning :) [15:48] ivoks : I did this https://bugs.launchpad.net/ubuntu/+source/ltsp/+bug/573975 [15:48] Launchpad bug 573975 in tftp-hpa "pxe image fails to boot: "Forbidden directory" (dup-of: 531123)" [Medium,Confirmed] [15:48] Launchpad bug 531123 in tftp-hpa "pxe boot with tftpd-hpa failed with error: Forbidden Directory" [High,Fix released] [15:48] take a look [15:49] excellent [15:49] you found a bug [15:49] I dont knwo did i? [15:49] Anyway -- there's a lot of setup that goes into setting up UEC and it would be really nice to have an easy-to-try virtual appliance available for testing and development. [15:49] so, you changed /etc/default/tftpd-hpa? [15:50] yes [15:51] was i on the right track [15:51] it was a pot luck go [15:52] hm... [15:52] can you paste it on pastebin.ubuntu.com [15:58] paste what? [15:58] ivoks [15:59] content of /etc/default/tfpd-hpa [16:00] yeah no probs [16:01] and paste the http link here [16:01] ivoks : http://pastebin.ca/2011217 [16:01] I added --secure [16:01] without --secure it doesn't work? [16:03] well it didnt [16:03] News back after adding it it worked [16:03] ok [16:04] Have to see for myself when i goto office in morning [16:04] i guess ltsp requires secure connection to tftp [16:04] i'd like to hear reasons for that :) [16:04] can they log in and use computers? [16:05] there was one person left at office and they could login and use the computer [16:05] after i added secure [16:06] ok then [16:06] all files in place? [16:07] yes seem to be [16:07] Ill know for sure in the morning [16:07] awesome [16:07] all your problems are gone [16:07] make sure you maintain that server [16:07] apt-get update ; apt-get dist-upgrade [16:07] that for updating security fixes [16:07] yes i will [16:08] cheers === 5EXABVMZK is now known as cokegen [16:26] could it be that php in ubuntu maverick is broken for memcache ? anybody here who is able to save values in memcache over php. i am using this test script here on my ubuntu maverick server and it does not work! => http://paste-bin.com/view/raw/ffe58b69 [16:26] it should however as it worked before the upgrade === doko__ is now known as doko [16:34] hey all [16:35] anyone know how to connect to ubuntu ec2 instance with out the default ubuntu username? [16:36] xperia: where does it fail? [16:37] ivoks: i have used this howto here https://wincent.com/wiki/testing_memcached_with_telnet to test memcached ovet telnet and it works great in the console. all values are stored like it should be [16:38] if i use however this php script that worked before the upgrade however none of the Values are stored in memcache http://paste-bin.com/view/raw/ffe58b69 [16:38] xperia: so, again, where does it fail? this script prints different errors depending on where it fails [16:38] the line $memcache->set('MyTimeStamp', $timestamp, 0, 0) does not works for some reason in php after the ubuntu upgrade [16:39] i do not get the print("SUCESS3"); line and becouse of this no value is somehow stored after the upgrade. for me it looks like php memcache is broken in maverick [16:41] SUCESS1
PHP Notice: MemcachePool::delete(): Server 127.0.0.1 (tcp 11211, udp 0) failed with: CLIENT_ERROR bad command line format. Usage: delete [noreply] [16:41] this is what i get with this script [16:42] problem with this script [16:42] ivoks hmm how did you get this additional debug messages. i just get outputed SUCESS1 and Deleted here [16:42] is that it dies if MyTimeStamp doesn't exist [16:43] with php-cli [16:43] when i remove: [16:43] if($memcache->delete('MyTimeStamp')) print("SUCESS2
"); [16:43] i get: [16:43] SUCESS1
Deleted
SUCESS31291567348
[16:43] then again: [16:44] SUCESS1
Not Deleted
FILETIME = 1291567348SUCESS31291567397
[16:44] ivoks woow you are great [16:44] man you solved the problem big compliments but really strange that it does die if you want delete something that do not exist. this was not allways that way [16:45] is this a bug ? i guess yes ! [16:45] must be in package php5-memcache [16:46] don't know [16:46] i hate php [16:46] this connection breaks to memcache in the php script make sure a lot of scripts non functional i guess [16:47] ivoks really ? woow how do you then programm websites ithout php ? [16:47] i think logic in this script is bad [16:47] deleting an object without checking if it exists... [16:49] okay i will fix it here then. ivoks i thank you really for this great help. you are awesome man ! [16:49] maybe [noreply] would help [16:49] Usage: delete [noreply] [16:49] i guess it wouldn't fail then [16:49] np [16:51] ivoks this feature does not exist in the package php5-memcache => http://www.php.net/manual/de/memcache.delete.php [16:51] ehh sorry => http://www.php.net/manual/en/memcache.delete.php [16:51] my german is rusty [16:51] geoip sorry [16:52] here it show full english even of the geop [16:52] well what we can say that php breaks memcache 100% when you delete something that does not exist [16:53] will just inform the php group about this [16:53] don't know... but that's where you are [16:57] ivoks: I'm around now ;) [16:57] stgraber: too late :D [16:57] stgraber: we had an issue with ltsp client not booting [16:58] stgraber: turns out tftpd did do 'secure' [16:58] stgraber: is there a reason why client would require secured connection? :) [16:58] tftp and secure in the same sentence always sound a bit weird ;) [16:58] yeah [17:01] anyway [17:02] guy added --secure to tftpd and things worked out [17:04] it was an update from karmic to lucid [17:11] hi anyone have any exp with ubuntu on ec2? [17:11] ssh to ubuntu@ec2instance works [17:12] but trying ssh newuser@ec2 does not work, even when pointing to the key file. for example: === lavish_ is now known as lavish [18:42] Hello folks :) [18:46] New bug: #685590 in dovecot (main) "dovecot 1.2.9 not compatible with thunderbird" [Undecided,New] https://launchpad.net/bugs/685590 [18:52] hey folks [18:53] I installed kernel sources but can't find the build directroy - any idea where it gets installed? [19:05] SpaceBass: I think the source package just installs a tarball. try dpkg -L linux-source-(version) [19:08] shauno, thanks [19:08] found the tarball [19:09] the package Im trying to compile still insists the kernel source doesnt exist, even when I give it the explicate path.... think its the package [19:10] space bass, as in the fish or the sound? [19:11] hello to all. i have upgraded my ubuntu server but for some strange reason apache or php are searching now files in the directory /usr/share [19:13] where is this strange path defined to search my files there [19:13] as a examle if i call this url here => http://wificom.ch/javascript/test.js [19:13] i get this strange apache error [19:13] [Sun Dec 05 20:05:03 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.js [19:13] together with the message file can be not found [19:27] AndyGraybeal, bass as in bass, not bass [19:28] xperia_, it sounds like the web root has been changed [19:28] xperia_, whats defined in /etc/apache/config [19:33] SpaceBass: will just look at it. maybe /etc/hosts was changed too after upgrade [19:33] SpaceBass: thanks for clearing that up. [19:33] AndyGraybeal, :D - its bass as in the insturment [19:33] :) [19:34] i liked the dry british styled humor. [19:37] SpaceBass: i have nothing with webroot in /etc/apache2/apache2.conf [19:39] xperia_, sorry, its documentroot and its defined in /etc/apache2/sites-available/default === squishy is now known as SquishyNotHere [19:52] SpaceBass: http://paste-bin.com/view/a7f68601 [19:54] it is set right and i am able to call the files in this document root but calling files in sub directory lead allways to the message file could be not find and i get the error message in apache error log [19:54] [Sun Dec 05 20:05:03 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.js [19:54] the file http://wificom.ch/javascript/test.js exist however [19:55] it is heavy strange and i dont know what exactly really the problem is [20:00] SpaceBass: example => http://wificom.ch/test.html works ! => http://wificom.ch/javascript/test.html does not work [20:00] File exist however ! => ls -la /var/www/wificom/javascript [20:00] -rw-r--r-- 1 root root 81 2010-12-05 20:04 test.html [20:00] any help for this heavy strange error. before the upgrade everything worked like it should ! [20:01] looks like apache search allways in usr/share when it comes to subdirs === xperia_ is now known as xperia [20:18] you again [20:19] ivoks yeah its me :-) have problems after server upgrade :-) [20:19] what's up? [20:19] anybody good with WAN here ? [20:19] ivoks: well as said this url => http://wificom.ch/test.html works ! this url however => http://wificom.ch/javascript/test.html does not work [20:20] i get allways the error message [Sun Dec 05 20:05:03 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.js [20:20] You don't have permission to access /javascript/ on this server. [20:20] ok [20:20] the file exist however and path is right => ls -la /var/www/wificom/javascript [20:20] -rw-r--r-- 1 root root 81 2010-12-05 20:04 test.html [20:20] is there an .htaccess file in /var/www/wificom/javascript [20:21] i have changed allready permissions but it does not help as apache looks for some strange reason in /usr&share instead of the document root directory http://paste-bin.com/view/a7f68601 [20:22] ivoks: .htaccess file does not exist in /var/www/wificom anywhere [20:22] there's never a strange reason [20:23] are you sure it's looking in /usr/share? [20:23] you pasted a link to ..javascript/test.html [20:24] and then pasted the error for javascript/test.js [20:25] ivoks: based on the error log it apache2 look to find this files in /usr/share [20:25] [Sun Dec 05 20:05:03 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.js [20:25] it's test.js [20:25] but the link is test.html [20:25] well okay test.html exist also but dont load also [20:25] therefor, i'm sure you have rewrite somewhere, with broken rules [20:25] do you see test.html in error log? [20:26] tail your log now [20:26] i've asked for nonexistingfile.html [20:26] do you see that in error log? [20:27] ivoks: okay here are the files http://paste-bin.com/view/a271382e [20:27] and here is error log [20:27] [Sun Dec 05 20:04:39 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.html [20:27] [Sun Dec 05 20:05:03 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/test.js [20:28] ok [20:28] how about nonexistingfile.html? [20:28] http://wificom.ch/javascript/nonexistingfile.html [20:28] in your logs [20:28] [Sun Dec 05 21:28:26 2010] [error] [client 192.168.1.1] File does not exist: /usr/share/javascript/nonexistingfile.html [20:28] [Sun Dec 05 21:28:29 2010] [error] [client 93.138.88.10] File does not exist: /usr/share/javascript/nonexistingfile.html [20:29] ok [20:29] last ip is probably yours [20:29] and there's no /var/www/wificom/.htaccess? [20:29] it drives me crazy this problem most of my ites are broken but they worked without any problems before the upgrade [20:29] most of your sites? [20:30] so, not just this one? [20:30] then go to /etc/apache2 [20:30] and run: [20:30] grep -sr javascript * [20:31] ivoks: all of my vhost sites on this server have this problem [20:31] ls -la /var/www/wificom/ [20:31] drwxr-xr-x 3 root root 4096 2010-12-05 20:57 . [20:31] drwxrwxrwx 26 root root 4096 2010-12-04 18:01 .. [20:31] -rw-r--r-- 1 root root 0 2010-08-13 21:03 index.html [20:31] drwxr-xr-x 2 root root 4096 2010-12-05 20:06 javascript [20:31] -rw-r--r-- 1 root root 81 2010-12-05 20:57 test.html [20:31] so no .htacces file anywhere [20:31] if all have problem, then your system wide config is broken [20:31] go to /etc/apache2 [20:31] vhost works but as you try to access subdir nothing is loaded [20:32] ivoks okay i am in /etc/apache2 [20:32] grep -sr javascript * [20:33] I've just finished migrating my home server. Thing is, I get dropped to an (initramfs) prompt because my software raid-10 partition backing my lvm volume group isn't started automatically. I simply need to mdadm --assemble my raid partitions and then the boot continues. Do you folks have an idea of what I could try? [20:33] "/etc/apache2$ sudo grep -sr javascript *" [20:33] conf.d/javascript-common.conf:Alias /javascript /usr/share/javascript/ [20:33] conf.d/javascript-common.conf: [20:33] mods-available/deflate.conf: AddOutputFilterByType DEFLATE application/x-javascript application/javascript application/ecmascript [20:33] there you go [20:33] conf.d/javascript-common.conf:Alias /javascript /usr/share/javascript/ [20:33] I'm not entirely sure that my raid metadata is version 0.90 -- I suppose I should check that ... [20:34] ohhhhh man how did that happen ! [20:34] xperia: check [20:34] its crazy such thing [20:34] xperia: dpkg -S /etc/apache2/conf.d/javascript-common.conf [20:34] I should also mention that my dmesg says that the requisite raid modules are loaded :) [20:35] ivoks: "/etc/apache2$ sudo dpkg -S /etc/apache2/conf.d/javascript-common.conf" [20:35] dpkg: /etc/apache2/conf.d/javascript-common.conf could not be found [20:36] ivoks: file exist however ! [20:36] i try it again to recheck [20:36] xperia: that check is only for packages [20:37] xperia: it means that no package created that file [20:38] Hrm. I should probably use --update the next time I assemble the arrays. [20:38] hmmm this knoweledge that you have is beyonf my horizont sorry asking such dumb questions as i am new user and learning but what does that mean. it was not cleared after the upgrade and need now to be removed ? [20:39] hmm i really asking me how this files where installed. it must be done by the upgrade process as i have never installed this files [20:42] ivoks: thinks know now what the reason is for this files where installed ! [20:42] "/etc/apache2/conf.d$ ls -la" [20:42] lrwxrwxrwx 1 root root 24 2009-11-01 16:02 gallery -> /etc/gallery/apache.conf [20:42] lrwxrwxrwx 1 root root 45 2010-12-04 19:24 javascript-common.conf -> /etc/javascript-common/javascript-common.conf [20:43] its gallery. it was updated and asked me if i want to use the new config files for the new package and it has then installed this link javascript-common.conf -> /etc/javascript-common/javascript-common.conf [20:43] in apache2/conf.d [20:44] well maybe also not [20:44] http://packages.ubuntu.com/natty/javascript-common [20:46] well, there you go [20:46] now you know everything [20:46] why would anyone package systemwide alias for /javascript is beyond me [20:46] wasn't me :) [20:47] i have just done a "/usr/share/javascript$ ls"and got this here as result [20:47] /mootools$ ls [20:47] mootools-core-jm.js mootools-core-server.js mootools.js mootools-more.js mootools-more-yc.js [20:47] mootools-core-nc.js mootools-core-yc.js mootools-more-jm.js mootools-more-nc.js [20:47] mootools [20:47] i like the name :) [20:47] footools [20:47] bartools [20:47] kidding [20:47] looks like this need files/scripts use this global path [20:48] libjs-mootools appears to recommend javascript-common [20:48] Web applications that use JavaScript need to distribute it through HTTP. Using a common path for every script avoids the need to enable this path in the HTTP server for every package. [20:49] what are these guys smoking? [20:49] all my vhost sites are broken because of this since days after the upgrade ! winning heavy [20:49] it seems that installing webapps from repo is going to end up an all-or-nothing approach [20:50] yeah [20:50] stupid [20:51] Javascript maintainers [20:51] xperia: so, as an easy way ouy [20:51] cool i am hearing [20:51] open up /etc/javascript-common/javascript-common.conf [20:51] yes [20:51] and rename that alias from /javascript to /braindeadpackage [20:52] and then, in gallery, if you are using it, find /javascript/ paths and replace them with /braindeadpackage/ [20:52] pure existance of this package is a bug [20:53] and next time every domain you host gets clobbered, change them again ;) [20:53] ha ha ha okay will do and then restart the apache server. about gallyer i have installed it but dont know really if it is becouse of gallery or other package was the reason. but i will check it out [20:53] my rule: never install we apps from archive [20:54] we, developers, tend to be smarter than we are [20:54] and then fuck up [20:54] pardon my language [20:54] good working on fix now:. ivoks you are a awesome person really. such expert people like you should have more in the world: [20:54] shauno thank you also a lot for the help. thinked since hours i am lost becouse of this strange problem [20:55] thanks [20:55] i'm going out for a smoke now [20:56] ivoks: have nice pause and time [20:56] hi guys i install dnsmasq but the problem is when i used the linux box as dns aside for gateway- PC will timeout and wont connect to internet if i make it dhcp it will connect. === Pici` is now known as Pici [21:00] that's a little unnerving. the only packages I can find (on .04) that depend on libjs-mootools are myphpadmin and midori. [21:01] midori's a gtk-webkit browser. why's it pulling in packages that'd alter your apache config? [21:01] shauno: well while the upgrade myphpadmin was updated and i was asked if i want to install the new config files. based on your research i would say it is becouse of phpmyadmin then with a possibility of 80% [21:02] i will just restart the webserver and test if phpmyadmin still works [21:03] xperia: most likely myphpadmin, yes. not many people install gui webbrowsers on -server ;) [21:03] ruben23: i itself use bind9 and dont use dnsmasq. resolves everything on bind9 here =>https://help.ubuntu.com/community/BIND9ServerHowto recomend using also bind9 instead of dnsmasq [21:19] ivoks: shauno:my sites works now great. so the fix recomended by ivoks worked great. all my sites now are running the way like before. checking now phpmyadmin how it works. [21:19] shauno: well everytime i install something it is recomended to install it from the repo. all external packages are a security treat. i never thinked that webapplication would make anyproblems and till yet did also not had any problem. [21:20] idea is to create common packages, that all web apps could use [21:20] mixing non-archive with archive web apps then results in this kind of problems [21:20] fwiw, webmails are much smarter [21:21] they create /webmail or /roundcube aliases [21:21] but leave it disabled by default [21:21] so you have to enable it; therefor, you are aware of that alias [21:21] this, with /javascipt is just nonsense [21:28] ivoks: tested now also phpmyadmin and it works still same. so the change in the alias does not affect really a lot phpmyadmin. looks like this mootools is not really used a lot here. [21:28] can I use apparmor to allow users to run programs with elevated permissions ? (similar to pfexec or sudo) [21:29] aarcane: afaik, apparmor is for restricting [21:31] ivoks, I know it's primary use case is restricting, but it'd be nice if it could simplify the use of certain semi-administrative apps (let certain users use certain apps in elevated mode automatically) [21:32] * ivoks shrugs [21:33] * ivoks 4 [21:33] ups [21:35] oh well. [21:36] * aarcane off to do some programming school work~ [21:36] thanks ivoks. [21:54] see you all next time need to write at my websites. have all nice time. bye [22:05] hi, does anyone knows a way to setup a mailserver with a dinamic ip using bind9 as dns? [22:25] aristo, in order for a mail server to work reliably it needs a static IP [22:26] DNS can take up to 48 hours to propogate, and during that time you won't get any mail [22:27] qman__ i'm aware of that, the problem is that currently our isp is not able to give static ip [22:28] i tried with DynDns and others services but i cannot receive any incoming mail from outside [22:28] ok, but installing bind on your server won't really help in this case, unless you already host your own internet DNS [22:28] is your ISP also blocking port 25? [22:29] most residential ones do [22:29] lemme verify that [22:30] also, most spam filters will block you [22:30] static IP is basically a requirement to get taken off their lists [22:34] nmap online shows only 22, 80 and 110 ports opened [22:40] aristo, do you have a choice of isp ? [22:41] well basically not [22:41] does basically not mean cost to high? [22:48] means that here are only 3 mayor isp [22:49] all 3 sucks [22:54] eh, there must be access providers that sell internet access to companies, right? [22:54] and none of them have fixed ip addresses available? [23:33] I have a concern... I started out with 3 2TB Samsung drives in RAID-5... each drive has 4x500GB platters. I just bought another of what I thought was the same drive... but I discovered after receiving it in the mail that is is actually a new model... with 3x666GB platters. What are the implications of adding it to my RAID5? [23:35] do adaptec RAID cards care? I have a 5405 [23:40] electrofreak: I think none [23:41] cokegen, can it affect performance? [23:41] i wouldnt think it could decrease it, since the 3x666GB would be faster... i'd guess that the array would be limited by the slower drives you already have [23:43] and by "faster" i dont really mean by that much ;) [23:44] c0nv1ct, yea... performance wise, there isn't that much difference... but yes, the newer model is apparently faster [23:44] electrofreak, the physical construct of the drive is completely invisible to the rest of your hardware [23:44] it's handled entirely in the disk's own controller [23:45] qman__ is right [23:45] there are no problems but the "speed" thing [23:47] right [23:48] I just removed the quiet and splash part of my grub entry [23:48] your RAID will only perform as well as the slowest, smallest drive you have [23:48] I'm trying to get console output at bootup [23:48] and having a wide variation in performance can cause spikes and such [23:48] but just one newer model shouldn't be a problem [23:48] but that didn't worked ... anyone knows how ? [23:49] cokegen, that's because of upstart [23:49] or rather, plymouth [23:49] in order to get console output at boot you have to modify the plymouth upstart scripts [23:50] qman__ thx [23:50] I come from old linuxes [23:50] I really hate it too [23:51] here's what I did [23:51] in /etc/init/plymouth-splash.conf, comment the exec line at the bottom [23:51] exec /bin/plymouth show-splash [23:51] cokegen, qman__, I just thought there might be some timing thing that can throw the RAID controller off... [23:52] and the same line in /etc/init/plymouth.conf, near the bottom [23:52] I think that's all I had to do [23:52] qman__, ooo, I hate that splash screen... but that plymouth thing is 10.10, right? I'm running server 10.04 [23:53] no [23:53] it's actually new as of 9.10 IIRC [23:53] and it's in 10.04 [23:53] oh, so disabling it would be the same... [23:53] it really should be configurable [23:53] removing quiet splash really should disable it [23:53] but it doesn't, you have to go modify those files [23:55] qman__, just comment out this whole post-start script block, right? (in /etc/init/plymouth.conf) [23:56] no, just the exec line [23:56] I mean, you could [23:56] I just did the one line [23:56] qman__, well, the whole if statement is just for that 1 line, heh. I figured it'd be fine to just comment out the whole block. [23:59] and really, the whole plymouth-splash.conf could be commented/removed/disabled [23:59] but I don't know of a way to disable things in upstart