[00:00] <sarnold> yeah; but containers can overprovision even further because everyone shares the same kernel, probably shares libc6 and so forth...
[00:02] <mvhenten> ok let me look around in their web panel thing
[00:02] <mvhenten> I might now know where to start looking...
[00:03] <mvhenten> parralells...
[00:05] <mvhenten> thing is
[00:05] <mvhenten> nmap tells me smtps is closed
[00:05] <mvhenten> i guess because ufw is not running
[00:05] <mvhenten> enabled
[00:08] <mvhenten> ok i'm gonna contact the hoster. maybe I can upgrade the kernel
[00:08] <mvhenten> seems like that could solve the problem
[00:09] <sarnold> they may just be blocking standard smtp
[00:09] <mvhenten> and explain why every hit on google refers to issues around 2009~2010
[00:09] <sarnold> it wouldn't hurt to have a converation :) hehe
[00:09] <mvhenten> yeah, trying to use smtps actually
[00:09] <mvhenten> really sucks using mutt over ssh
[00:20] <mvhenten> hm.... plain old iptables don't work either
[00:20] <mvhenten> must be something larger at work
[00:27] <mvhenten> ok thanks you guys, off to bed then
[00:28] <keithzg> Hrmm. New laptop at work, trying to set it up...but no response from our DHCP server (running 12.04).
[00:30] <keithzg> Wireshark shows it calling out requests, and no one answers.
[00:31] <sarnold> do the logs on the dhcp server indicate why it might be neglecting to respond? can you run wireshark on the server and see if it receives the requests?
[00:32] <sarnold> can you throw a hub or span port between the two systems and see the packet on the network? One common source of dhcp problems is from silly nic offloading not properly filling in the udp checksums..
[01:03] <keithzg> I've been running wireshark on my desktop, and I see the requests, but no response. Hmm. I could probably go and hook them up if there's space on the switch in the server room...it IS a laptop, after all.
[01:04] <keithzg> To clarify, from my desktop, I see the DHCP requests from the laptop, but no ACK from the server.
[01:04] <sarnold> hrm, the response might be sent back directed to the ethernet MAC that requested the address...
[01:07] <keithzg> Yeah, fair enough.
[01:08] <keithzg> Installing Wireshark on the server now (strange that I hadn't had to before, actually, heh).
[01:08] <sarnold> you might have just used tcpdump? :)
[01:11] <keithzg> ...d'oh!
[01:18] <keithzg> Hmm. I'm still only getting the requests from the client, but perhaps I'm only filtering for that? I must admit I'm a bit inexperienced with tcpdump.
[01:19] <keithzg> tcpdump -i br0 -vvv -s 1500 '((port 67 or port 68) and (udp[38:4] = 0x6E11BA57))'
[01:19] <keithzg> The laptop's MAC is 08:60:6E:11:BA:57
[01:28] <keithzg> Hmm, no, that's not the issue. Even if I just fire up Wireshark on the server now, and filter purely for "bootp", looking at it I can see the DHCP Discover requests from the laptop (which is now hooked up to the same switch as the DHCP server) but no Inform or ACK with the same transaction ID.
[01:29] <sarnold> keithzg: cripes :(
[01:42] <keithzg> :(
[03:09] <keithzg> Still stumped. Sigh. Guess I'll try and tackle this again tomorrow. Hope this isn't the herald of a deeper issue . . .
[04:11] <rostam> HI I have installed ubuntu 12.04 update 3. when system  boots , I can see the Ethernet MTU size is set to 574 bytes (ifconfig), Is this value set by defaults? thx
[04:12] <jdstrand> sarnold: not sure if you mentioned it, but I like to point people at 'sudo /usr/share/ufw/check-requirements'
[04:12] <jdstrand> sarnold: some VPSs don't give everything unfortunately
[04:12] <jdstrand> fyi...
[04:30] <babinlonston> I Have Installed Ubuntu 12.04 in LVM and if i need to Create a New Virtual machine , How can i take snapshot and what are the steps there to take a snapshot , How can i create a New VM from snapshot
[07:26] <sarnold> jdstrand: oh right! I forgot about ufw's check-requirements, that's way easier than trying to recall what's used :) hehe
[07:26] <sarnold> jdstrand: thanks
[07:28] <makara> i want browse (Firefox) to a certain server, so the curl equivalent: "curl -H "Host: specialserver.website.com" website.com
[07:28] <makara> how 2 do that?
[07:43] <andol> makara: I assume the reason for you not simply putting specialserver.website.com in the url bar is the lack of such dns entry? How about just making an entry in your /etc/hosts file?
[07:44] <makara> andol, correct, but there are multiple servers.
[07:44] <makara> so hosts file is clumsy
[07:45] <sarnold> then set up dns?
[07:45] <makara> I see the Modify Headers extension does the trick
[07:45] <makara> at least I don't need to root everytime I want to switch
[07:45] <makara> sarnold, corporate hubris
[11:23] <makara> sarnold, can you recommend a DNS server?
[11:27] <ikonia> ISC bind is pretty much the most common usd
[13:07] <zetheroo> how can you see if the network connection on a server is being overloaded?
[13:07] <remix_tj> zetheroo: i use iptraf
[13:08] <zetheroo> is there a tool which can show network bandwidth usage live?
[13:08] <remix_tj> iptraf
[13:08] <remix_tj> but is a good idea to keep monitoring configuring mrtg
[13:08] <zetheroo> ok
[13:10] <zetheroo> can iptraf show a graph of the bandwidth being used?
[13:11] <remix_tj> uh, a graph no, but some infos for sure
[13:11] <zetheroo> ok
[15:32] <jrwren> rostam: re the 574 MTU, i've seen some DHCP servers suggest that. I usually override it. Is it your cable modem interface doing that?
[15:35] <i3luefire> i am having trouble logging in via ssh on my 12.04 server. it has worked plenty of times in the past with various forms of auth but now nothing works.
[15:57] <jamespage> zul: how many new MIR's so far for icehouse
[15:57] <jamespage> ?
[15:57] <zul> jamespage:  two so far not including deps of the new MIR
[16:03] <jamespage> zul, a few on this list - http://people.canonical.com/~ubuntu-archive/component-mismatches-proposed.svg
[16:03] <jamespage> I think most of those are openstack deps right?
[16:04] <zul> jamespage:  afaik roaksoax is handling kombu but ill doublecheckdjango-compressor
[16:04] <jamespage> sweet
[16:04] <jamespage> zul, just thinking about icehouse1
[16:04] <zetheroo> we have a server with bonded ethernet's  ... I followed this to setup https://help.ubuntu.com/community/UbuntuBonding#Ethernet_Bonding_modes
[16:04] <jamespage> start of december
[16:04] <zul> jamespage:  yeah
[16:05] <zetheroo> however I am not sure about this line under Mode 4: Prerequisites: Ethtool support in the base drivers for retrieving the speed and duplex of each slave.
[16:05] <zetheroo> how does one find out if Ethtool is supported by the driver being used?
[16:05] <jamespage> zetheroo, query your physical interfaces using ethool
[16:06] <jamespage> if it reports information about speed and duplex - alls good
[16:06] <jamespage> :-)
[16:06] <zetheroo> never used ethtool ...
[16:06] <zetheroo> installing it now ...
[16:08] <zetheroo> ethtool output ... looks good no!? http://paste.ubuntu.com/6490117/
[16:10] <Gorash_> Dag! Ik vroeg me af of iemand zin heeft om mijn settings voor postfix (vps) na te lopen. Sommige emails komen niet aan namelijk!
[16:10] <Gorash_> I mean, does someone have time to check postfix settings with me, some of mail emails are not being delivered. Sorry for dutch ;)
[16:11] <zetheroo> using iptraf it just seems like eth0 is getting 90% of the network traffic ....
[16:11] <zetheroo> jamespage: ^
[16:11] <jamespage> zetheroo, yeah - that looks OK
[16:12] <jamespage> zetheroo, might be misconfiguration switch side/
[16:12] <jamespage> ?
[16:12] <zetheroo> eth0 will be having about 1800kbits/sec while eth1 is sitting at less than 100 ...
[16:13] <zetheroo> I guess ... could be ... though on the switch side it's pretty much just LAG groups ... from what I saw anyhow ...
[16:13] <jamespage> zetheroo, can you check /proc/net/bonding/bond0
[16:14] <zetheroo> http://paste.ubuntu.com/6490144/
[16:15] <jamespage> zetheroo, that output looks OK - I would suspect switch config
[16:15] <jamespage> but that's obviously harder to diagnose
[16:16] <zetheroo> ok
[16:56] <rbasak> zul: working on apache2. Just noticed that you TIL.
[17:37] <zul> rbasak:  be my guest
[19:37] <Maddeth> Hi all, I am looking for advice on capacity planning. I am looking for some sore of software, or even a nagios plugin that can help me predict if and when my VSphere datastores are going to run out of CPU, RAM and Disk capacity. I have been writing a web front end to pull disk data from nagios based on the host and graph it, I could do the same for RAM and CPU easily enough, but I want to compare it to the clusters they are 
[20:05] <blueking> remote desktop connection from windows to ubuntu-server (installed gnome) with xrd
[20:05] <blueking> xrdp
[20:05] <blueking> doesn't work
[21:17] <blueking> hello again
[21:17] <blueking> ubuntu server almost finished :)
[21:18] <zerick> blueking, from Windows to Ubuntu, VNC maybe suits for you ?
[21:18] <blueking> just miss some stuff,  thinking about to install something so windows can map network drives on ubuntu   what are recommended to use ?
[21:19] <blueking> zerick:  managed to get xrdp work
[21:19] <zerick> blueking, samba
[21:19] <blueking> ok
[21:20] <zerick> or NFS,  aswell
[21:20] <blueking> what's fast reliable ?
[21:21] <zerick> AFAIK NFS it is
[21:29] <pmatulis> windows desktop can act as a NFS client?  nice
[23:18] <zerick> pmatulis, yes, it can. You just need to instal (or enable) NFS
[23:19] <zerick> s/instal/install