[00:00] not all folders are 4k in size, folders with a lot of files in them might be (a lot) larger === jjohansen is now known as jj-afk [02:18] hi! [02:18] forgot there's a special channel for _server ;) [02:18] on 10.04, my dm_crypt randomly crashes when writing ~300-700GB large files, need to reboot the machine... any idea? should I try maverick kernel? here's the trace: http://www.abload.de/img/captureoby2.png [02:28] hmm, I have no single files that large [02:28] but I have a 6tb drive that is dm_crypt, and it's happy with 20gig files [02:29] but that really should go to the kernel channel :) [05:56] http://pastebin.com/HQzf9fxk [05:56] Whats this error mean? [05:56] permissions problem? [06:05] DjKiDD: run as root? === Psi-Jack_ is now known as Psi-Jack [10:45] I have a public static IP and a subnet routed behind it. I configurate it on ubuntu server and I have internet on my PC, but the server gives me the IP address of the linux not the one I picked up from the subnet. What to do ? http://img573.imageshack.us/img573/1244/capturewn.jpg === jdrab is now known as cortex|sk [13:01] New bug: #688922 in samba (main) "package samba 2:3.4.0-3ubuntu5.7 failed to install/upgrade:" [Undecided,New] https://launchpad.net/bugs/688922 [13:23] just wondering which virtual server deployment is best, kvm vmware, openvz? other? [13:25] i've been having some issues with kvm, and it seems to be less universal cause it requires vt on the processor, so i guess it's less universal.. [14:01] Hey. Does anybody know a web frontend for apt-get? :) [14:06] gorkhaan: landscape [14:09] well ty... but I prefer a free solution. === smoser` is now known as smoser [15:06] Dire need to assistance. After receiving so much network traffic, my headless server stops responding to ping/ssh/ftp.. everything. Ubuntu Server Maverick 10.10. I'm using it to back up some 200GB of files over FTP on local network but this isn't the first time this happens - it seems to lose connection quite often. Any tips to help remedy it? [15:07] What kind of network card? [15:07] Hm, One second. [15:07] PCI.. uh.. [15:08] Oh. For the first time i've plugged a monitor into the headless server before starting it and it's not responding to anything via keyboard. [15:08] What kind of keyboard? [15:09] PS/2 keyboard. Screen went blank somewhere along the way and the keyboard didn't do anything. Restarting it now to check the network card details. this is new, though [15:10] 00:12.0 Ethernet controller: VIA Technologies, Inc. VT6102 [Rhine-II] (rev 78) - that's my network card [15:11] Ethernet is onboard? Or a PCI card? [15:11] It's onboard. [15:11] Maybe it isn't to do with the network activity after all - maybe it's just general activity. Sometimes the server sits on for a few days and I don't access it much bar to upload a small text file or the like. Other times it'll just stop responding. [15:12] Wait, sorry. I'm pretty useless. I used "lspci" to view the network card. That means it's PCI, right? [15:12] No neccesarily [15:12] lspci shows the pci bus [15:12] onboard cards can be hooked through the pci bus [15:13] Aah. Of course. In that case i'll go with my answer of "onboard" i've had the tower open before and it doesn't slot into a PCI slot. [15:13] what does lspci | grep -i ethernet say ? [15:13] oh, you had it above, the Via [15:13] Yeah, that's the one. [15:13] Ken: So you have a keyboard/monitor on it [15:13] And you say it's locking up randomly? [15:13] penguin42: I do now, yes. Normally I ssh. [15:13] ken: When it stops responding what is shown on the monitor? [15:14] DevoKun: It is. I can't seem to find any pattern to be honest. It's usually during heavy network traffic, either locally or ftp access via a dyndns name [15:14] penguin42: I've never had it connected in time to see. I'll try and trigger a lock up, though. It was connected prior to the most recent (a few minutes ago) but the screen was blank and it didn't respond to keyboard. [15:14] You're backing up 200G over FTP? Tried rsync? [15:14] Ken: Does it come back or does it stay dead? [15:15] It stayed dead. [15:15] Ken: Anything in dmesg / /var/log/messages or the like? [15:15] I'll check now [15:16] dmesg is just showing me information from the reset I just carried out. How can I view a longer log? [15:17] Sorry for the really newbie questions. Troubleshooting a server is well beyond what i'm used to. I've got it reasonably well set up but I'm not even sure where to start with this. [15:17] /var/log/messages and /var/log/kern.log [15:17] however, if it died in an odd way then they won't make it to ther [15:18] It's giving an error for my wireless card, which I don't ever actually use. [15:18] meh [15:19] May or may not be of use: [15:19] [ 9.467313] ACPI: resource vt596_smbus [io 0x5000-0x5007] conflicts with ACPI region SM06 [??? 0x00005006 flags 0x36] [15:19] [ 9.467319] ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver [15:19] That's on boot though [15:19] wouldn't worry too much about that one [15:20] I'm seeing if I can make it do the same again. I'm transferring some large files over [15:20] This is part of the problem. The only pattern I seem to find is it's after a period of heavy activity. Time makes no difference. As i've said, it can happily sit on for days, barely used, with no issues. [15:21] Other times It'll be on for no more than a few minutes whilst I'm doing a lot of heavy traffic file transfers and it stops responding to ssh, ping, etc. Resetting manually is the option i've used. [15:21] Ken: this sounds like bad hardware to me. You should try running memtest86 on the machine to see what happens. [15:22] Memtest using a Live CD, right? [15:22] yes [15:22] I did have a hard drive die. It was years old though and it's since been replaced. The change of hard drive hasn't made a difference though - this happened before and after the replacement. [15:23] Thanks, DevoKun and Penguin42. I really appreciate you taking the time to help me. [15:25] Hmm. 2GB file just transferred without a problem. [15:25] I'll send another few larger files then run memtest. [15:27] Ken: If the machine survives memtest86 without crashing, then you can run some other hardware testing tools like cpuburn or smartctl (part of smartmontools). However, if you suspect the problem may be with your Linux installation, then download and boot from the "System Rescue CD" project liveCD: http://www.sysresccd.org/Main_Page you can startup the network then copy files on mass to see what happens. [15:29] DevoKun: Brilliant. That should really narrow it down. I've already switched out my router to see if that made a difference (I wanted to change my router anyway because my provider-branded one doesn't allow Wake On LAN), so I can rule that out as a possible problem. [15:32] Another question, somewhat unrelated to servers. Fastest protocol for transferring files over local network? I'm assuming FTP? [15:32] Fastest? [15:33] Yeah. To transfer from my laptop to my server. Two wired connections and ftp? [15:33] In terms of transfer speeds that is. [15:34] Actually, there is some data out there to suggest that Samba-to-Samba (smb) connections are the fastest. However, I always use, and recommend, rsync over ssh or scp. [15:34] You should train yourself not to use FTP. [15:34] I've been following an rsync tutorial to have incremental, automated backups [15:34] It's a bad habbit ;) [15:34] Haha. I'll heed that advice. [15:36] rsync is dead simple. Create a cron job that does: rsync -avz -e 'ssh' @: [15:38] What to do if dnsmsq is using the same port as named ? [15:38] something like: 0,15,30,45 * * * * /usr/bin/rsync -avz -e 'ssh' /home/ken ken@192.168.0.5:. [15:39] DevoKun: -e ssh is implicit since five years ago or so [15:40] Old habbit [15:40] I followed a tutorial and made a bash file, then i'm going to link it to cron. [15:40] "rsync -e ssh -varuzP /media/DATA/College/ :/home/kenneth/backup/College/" [15:41] Ooh, I just got an error although nothing bad has happened to my network connection [15:41] Segfault in miniserve.pl - that's webmin as far as I know [15:41] Okay, woah [15:41] I think we have success. [15:42] I see an error on the server screen. It's talking abut er... kthread+0x74/0x80 and bdi+start_fn [15:43] I'll have to do a reboot and then i'll try to dig that out of the dmesg log if anyone can advise me how I might go about it. [15:43] Ken2: Can you get a full copy of the whole error? [15:43] Hmm. [15:43] Is it possibel to have Squid to act as a transprent proxy with oly one network card [15:44] Ken2: -r is part of -a [15:44] penguin42: I'll give it a go. I used watch "dmesg | tail" on my server and it all appeared on the monitor. I'll have to reboot then i'll try to ssh and view the dmesg log fully to get a copy of it unless you have a better suggestion [15:44] adrianrally: you probably don't need both DNSmsq and named running at the same time. If you are trying to use just DNSmsq, then stop and disable named. [15:45] Ken2: If you can't get it any other way then try a picture! [15:45] penguin42: Haha, good idea. [15:45] Is it possible webmin is causing the issue? It showed a segfault with miniserv.pl which, as far as i know, is what webmin runs on. [15:46] Then a lot of scroll happened (rebooting to find it) [15:47] !webmin [15:47] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. [15:47] DevoKun. The thing is like this. I use my saver as a webserver. I have domains parked on it. [15:47] I use bind9 for that. [15:47] hello, are there any Ubuntu Server admins for hire out there? [15:48] I have a project I need an ubuntu server OS expert right away to help me set up a custom config by Monday [15:48] VirtualBox is already setup on my server but I need to build some VM's and and then configure some special software that has full documentation and tech supper. However I need an expert with Ubuntu Server OS to make sure it is all running optimally. I must place this server into a client's location on Monday. [15:48] DevoKun: This is my problem: http://img529.imageshack.us/img529/9469/captureqf.jpg [15:49] Okay, the first error I saw: [15:49] [ 38.155224] miniserv.pl[1519]: segfault at 19 ip 080cdcb3 sp bfefc100 error 4 in perl[8048000+12d000] [15:49] adrianrally: looking now [15:50] It doesn't give me much else aside from that. I'll try to dig out the rest from somewhere and use pastebin. [15:50] That's my server's ip addres. I should have an ip from the subnet(62.231*). [15:50] Does /var/log/kern.log be over written every time I reboot? If so, can I find older logs? [15:52] Ah. Scroll bad was limited in my terminal. [15:54] iptables -t nat -A POSTROUTING -o eth0 -j MASQUERADE [15:54] If I put this rule... I get the server's ip address. If not, I can resolve dns but no incoming traffic. What am I missing here ? [15:56] If I ping the gateway, it's ok. If I ping the PC from the server... no ping. [15:57] Alright, having trouble locating the previous log when I crashed. It's showing mostly reboot information and nothing before it. I am getting a segfault from miniserv.pl though [15:57] adrianrally: you need to be able to forward packets to the aliased subnets, so the masquerading rule is required. [15:58] Yes but it makes my subnet work as a intranet. [15:58] The subnet is routed behind the public ip address. [15:58] 62.231.69.58 should be my public ip address. [15:59] Ken2: It's possible, but the bdi+ blah error is a kernel oops, and that looks like the real issue [15:59] adrianrally: do you have multiple ethernet ports in the machine? [15:59] Yes. [15:59] But I use only one. [15:59] Is this your firewall? [15:59] No. [16:00] penguin42: Ah. And the kernel problem could either be hardware or an install issue? I'm trying to run another large transfer to encourage another crash [16:00] adrianrally: You're connecting your webserver directly to the Internet? [16:00] Ken2: Either hardware or a bug [16:00] Yes. [16:00] penguin42: Okay. Praying for a bug then. [16:00] DevoKun: On eth0 it's the public IP address the ISP gave me. [16:01] Ken2: You could try changing the font used on the terminal to give a better chance of it all fitting on [16:01] I've changed the scrollback to "unlimited" right now. I'm sshing and hoping it'll capture it all [16:01] Ken2: OK, is the ssh surviving the problem? [16:01] Not usually. [16:02] adrianrally: You're getting (eth0)86.122.123.155 as your public IP because that's the interface that traffic is being routed out on. Traffic may come in on (eth0:0)62.231.69.57, but it will go out eth0 because that's how you have your routing setup. [16:02] I'm just an optimist. How do the logs work? If I can access the log from before I had to reboot, I should be able to find the line that show the issue. [16:03] Ken2: It depends on the failure, if the kernel got upset it might not make it to the file [16:03] Oh. Hmm. [16:04] Running "dmesg /var/sys/log" shows only what has happened since boot, not since the previous boot. [16:04] DevonKun: So how do I send traffic from eth0 trough eth0:0 to my pc? [16:04] adrianrally: I recommend you install a firewall between the Internet and your webservers. pfSense (http://www.pfsense.org/) is a good option for this scenerio. Web based interface, quick and easy to configure. Then you can setup 1-to-1 NAT to server based on IP. [16:04] Why are you using two subnets? [16:05] You should really only need 1 in a hosted environment. [16:05] 62.231.69.56/29 is behind 86.122.123.155 [16:06] I had what I wanted but the harddrive exploded last night. [16:08] I didn't even changed the config on my PC. It's like you saw it in the screenshot. It's something with iptables I think. [16:09] adrianrally: So your ISP told you to use 86.122.123.155 as an IP and to use 86.122.123.129 as a router. [16:09] adrianrally: Where did 62.231.69.56 come from? [16:11] My ISP told me to put on eth0 86.122.123.155, and on eth0:0 62.231.69.57. Then on my PC 62.231.69.58, 62.231.69.59 on the wifi router, 62.231.69.60 on mom's laptop. [16:11] Ah ha. [16:11] http://pastebin.ubuntu.com/542312/ [16:11] is that it? There is normally a lot more after that [16:11] They don't offer support for linux.... they said to me to make internet connection sharing. [16:12] Just That! [16:12] I'm viewing it in Nano and that's all I can see. [16:12] Line 3, [cut here], is pretty much where the problem happened. [16:12] Ken2: My bet is there is a lot more on the console [16:13] silly Q' how to you comment out multiable lines in a configuration file [16:13] penguin42: What's the best way for me to view it on the console? It might be really obvious but I couldn't see it. Can you use dmesg with a parameter to have huge scroll back? [16:13] Right now Im just adding # to every line; surly there is a beter way [16:14] Ken2: are you getting it after reboot? [16:14] This is killing me DevoKun.... [16:14] Yes. I opened it in Nano and it's shown me all of the kern.log since 4 dec. [16:14] When I open it using "dmesg" it just shows since reboot. [16:15] DevoKun: Might be bind9 screwing up something ? [16:15] Ken2: It would be better to get the whole message as it dies on the console [16:16] adrianrally: no. this is not a DNS issue. your problem is with IP routing. [16:16] I'm trying to make it crash again. My monitor resolution is quite bad though so i'll be stuck with typing it out if it doesn't go into a log. [16:16] adrianrally: 62.231.69.56 can not be "behind" 86.122.123.155 because it is a *very* different network. [16:17] It's a 8 IP subnet behind that ip address. [16:17] That's what they said and it worked till now. [16:18] Ken2: Look at bug 670117 [16:18] Launchpad bug 670117 in linux "After some io activity kernel bug in dmesg, then server freezes (pings still working)" [Undecided,New] https://launchpad.net/bugs/670117 [16:18] Ken2: He has a similar inode.c:2099 but you see he has a lot more info after it? [16:18] penguin42: Yes. That's pretty much exactly what mine looked like [16:19] Ken2: Well the text of the backtrace is important [16:19] It appeared on my screen but isn't in the log. [16:19] any server admins for hire in here???? [16:19] adrianrally: see, if you wanted 192.168.0.10/24 routed through 192.168.0.1/24, that would be do-able. If you wanted 192.168.100.10 routed through 192.168.0.1, you could change the netmask to /16 and no problems. To route 10.10.100.10 through 192.168.0.1 you would need to setup your routing so machines new how to get to and from that IP. [16:20] Well, it's not in kern.log anyway. Is there anywhere else it might be? [16:20] Ken2: I doubt it, have you got a digi camera? [16:20] DevoKun I wanted more IPs. And that's what they gave me. [16:21] penguin42: Embarrassingly, not at the moment. [16:21] adrianrally: OK, so they probably gave you a separate gateway for that subnet also? [16:22] Now i'm just getting a lot of segfaults from python/miniserv.pl. Any suggestions as to what else I could use since Webmin isn't supported (If you don't have any off the top of your head i'll go google) [16:22] Let me translate your the mail. [16:22] Hmm, okay. I just sent 5GB of files over the network to the server and nothing happened. [16:23] Ken2: If your getting seg faults all over then I really do suggest the memtest86 run [16:24] I shall do that right now. [16:24] Maybe I should get a new hobby. [16:25] DevoKun: This is the email from the ISP(RDS): "We've routed 62.231.69.56/29 behind 86.122.123.155. The subnet can be configured on the local interface of the server if eth0 is from RDS(this is the ISP) and eth1 the local network. For example ifconfig eth1:0 62.231.69.57 netmask 255.255.255.248. And you can use on the local network, directly connected into eth1, addresses from 62.231.69.58 [16:25] to 62.231.69.62(with netmask 255.255.255.248 and gateway 62.231.69.57). Or you can add aliases for example: ifconfig eth1:1 62.231.69.58 netmask 255.255.255.248" [16:26] But. I know I didn't used another network card. I've put 86.122.123.155 on eth0, 62.231.69.57 on eth0:0 and the other ones I used them as aliass or on other equippments. [16:27] adrianrally: Sure. You need to setup a route for the 62.231.69.62 IP. [16:27] adrianrally: Aliases are fine, but they still need routes. [16:29] How do I do that DevonKun ? [16:29] adrianrally: add the gateway they gave you (62.231.69.57) to your config file for the eth0:0 interface [16:29] Hmm.. [16:29] Having a odd issue trying to push a proxy-configuration file over DHCP3 [16:30] DCHP3 crashes with /etc/dhcp3/dhcpd.conf line 43: unknown option dhcp.custom-proxy-server [16:30] Cant find the correct syntax though for that option [16:31] Done that. Now ? [16:32] adrianrally: restart the network [16:32] adrianrally: if you type "route -n" you should see a route for both networks [16:34] It died. [16:34] adrianrally: your machine locked up? [16:35] The internet crashed. [16:35] adrianrally: what are the contents of "route -n" ? [16:35] adrianrally: what are the contents of "ifconfig -a" ? [16:37] http://pastebin.ubuntu.com/542320/ [16:39] adrianrally: you have two default gateways. See the "0.0.0.0 ww.xx.yy.zz" lines? [16:39] adrianrally: you can have two default gateways, but the metrics need to be adjusted. [16:39] adrianrally: Ideally, you need to have routes (top 2 lines of route -n) with only 1 default gateway. [16:40] Right... soo.. what do I have to do to achive that ? [16:40] 0.0.0.0 62.231.69.57 0.0.0.0 UG 100 0 0 eth0 [16:41] This line appeard when you put me to type the gateway for eth0:0 [16:41] After I restarted. [16:42] Memtest is running. I'm going for a sandwich. Anyone want one? [16:42] Ken: good luck [16:44] Thanks, DevoKun. I appreciate the help both you and Penguin offered me. I'll be back shortly no doubt with more tedious question for you both. I guess it's a curse being so knowledgeable, huh? [16:45] adrianrally: you have two default gateways - that's bound to confuse the ip stack [16:46] So what do I have to do now ? [16:46] Wait a moment. [16:47] DevoKun: http://pastebin.ubuntu.com/542327/ [16:48] Is it possible to have 2 dhcp assigned ip's on 1 wan port? [16:50] adrianrally: why on earth do you have two ip addresses on different IP networks on the same NIC? [16:50] Doonz: the dhcp server will only assigne a single ip to a mac address unless the developer was drunk [16:50] Because the first one is the publick one, the second is the gateway for the subnet... [16:51] Or smth like that. [16:51] RoyK: is it possible to use mac spoofing on the one interface [16:51] Ah, Mem-test has found two errors so far. [16:51] Doonz: obviously, yes [16:53] The last routes -n looks very logic to me: http://pastebin.ubuntu.com/542327/ [16:53] RoyK: can you point me to docs on how to do that [16:53] Doonz: you'll need to script that up - basic config can't do it [16:54] k [16:54] thank you [16:54] mac spoofing is simple, though, ifconfig eth0 hw ether x:x:X:x:xx:X:X:X [16:54] something [16:54] My ip address is the same on my PC and my server(86.122.123.155). I need to have 62.231.69.58 on my PC. http://img529.imageshack.us/img529/9469/captureqf.jpg [16:55] adrianrally: you're using the _same_ IP on the PC and server? [16:55] * RoyK sends adrianrally into google to read up on IPv4 and general network basics [16:56] RoyK: Yes. It's something wrong with the server. Look at the screen shot http://img529.imageshack.us/img529/9469/captureqf.jpg [16:56] this is increadably fustrating [16:56] has anyone got a Proxy-config push working with dhcpd3? [16:58] I;ve got the file scripted but it doesnt seem to be getting pushed out to clients via dhcp3d [17:00] adrianrally: something is wrong with the config and perhaps your understanding of networking [17:10] RoyK... do where do I get started? [17:12] http://tinyurl.com/22sxxfs [17:12] I don't get it. What's wrong ? [17:12] your understanding of it all. :) [17:12] Seems to me you just use the server as gateway. [17:13] That's why the servers IP is shown in your browsers "whatsmyip" [17:13] A few hours ago it was working. I was getting on my PC 62.231.69.58 aka the IP i wanted. [17:14] Something happened... and I don't know why. [17:14] And what. [17:14] adrianrally: but what's wrong? [17:14] adrianrally: our screenshot shows your ip on the pc, exactly the one you say you want. [17:15] +y [17:15] Think Im just about to give up on this haha [17:15] air^ I don't need the server's public IP. I want my private one. [17:15] so don't route the traffic through your server. [17:21] I followed this tutorial in the first place: http://www.ubuntugeek.com/sharing-internet-connection-in-ubuntu.html [17:22] And at some point it worked.. but I did something wrong. [17:22] penguin42: In case you were curious, memtest has turned up two errors. This might be what's causing my server to stop responding. [17:24] Ken: Ah right, go get that fixed and life should be a lot better [17:25] penguin42: Memtest is still running. I'll use their site to see if I can figure out how to fix it when it's done. [17:25] Ken: Are you running over clocked or normal? [17:26] penguin42: I'm not sure. i'm going to assume normal. I just selected it from the grub2 boot menu. [17:26] Ken: No, I mean your hardware, is it normal or have you done anything odd with it? [17:27] penguin42: Ah, sorry. No, it's normal for the most part. It's a really old system. The newest thing is the hard drive. One is ID, the other is SATA connected via a PCI SATA card. [17:27] RoyK: any idea what he tried to achieve? :) [17:28] Ken: Make sure your cooling etc is OK but chances are it's probably just a dead DIMM [17:28] air^: not really, although I think he was in desperate need for basic network knowledge [17:28] penguin42: It's cool enough right now. I have a spare RAM module about somewhere so, if I can figure out which is dead, I'll try switching it out. [17:28] RoyK: seems so. the guide was about sharing internet... that is.. exactly what he had done (server acted as gateway) :D [17:28] Ken: clean the contact areas both on the DIMMs and their sockets - on old systems, those often corrode [17:29] RoyK: I will do. Thank you [17:29] Ken: use isopropanol or something [17:29] completely unrelated but a tricky thing about bad memory that I recently had. [17:29] * Ken puts down the bottle of vodka again, "Ah, okay." [17:30] it'll work with vodka as well :) [17:30] installed os x server on an old power mac g5. it failed on "bad media". [17:30] tried everything, still failed, tested the harddrives and install media in another g5. worked nicely. [17:30] put the drives back, g5 worked nice, run updates, update failed on bad media and fscked the system :O [17:30] turned out the memory was bad. seems it corrupted something in uncompressing stage. [17:31] never seen anything like it, pretty fun though when everything started working with other memory :) [17:31] Fun? Hmm. I gus this is all a learning experience but i'm not sure i'd rush in and say "fun" [17:32] I mean it's not hard for me or anything. Penguin, RoyK and DevoKun are pretty much holding my hand through it [17:33] Ken: Working with broken hardware really takes the fun out of it [17:33] Ken: You just never know when you're going to walk into a bear trap [17:35] penguin42: I do enjoy the whole thing though. I have this old tower that i've turned into a server. It's running apache with joomla, webmin, I've set it to Wake on LAN, has samba (although I never use it), FTP, SSH, runs a headless virtual machine that I can rdesktop into, uses dyndns so I can connect through a host name. I was in the process of setting it up, between my laptop and the server, for rsync backups. [17:35] I've learnt pretty much all of it from scratch using the internet. [17:36] Not that impressive compared to some setups i'm sure but i'm gradually learning more and more. Next port of call, if I get it working, will be to set up VPN [17:37] * Slyboots is far from having fun [17:38] * Slyboots is just about loosing his mind over it :P [17:38] this is rather nice http://pastebin.com/7gtaGRwc [17:38] I'd offer some help, slyboots but i'm afraid i'll well out of my depth dabbling with servers. I'm not bad at helping in the desktop channel but this is where i'm completely lost. [17:39] Woah, is that a machine you have, RoyK? [17:39] Its more DNS/DHCP issues [17:39] Ken: yes, I have two of those :þ [17:39] RoyK: I can't possibly imagine ever needing that much space. [17:40] Royk: Impressive, though. [17:40] disk-based backup will require lots over the years [17:40] and we _do_ have a lot of data [17:40] I guess that makes sense. I'm just a home user with paltry needs. [17:41] I have no need for a server at all anyway but I figured I had an old tower sitting around doing nothing so I might as well do *something* with it [17:41] I'll be installing a 6TB home server for my brother for christmas :) [17:42] :) [17:42] I figure I'm not doing bad. I have 1.5TB but only because I happened across a terrabyte that was pretty cheap and figured "why not?" [17:42] I'm having more success with this tub of haagen-dazs than with the server right now though. [17:43] btw, terra = earth, tera = 10^12 [17:43] I was being hippy (I definitely didn't just make a typo because I was stuffing a spoonful of ice cream into by gob) [17:43] What is the best way to get php5 --enable-pcntl? I tried the instructions at http://ubuntuforums.org/showthread.php?t=549953, but make test fails 100%. [17:45] Okay, memtest is still running and it's got four errors now. Any suggestions as to how I might work out which DIMM these might be on? [17:46] Ken: unplug one dimm, run memtest... [17:46] or unplug all but one [17:46] RoyK: Ah, of course. Thank you. [17:47] On another note. I'm using Webmin which isn't really supported by Ubuntu. Any suggestions as to what else I might use? [17:47] Amgine: if that flag isn't compiled into the php5 that comes with ubuntu, apt-get source php5, add it, and build a new package [17:47] !webmin [17:47] webmin is no longer supported in Debian and Ubuntu. It is not compatible with the way that Ubuntu packages handle configuration files, and is likely to cause unexpected issues with your system. [17:47] Isn't that the truth. [17:47] Segfaults all over the place. [17:48] RoyK: compile the full php5? [17:48] Amgine: yeah, doesn't take too long on most hardware [17:48] I'll give it a shot... [17:50] configure: error: xml2-config not found. Please check your libxml2 installation. [17:50] Amgine: apt-get source php5, unpack the tar.gz, patch it with the patch, edit debian/rules and add that flag [17:51] run dpkg-buildpackage [17:51] done [17:51] Amgine: if you compile it manually, it won't install in the ubuntu standard place - better make a package [17:52] anyone else using Webmin? [17:52] optikalmouse: I am. [17:52] !ebox [17:52] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [17:53] !zentyal [17:53] Silly bot. [17:53] eBox has been renamed so if you're googling it, you'll need to search for that [17:53] to everyone using a web-based admin tool like webmin or ebox or whatever - please don't! it doesn't take very much to learn how to administer a server using the commandline, and it'll be faster over time, way faster, and better, and ... [17:53] Ken: is it slow for you? I see two perl scripts being run in the cronjob [17:53] every 5min. [17:54] and i noticed that the comp that I'm sshing into that runs webmin slows down every few min... [17:54] RoyK: I would love.. LOVE to have a webmin interface for squid proxying lol [17:54] Im about to *snap* here :P [17:54] optikalmouse: I see the same thing, Optikal. It is very, very slow sometimes. [17:54] Slyboots: Er, I think there is a squid proxy webmin interface.. [17:54] Slyboots: why? squid config is simple :) [17:54] RoyK: There's rather a large number of unmet dependencies; should they be installed via apt-get? [17:54] RoyK: Also, I try to use SSH where I can. [17:54] Oh I got Squid and DansGuardian going [17:54] Amgine: I guess you'll need a bunch of -dev packagages [17:54] I can-not; NOT; for ANY reason; get WPAD Proxy-config pushing working [17:55] At all [17:55] Yep. [17:55] RoyK: But i freely admit that jumping into a nice GUI is sometimes better, even for tiny things like viewing disk usage, etc. [17:55] Its setup; from everything I can see it *should* be working [17:55] Ken: I have Icinga alarming me if disks fill up [17:56] Ken: how much do things slow down for you? [17:56] RoyK: I might just have to look into that too. [17:56] Ken: also, ssh'ing into a server and typing df -h isn't very much work :) [17:56] optikalmouse: Well, it's usable most of the time. It's a bit laggy but nothing too bad. [17:57] optikalmouse: Having said that, for most things I try to use SSH. [17:57] RoyK: But it doesn't have pretty colours. [17:58] * RoyK hands Ken a bucket of yellow paint [17:58] I work for a non-profit organisation and we have a split site. In the coming months I hope to verse myself well enough so I can set us up with a proper server system so we can share files over VPN and.. y'know, stuff. [17:58] Ken: does it slow down while you're using ssh? [17:59] optikalmouse: No, I guess not. I mean my server is crashing but I think that's unrelated. SSH doesn't ever be so slow that I can't use it unless I'm doing a huge amount of file transfers over ftp. [17:59] Aye; I need a break [17:59] * Ken hands slyboots a beer and a tub of ice cream. [18:00] Slyboots: It's my solution when things aren't going well. [18:00] hmm ok [18:00] optikalmouse: I'm getting a lot of segfaults from miniserv.pl recently though, which, as far as I know, is the "server" that webmin runs on. [18:00] the sysadmin for the server was thinking that rsync slowed down the server [18:01] optikalmouse: I'm going to try the eBox platform since it's better supported. [18:01] but I can clearly see webmin's perl scripts running every few min :/ [18:01] hm [18:01] optikalmouse: Yeah, when I was trying to find the problems to my server, perl scripts were running every few minutes like you said. I don't know if that was doing any harm (e.g. slowing things down) but they seemed to be firing very frequently. [18:03] I checked `top` when one of them ran [18:03] and it hit the cpu for 2-4% [18:03] and the memory for 2-5% I think [18:04] Ken: Much appericate :D [18:04] A beer would be good right now actually [18:05] Maybe if the module using those perl scripts could be isolated then you might find that it doesn't hang so much? My server only ever has me accessing it and it's sitting with about 1GB of RAM (which we've newly discovered, some of it is bad) but it's never slowed as badly as you've experienced. [18:05] yeah I don't know [18:06] * Ken hands Slyboots a bud. [18:06] Slyboots: I don't even drink that often so one beer will result in me on my back, so if my typing becomes incomprehensible feel free to get an op to kick me if I'm not responding to a polite request to shut up and leave. [18:07] lol [18:07] holy crap [18:07] I ran free -m and apparently the server has 192mb of ram free :| [18:07] optikalmouse: Discovered what's causing the problem? [18:07] wtf is using up all that mem? :| [18:08]