[00:01] New bug: #420813 in openssh (main) "ssh blacklisting of private keys 9.04_64" [Undecided,New] https://launchpad.net/bugs/420813 [02:49] Does anybody know of a VPN server that's easy to set up? Doing some packet sniffing on my local network has made me paranoid of what could be accomplished in public [03:07] jumbers, I use openVPN, though I wouldn't say it's particularly easy to set up [03:08] it helps if you understand how SSL and certificates work, and routing [03:09] if you're just looking for something simpler for a particular use, ssh tunneling goes a long way [03:10] Hear, hear. [03:36] If I SSH tunnel, can I do it so that it tunnels traffic on all ports? [03:37] I know it can be done on a single port basis [03:38] no, only one port per tunnel [03:38] for more than a couple, you'll need a full-on VPN, like openVPN [03:39] SSH tunneling also can't do UDP ports (to my knowledge) [03:40] Hmm, then it looks like I'm going to need openVPN [03:43] ssh is also tcp-based, so you get the tcp-over-tcp resend hysteresis problem. [03:44] However you CAN do UDP over ssh tunnels (-w), but not simple port forwarding (-R/-L). [03:44] I did that for syslog once, it was fugly. [04:05] * ScottK waits impatiently for lamont to upload the new Postfix release. [04:36] I have a server setup with ubuntu server edition, but I'm having trouble with the network. I connected directly to it via a crossover ethernet cable, but when I copy more than a few hundred megabytes of files, somewhere randomly in the middle, the connection will drop and I get a "No route to host" error until I restart the server. (I've tried restarting ntp, ssh, and proftp all to no avail) I've used sftp to try to copy files as well as [04:36] scp and both freeze and drop connection. [04:37] How do I stop it from freezing? Is there a setting which prevents more than a certain amount of data? [04:37] I still have 30GB of hdd space on the server, and I've only used about 16 so far copying on to it. [04:39] The server is running Ubuntu Server edition 9 I think, and I'm trying to copy data onto it from Fedora 11. [04:41] Ubunty versions have a month and a year. Do you mean 9.04? [04:45] I believe so [04:46] Yes, cat /etc/issue says 9.04 [04:46] This is generally a pretty quiet time of day. [04:47] I'd recommend seeing what network card you have using lspci and then Google. [04:48] Is 802.11bg a wireless thing or just a network thing? (I see two, and I want to see which is for the wired) [04:50] 802.11bg is WiFi (wireless) [04:51] lspci | grep Ethernet [04:53] Nothing online will tell me what the ethernet card is, I suspect it's onboard. lspci tells me "Realtek Semiconductor Co., Ltd. RTL-8139... [04:53] (It's a Toshiba Satellite A105-S2061 btw) [04:54] that's a common on-board chipset. [04:54] my old router has an RTL-8139 apparently. [04:55] I don't suppose there's simple setting I can change that will allow me to copy lots of data onto the server without disconnecting? (As in, is there something that prevents large data transfer?) [04:56] over a LAN? you need to figure out what's causing the disconnections first. [04:56] over a direct cable connection via a crossover ethernet cable [04:57] the obvious thing to check is the cable. [04:57] I can move gigabytes of data on my LAN (through a switch, mind, not computer-to-computer) without interruptions. [04:57] Do you have any ideas as to what could be doing it? Any ideas on how I could test it in other ways? What can I check about the cable? It's hardly been used, it should be in great shape. [04:57] there's nothing wrong with the RTL-8139. [04:58] It could be a defective cable. It's not a common fault, but it's possible. [04:58] I gues one thing you could do is to install ethtool, and see if your NICs are in full duplex mode. [04:58] connecting directly, they might have issues autonegotiating speed and duplex. [04:58] (just a theory.) [04:59] If it were the cable, wouldn't it be an all or nothing thing, rather than failing after copying tons of data successfully? [04:59] not necessarily. [04:59] but usually, you're right. [04:59] I'll switch to a standard cable and try again [04:59] If neither NIC is 1000baseT, you can't rely on them to autonegotiate sensibly [04:59] a standard cable won't work unless you use a hub or switch. [04:59] PhotoJim: wrong [05:00] twb: did your mother not teach you how to be polite? there are more respectful ways to disagree. [05:00] Even recent 100baseT NICs will negotiate crossover correctly. [05:00] twb: unless you know he has two of those, then you don't know that that's relevant. [05:00] A standard cable *might* work as a crossover cable, if at least one NIC is sane. [05:00] alright. might, not won't. [05:01] twb: Well since he HAS a crossover cable, how is that relevant. [05:01] ScottK: because the cable might be bad. [05:01] I suspect it's a speed/duplex mismatch. probably duplex. ethtool will tell us. [05:01] PhotoJim: I agree. [05:02] I do have to do "sudo ifconfig eth0 192.168.2.101" to get my IPs to match so that I can connect at all. [05:02] phylogenesis: nothing wrong with that [05:02] damn [05:02] I'm not a big fan of connecting NIC to NIC, but I can't tell you that I feel that way based on logic. I just like the flexibility of being able to add additional machines to the network. [05:03] PhotoJim: I don't like it because all my gear is too old and shit to autonegotiate, and a switch is easier to find than a crossover cable [05:03] Does the fact that I have to completely restart the server after a failed transfer say anything? [05:03] If I don't I just keep getting a "No route to host" error [05:03] phylogenesis: are you running NetworkManager? [05:04] Man I wish I hadn't left my router at home :-/ [05:04] twb: good reason. I have everything from stuff on AUI 10BaseT to gigabit stuff on my switch. [05:04] twb: on the server I only have command line so I'm not sure. I'm not specifically running that but it may be running automatically. [05:04] phylogenesis: you need to get yourself a small switch and keep it in your bag o' bits. :) [05:04] phylogenesis: pgrep Network [05:04] phylogenesis: do you get a number? [05:04] I think Network Manager runs by default on Ubuntu. [05:04] one sec, gotta wait to the machine starts back up [05:05] PhotoJim: only on desktop installs IME [05:05] proftpd is taking a very long time to start now [05:05] Use SFTP and/or vsftpd, not proftpd. [05:05] twb: that could be. I'm not 100% sure it installed on my server. I had issues with it, so I think it did but I didn't make notes. [05:05] I was, but I installed proftpd when I set up the machine [05:05] so it starts up at startup [05:06] "pgrep Network" returns nothing [05:06] phylogenesis: good. Then the problem is probably something NEARLY as dumb. [05:06] DHCP? could it interfere? [05:06] it should fail elegantly. [05:06] phylogenesis: instead of doing ifconfig, configure the static network in /etc/network/interfaces [05:07] Oh, my server is using a static IP, I set it to use 192.168.2.100 when I had my router, and it seems to have kept that up (I don't see why it wouldn't just pointing it out) [05:07] PhotoJim: dhclient will break him if he's doing-it-wrong, though it's surprising that he's tripping it. [05:07] twb: might be wise to disable it here, to rule it out. [05:07] phylogenesis: "no route to host" means that something is eating your point-to-point route [05:08] I'm getting behind here: What should I try next? [05:08] edit /etc/network/interfaces? [05:08] phylogenesis: yes. [05:09] Back it up first [05:09] phylogenesis: oh, first: do you have physical access to this machine? [05:09] Edit it how? should I just comment out the iface eth0 inet static and following lines? [05:09] yes [05:10] I have two laptops in front of me, this one is my main one (client, running Fedora 11), and the other is the server (running Ubuntu Server 9.04) [05:10] You want it to basically contain a single interface, "iface eth0 inet static" "address 192.168.2.100" "pointtopoint 192.168.2.101" where .2.101 is your other box's IP [05:11] Ah, I guess you need "netmask 255.255.0.0" or so, too. It's a bodge, but it should suffice for now. [05:11] Can you pastebin what I should have in it? I have pretty close to what you said already. [05:12] phylogenesis: pastebin what you have now [05:14] http://pastebin.com/d24f1ccc1 [05:15] My router was 192.168.2.1 when I had it here [05:17] twb, should I install that eth tool? [05:18] No, that makes sense. [05:18] I dunno what's going wrong. [05:18] You definitely got "no route to host", and DIDN'T get "destination host unreachable"? [05:19] "scp -r /files/Music phylogenesis@192.168.2.100:/home/phylogenesis/backup/Music" returns "ssh: connect to host 192.168.2.100 port 22: No route to host\nlost connection" [05:19] "ip route" ... ? [05:20] 192.168.2.0/24 dev eth0 proto kernel scope link src 192.168.2.101 10.32.2.0/24 dev eth1 proto kernel scope link src 10.32.2.55 metric 2 default via 10.32.2.1 dev eth1 proto static [05:21] I'm sorry, that's on my client machine [05:21] so is the "no route to host" message [05:21] is that what you were asking for? [05:22] on the server I get: "192.168.2.0/24 dev eth0 proto kernel scope link src 192.168.2.100\ndefault via 192.168.2.1 dev eth0 metric 100" [05:31] twb, someone in #fedora suggested that it's an sshd issue [05:32] Sorry, I'm a bit busy [05:32] phylogenesis: I assume you try ping(1) when ssh fails. Does ping work? [05:33] Does your route table change before vs. after the failure? [05:33] Are you transferring data with scp, or doing something more involved with ssh directly? In the latter case, perhaps you have an idle connection that is being interrupted. [05:34] ping 192.168.2.100 returns "Destination Host Unreachable" whereas it works when I first setup the connection, before I transfer any files [05:34] I'm also assuming you're checking BOTH ends of the connection -- have you confirmed that the machine you're connecting the Ubuntu server *to* is not borked? [05:35] np about being busy, I'm patient [05:35] Yeah I went and asked about that, and am trying to talk to them at the same time, which is pretty hard :-P [05:37] twb, How can I check my routing table? I have standard internet going via wireless at the same time (on the client (fedora) machine only). I'm transferring with scp and not doing anything else with ssh. [05:37] phylogenesis: route -n [05:38] What should I see there? I see two connections: Destination=192.168.2.0 and 0.0.0.0 (I don't see 192.168.2.101 anywhere) [05:51] pastebin it [05:51] Or better, compare it before and after the failure [05:52] Okay, but that'll take a bit. (gotta restart, get it to work, and then fail) [05:52] I'll post when I have that [06:05] twb, sigh... How can I tell if my computer can connect to the internet? I tried sudo apt-get install ethtool and it failed to connect. I'm thinking, gee maybe this has something to do with it? I'm connected to the internet via an ethernet cable (yet another one). [06:06] ping www.google.com gives me "Destination Host Unreachable" error [06:06] *errors* [06:07] You have two NICs? [06:07] Nvm, it gives me nothing, but ping 66.102.7.104 does (which is googles IP according to this computer) [06:07] I have one onboard ethernet nic and a wireless one [06:08] I dare not try to connect wirelessly as the school's wireless is very difficult to get working on any OS [06:09] and so far I can't connect with my own school credentials (right now I'm on wireless with my client machine thanks to a friend putting his in). However, the wired connection works immediately and easily on my client machine when I use it. [07:10] twb, it failed again. I plugged in my server to the internet via ethernet. I changed /etc/network/interfaces to use iface eth0 inet dhcp, then downloaded updates and ethtool. I then restarted the server and ran sudo ifconfig eth0 192.168.2.100. On the client I ran sudo ifconfig eth0 192.168.2.101, then "scp -r /files/backup phylogenesis@192.168.2.100:/home/phylogenesis". It ran for about ten minutes then stopped with: "[DB]_Bleach_182_[ [07:10] 1BD395A7].avi 80% 117MB 0.0KB/s - stalled" The routing table on the server was the same before and after, I forgot to check the routing table on the client before, but the after one is listed: http://pastebin.com/dbcf823 [07:13] ifconfig on the client still tells me inet: 192.168.2.101 for eth0 [07:17] neither routing table changed [07:19] hi, im trying to setup a irc server, but i cant conect to it from the outside , ive already forwarded port 6667, any ideas? [07:20] samd: forwarding the port, and then connecting to 127.0.0.1, should be sufficient. [07:20] samd: assuming, of course, that the IRC daemon is running and listening to lo [07:21] twb, i have to be connected to 127.0.0.1, from the server in order to connect from the outside? [07:23] samd: OK, suppose your server's name is "fs", and it's accepting ssh connections from anywhere, but only accepting loopback connections to the IRC server. [07:23] Then ssh fs -xfL 6667:127.0.0.1:6667 && irssi -c 127.0.0.1 [07:24] ill try that [07:35] twb: if samd connects to the irc server on the localhost interface there is no routing or port forwarding needed. unless you want it accessible outside [07:35] ight, so that command basicly connected me trough ssh then connected to the irc as local? [07:37] Boohbah, twb, yea, i want it to be open to some friends,, is there any other port supposed to be opened/routed apart from 6667? [07:39] samd: the command I gave will connect you to the IRC server from anywhere, without opening access to the IRC server to the world. [07:39] It assumes you already have a shell account on the box, of course. [07:40] If you WANT to let arbitrary people connect, you only need port 6667 open AFAIK. [07:41] twb ight, ill chek on that then, yeah,i have ssh acc on the serv, its a old computer running as personal server here at home [07:58] * Boohbah missed the ssh proxy part [08:11] New bug: #420957 in samba (main) "package libpam-smbpass 2:3.3.2-1ubuntu3.1 failed to install/upgrade: subprocess post-installation script returned error exit status 139" [Undecided,New] https://launchpad.net/bugs/420957 [10:26] New bug: #421005 in freeradius (universe) "can't compile version from repository (on 64bit)" [Undecided,New] https://launchpad.net/bugs/421005 [10:36] New bug: #408306 in dhcp3 (main) "Wireless not connecting to AP, no IP address assigned through DHCP" [Undecided,New] https://launchpad.net/bugs/408306 === xorigin_ is now known as xorigin [11:51] hi there somebody now about pount or another web proxy server [11:56] roxy09: a little? [11:58] i need to install a web mail server and i would like to know wich proxy or another extra web tools i need to install to make safety and fast the access [12:25] roxy09: web mail server? [12:27] yes [12:28] roxy09: are you talking about squirrelmail or something to that effect? [13:33] hi there somebody have experience setting a email server with security staff ? i mean antivirus, antispam, etc? [13:39] also which is the best mail delivery agent? [13:40] ao¿nother question amavis-new or greylisted? [14:06] Anyone here who can help me with an SMB-issue... the samba channel is pretty dead... [14:32] ScottK: 2.6.5 already? sigh [14:32] this weekend is swamped - prolly early next week [14:40] I'm reading the manual, and i'm not sure to understand what the line ( domain Local domain name ) is for in the resolv.conf. Can someone help ? [14:41] I understand that nameserver retrieves the server by its IP [14:42] slap: that is the search domain appended to any domain names not found at your resolver [14:42] slap: for example, an internal domain [14:43] that's waht i'm trying to do... [14:43] So if I have a mynetwork.lan... [14:43] search my.local int.my.com [14:44] And I look for mycomputer, the resolver will try to resolv mycomputer.mynetwork.lan ? [14:44] yes [14:46] But if there's already a nameserver 192.168.1.1, do I still need the line domain, or resolv will try with the default server address to map a local machine? [14:47] no, if you have DNS records for mynetwork.lan on your NS at 192.168.1.1 you don't need the search line [14:48] Right to the point. Thanks a lot. [14:48] welcome [14:55] any one here having odd Pam login issues since latest patches in LTS 8.10? [14:57] hmm I mean 8.04 [15:51] hi everyone [15:51] lamont: Yep. [15:52] anyone know how to get procmail to run on messages after amavis has filtered them? [15:52] I want spam messages to go into user's junk mail folder [16:27] I have a VPN router in front of a ubuntu box and I want to give users rights on shares and so on when they VPN to it... what should be the best way for this ? users use windows clients [16:37] does the router handle all the vpn login stuff? [16:47] anyone know how to get procmail to run on messages after amavis has filtered them? [16:57] <_ruben> tosh: that'd be default behaviour for postfix configured with amavis as content_filter and procmail as delivery agent [17:07] yeah [17:08] _ruben, I don't think the issue is with amavis, but that postfix isn'ta using procmail for the virtual mailboxes [17:40] i need help guys [17:40] ? [17:41] any one ? [17:41] ? [17:54] Abdullah9, whats the problem? [17:57] i installed a new drive for my wireless adapter , "rt2870 " , it seems it don't work [17:57] where did you get the driver? [17:59] http://www.ralinktech.com/ralink/Home/Support/Linux.html [18:00] did you try just going to System->Administration->Hardware Drivers to see whats there? [18:02] yes [18:03] is it listed there? [18:03] i can't find it [18:03] is this a laptop? [18:03] yes [18:04] but i am useing USB port [18:04] whats up guys? [19:20] Hi, I'm having trouble with a connection timing out. I have a laptop running Ubuntu Server 9.04 (server). I have another laptop (client), this one, which runs Fedora 11, that I try to connect to the server with via a direct crossover ethernet cable. Both machines use dhcp so first I run sudo ifconfig eth0 192.168.2.101 on the client and sudo ifconfig eth0 192.168.2.100 on the server. I then use scp to move files from the client hdd to th [19:20] e server hdd, and it works for about five minutes, but then stalls. If I attempt to rerun the command, I get "No route to host". How can I prevent the connection from timing out? [19:24] Both machines are fully updated, and I'm attempting to copy over 20GB of data. However, I've also tried using Nautilus in Fedora 11 to transfer data via ssh and it works the same way. Just as with scp, it can copy up to a couple GB before stalling, but even if I copy small chunks at a time, it will eventually stall on one of them. [19:25] phylogenesis: .. could be few things [19:25] I attempted the copy with neither computer connected to the internet and both having the wireless cards switched off, but the exact same thing happened. [19:25] simplexio, please, any ideas are welcome [19:25] phylogenesis: is dhvp server still running, sound like you manually define ip, and did you define default route [19:27] if those net stuff is ok, i would check next that you are moving files to linux fs ( those all support big files ) and not something like fat ) [19:27] I don't think I did. Although, initially the server was using iface eth0 inet static\naddress 192.168.2.100\n etc [19:28] simplexio, both machines are linux, with the client having a Windows 7 partition, but that isn't where the files are (they are on an ext4 partition I believe) [19:28] phylogenesis: but it should affect actually.. if you define ip + mask for inerface then that interface is default route for it [19:29] phylogenesis: does dmesg give anything funny about network interfaces [19:30] Can you explain what exactly to do? (I just spent about 10 hours on this yesterday, so clearly I don't know exactly what to put where) [19:30] Can I try dmesg without the connection being in place? [19:30] Unfortunately, my wireless reset on me and I cannot reconnect so I either have my internet connection here, or the crossover connection. [19:30] phylogenesis: at anytime.. dmesg tell about stuff that happens in kernel. so if somereason network interface dies or something it there [19:31] on which machine, the server? [19:31] phylogenesis: or both [19:32] what am I looking for, and would there be anything considering I have not retried the transfer since both computers were restarted? [19:32] (Btw I have physical access to the server) [19:33] phylogenesis: something about eth0/1 .. or about ssh server [19:34] phylogenesis: then you could offcourse use mii-tool to check that cards have negotriated right speed [19:35] I did dmesg | grep eth0 and I got this line for some reason: "eth0: Broadcom BCM4328 802.11 Wireless Controller 5.10.91.9" even though this is my ethernet card that I'm connected to the internet with right now [19:35] iwconfig says: "eth0 no wireless extensions." [19:36] phylogenesis: you probably have eth0 and eth1... [19:36] on my client machine, yes [19:37] I'm checking the server dmesg output now [19:37] what mii-tool says about speed [19:38] phylogenesis: btw... ubuntu changes interface names after boot, so that they can be same at every boot [19:38] Client: "SIOCGMIIPHY on 'eth0' failed: Operation not supported" (even as root): Server: "eth0: no link" [19:39] phylogenesis: you should get link up down notice from dmesg if you take wire off [19:39] ifconfig on the server gives eth0 and lo. on the client it gives eth0, eth1, and lo. I have no idea what eth1 is for. [19:39] checking [19:40] phylogenesis: client probably has wireless and wire network cards on [19:40] yes, but wouldn't that be wlan0? [19:41] dunno.. its dosent need to be names as wlan [19:42] Oh, you're right, even iwconfig shows it as eth1, anyway: On the client dmesg gives me "eth0: link down\neth0 link up", and the server gives "eth0: link down". However, they are not connected right now. I'd have to disconnect from the internet to connect them. Should I test it now? [19:43] More specifically: the last lines of dmesg on the server are: "[##.###] eth0: link down" then "[###.###] ACPI Error (evevent-0303): No installed handler for fixed event..." [19:45] Anyway, I think I'm getting off track. What should I test? Perhaps connect the computers, then set the IPs as before, then run mii-tool and pastebin the output? [19:45] should I set my server to have a static IP (in /etc/network/interfaces) instead? [19:46] phylogenesis: probably better [19:47] simplexio, should I also do that on the client? [19:47] phylogenesis: yeah [19:47] Any idea where to set it in Fedora 11? [19:47] I'll figure it out or ask elsewhere if not, I'm just wondering [19:48] phylogenesis: just call /etc/init.d/networking stop, then do ip addr add 192.168.0.1/24 dev eth0 (or what eth interface it is ) and same in server [19:48] phylogenesis: no idea.. easiest way probably is just call network starting init script to stop [19:49] phylogenesis: or if you want you could just add eth alias to those interfaces and define your own network into that === xorigin_ is now known as xorigin [19:50] How would that help? (I really mean to ask, not to suggest you are wrong) [19:50] phylogenesis: i have noticed that ubuntu dosent like me if i define manually stuff whit ifconfig, but ip addr works fine [19:50] Can I have my old dhcp configuration, then use the alias for a static one? [19:51] phylogenesis: if you still have dhcp call there when you define interface manually, it it could change to some default after it timeouts [19:51] phylogenesis: yes [19:52] simplexio, now I'm confused :-/ What should I be trying? The static IP on both, the dhcp with alias, replacing the dhcp with the alias? [19:52] phylogenesis: in buntu /etc/network/interfaces.. line: auto eth1:1 \n iface eth1:1 inet static .... etc etc . [19:53] in that case attleast dhcp stuff wont be reason why it dosnt work [19:53] phylogenesis: i assume that you wire interface is eth1, in that case you define that auto eth1:1 lines to both systems ( no idea howto do it in fedore ) [19:53] and I use eth1 even though ifconfig doesn't show eth1, right? [19:54] my wire interface is eth0 on both [19:54] phylogenesis: no, you use whatever is you wire interface.. i that ase eth0:1 [19:55] Okay, I will try that and come back. It'll take a good 20 minutes I think. Thanks. [19:55] phylogenesis: and chech that wire is ok [19:55] How? [19:55] I only have one crossover, and I can't even ping each other using standard ethernet. [19:59] This'll be my new interfaces file, is this correct or am I doing something stupid :-P: http://pastebin.com/d22a5ce85 [20:00] pretty much like that [20:00] and how do I indicate that the scp should use eth0:1 instead of eth0? [20:00] will the server just know to use that when it receives input from the connection pointed toward the static IP? [20:00] phylogenesis: scp user@host , just like before [20:01] you justkeep your old dhcp config on side and "working" when you define another ip on same interface [20:02] I'm gonna try to figure out the fedora side of this now [20:02] may or may not fix the problem : ) its computers [20:06] Should I use address=192.168.2.101, subnetmask=255.255.255.0, and gateway= for the client? In network configuration, I chose eth0, which is dhcp, but there is a tab "Route" that says: "Static network routes" so I clicked to add one and it asks for those things. [20:12] sure [20:15] andol: what's the magic for disabling recommends to be depends again? [20:17] andol: nvm. found it. [20:43] Here is the output from "sudo ethtool eth0" on both machines: using mii-tool on the client didn't work so someone suggested ethtool http://pastebin.com/d54b6387d [20:45] simplexio, sorry, I left a different channel and it totally disconnected me instead [20:45] phylogenesis: use irssi ;) [20:46] phylogenesis: its seems that its negotiated just right.. is it still doing those stalls [20:47] simplexio, I might, I don't like xchat much. As for the stalls, yes. This time it didn't take very long at all, maybe one minute. I didn't start the scp until being connected for a few minutes. It seems like it disconnects after a certain amount of time from being connected. [20:48] Also, I still had to do the ifconfig thing on both machines before I could even ping one another [20:50] phylogenesis: it should be that hard... [20:51] phylogenesis: i mean disable dhcp / manually configure interface to same subnet, (maybe restart openssh server ) and use scp * user@host [20:51] simplexio, as in, "figures" or literally [20:51] maybe fedore has some broken config :) [20:52] I worked on that last night, according to the people on IRC in #fedora, I have everything set up like I should [20:52] i mean have two buntu in use here and i havent see that kind problem for long time [20:52] phylogenesis: dmesg or cat /var/log/syslog.0 dosent tell anything funny ... [20:53] phylogenesis: do you have TCPKeepAlive Yes in sshd_config ... that pretty muhc only option that could affect [20:54] nothing that means anything to me. the last line of dmesg is eth0: link down, but I had to disconnect the two to get back on here [20:55] phylogenesis: does ifconfig report about collision or overruns [20:55] all 0 [20:55] (checking TCPKeepAlive) [20:56] then i have no other idea that you network cards are bad or their driver are bad [20:56] where is sshd_config? [20:56] /etc/ssh/ [20:57] I see "TCPKeepAlive yes" :-/ [21:00] Well, I'm baffled. Thank you for all your time. I'm gonna try it with a friend's switch. Hopefully that'll help... [21:01] phylogenesis: im just testing it. and it just works scp -4r user@host: . [21:02] What is -4r? (I use just -r) [21:02] force ipv4 addr [21:02] Mind that I've had it run for up to five minutes and transfer more than a few GB of data before failing. [21:02] Any chance that would help? [21:03] Also, do you know of a way I can test the cable itself? [21:03] if dmesg dosent say that hd timeouts its should be problem [21:04] you could try run rsync [21:04] attleast it can restart where it left it [21:04] btw.. id few GB is one big file or many small [21:05] im testing with many small [21:05] No thanks, I want to get it working. Who knows what other problems it'll cause if I don't. [21:05] it's both, I have tons and tons of small files plus several very large ones [21:05] i can test with big file [21:06] I may ask you to later, but I'm going to try one more time with -4r and if that doesn't work I'm gonna head to a friends place and have him help. [21:06] between 2 ubuntu boxes if it makes same on my computers [21:06] Thanks a million though, I'll let you know what happens if you wish. [21:08] 1% 2148MB 26.7MB/s 1:09:31 ETA.... going strong [21:09] nice connection, the best mine showed was 20MB/s, anywho I'm gonna try that -4r thing now, I'll be back on later [21:42] hi, i have a few questions regarding mail servers and ubuntu [21:42] !ask | linkxs [21:42] linkxs: Please don't ask to ask a question, simply ask the question (all on ONE line, so others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) [21:43] first of all, how would i make my ubuntu server forward to another smtp server, for example, i want to specify smtp.linkxs.org in the mail client, and my server would forward to smtp.sbcglobal.net [21:44] It depends a lot on what mail transfer agent you select. We generally recommend Postfix. [21:44] I've just finished installing dovecot [21:44] as far as i understand, dovecot doesn't do smtp [21:44] Dovecot is the mail delivery agent. [21:44] Postfix does. [21:45] so should i install postfix alongside dovecot and configure it to forward to another smtp server? [21:45] Yes [21:45] However .... [21:46] however...? [21:46] Any spam you forward is going to look to sbcglobal like it came from you, so if you don't want to get blacklisted by them, make sure you have that worked out first. [21:46] ah [21:46] good point [21:46] thanks [21:46] i think i might jsut get lazy and specify smtp.sbcgloabl.net [21:46] thanks though [21:47] another question, i've just installed dovecot(as i said), and I want to configure squirrelmail. #ubuntu threw this link at me: https://help.ubuntu.com/8.04/serverguide/C/email-services.html about dovecot, but i don't see anything for squirrelmail tehre [21:48] is there a good tutorial for squirrelmail? [21:48] I haven't setup squirrelmail. [21:49] ok [21:50] another question.. after I isntalled dovecot, I tried doing 'telnet 192.168.1.8 pop3s' from another computer [21:50] Trying 192.168.1.8... [21:50] Connected to 192.168.1.8. [21:50] Escape character is '^]'. [21:50] is what i got, but i can't quit it now [21:50] ctrl+] does nothing, just prints it [21:50] same goes for imaps [21:56] ok, another question: i've followed this tutorial: https://help.ubuntu.com/8.04/serverguide/C/dovecot-server.html , what else do i need to do to be able to use, let's say, imap? [21:59] dovecot does imap too [22:00] yeah, i know [22:00] i plan to use imaps and pop3s [22:00] oh [22:01] you have to set up certs for that [22:01] i did [22:01] self-signed [22:01] ok [22:01] but what else do i need to do after that tut? [22:03] linkxs: squirrelmail acts as an imap client [22:03] hang on [22:04] hmmm [22:04] that tutorial doen't have much in it [22:04] but it should work [22:05] yea its got protocols = pop3 pop3s imap imaps [22:05] spowers: i know, but before i deal with squirrelmail, i want to get dovecot to work with, say, thunderbird [22:06] tosh: i only enabled pop3s and imaps [22:06] and and the ssl stuff is there [22:06] yes, i did do the ssl stuff [22:06] ok [22:06] so it should work [22:06] don't i need to create users? [22:06] well yeah [22:07] how would i go about doing that? [22:07] you have postfis or some other mta? [22:07] uhm, no [22:07] i was planning on using squirrelmail [22:07] how are you actually receiving mail? [22:07] dovecot? [22:07] as i understand, dovecot is the pop3 server [22:07] yup [22:08] dovecot will read mail from a maildir or mbox mail store and serve it to imap or pop clients [22:08] therefore, it handles it [22:08] you need some way to get mail into those mail stores [22:08] e.g postfix [22:08] oh [22:08] but its not a mal transfer agent [22:08] ok [22:08] gotcha [22:08]