[00:51] smoser: http://paste.ubuntu.com/400294/ === jaraco1 is now known as jaraco === MTeck is now known as MTecknology === bladernr__ is now known as bladernr_ [01:42] i'm using lampp.. and i get two different hosts for " localhost " and " 127.0.0.1 " shouldn't they be same ? [01:46] not necessarily [01:47] 127.0.0.1 will serve you the first virtual host declared, while localhost will serve you the host declared with Servername localhost [01:48] yann2: ok.. [01:49] vhosts piss me off. [01:49] twb ? [01:49] I shouldn't have to add entries to my laptop's /etc/hosts or write GET requests directly just so that the site will work through an SSH port forward. [01:50] uh? what are you trying to do? [01:53] Access an internal vhost from outside a LAN by hopping through a bastion. [01:54] e.g. ssh -fNL 8080:porn:80 example.net && w3m http://127.0.0.1:8080 will fail, because w3m will send "Host: 127.0.0.1" which doesn't match the vhost "porn". [01:55] The godsawful workaround being to make "porn" resolve to "127.0.0.1" locally, and then tell w3m to go to "http://porn:8080/" [01:59] I'm afraid you might be right :) [01:59] on the other hand its your fault if porn doesnt resolve locally, as really it should ;) [01:59] smoser: http://paste.ubuntu.com/400316/ [02:01] yann2: it's a weak point if DNS failures also break the HTTPd. [02:03] twb, well you're right again but then that's how much of the internet work 'plus it's up to you to have enough dns servers :) [02:04] "That's how much of the internet works" isn't an excuse. [02:05] Rather, it's an invalid excuse. [02:13] kirkland, http://paste.ubuntu.com/400322/ [02:18] smoser: http://paste.ubuntu.com/400323/ [02:32] I just ran smartctl -t short on my drives, and it's giving me a LifeTime(hours) of 2600. Is that the estimated time until the drive fails, or just how long it's been running so far? [02:34] trimeta: you need a big bag of grok to understand the numbers that come out of SMART [02:34] But based on what you've said and my memory, I'd expect lifetime is the number of hours the disk has been running so far -- not how long it has left. [02:35] twb: Cool. [02:35] The drives I bought have some negative reviews, and I figure I should probably be monitoring them to make sure they don't die when I least expect it. [02:35] Unfortunately I don't have any good references to give you (re grokking) [02:36] Oh, these are new drives? Then 2600hrs runtime wouldn't make sense. [02:36] No, it's been a couple of months since I got them. [02:36] Beginning of December, I think. [02:36] That'd be about 2900 hours if they were running continuously [02:39] WolframAlpha gives me 2592 hours since December 5th, which is when the RAID says it was first created. So it's probably fine. [02:40] trimeta: LifeTime is how long they've been running for. [02:43] What is a WolframAlpha? [02:43] soren: thanks. [02:44] twb: Wolfram's "knowledge engine;" there was a bunch of hype when they released it about how it was better than Google, able to actually understand your searches and find the answer from its own database; in practice, it's useful if the question you've got can be answered definitively with some calculations. [02:45] So time between two dates, what day of the week something falls on, that sort of thing is useful. [02:45] I would normally use Emacs for those [02:45] Or GNU date [02:46] $ date -d '1 Jan 1970' +%a ==> Thu [02:47] I don't know how to do P intervals, but you can kludge it thusly: [02:47] Well, it does things other than date stuff, too...I can't remember anything useful, though. [02:47] $ sh -c 'TZ=GMT date -d @$(($(date -d "$2" +%s) - $(date -d "$1" +%s))) -Iminutes' -- '22 Jan 2010' '18 Oct 2010' ==> 1970-09-27T00:00+0000 [02:47] ...which means 9 months, 27 days (more or less) [02:48] Easier to let someone else parse my human input string and give me the answer. [02:49] FSVO someone = proprietary, unaccountable third-party [02:49] This is purely factual stuff, I'm not concerned about them fucking up horribly. [02:49] Plus this is the company behind Mathematica. They know math. [02:50] I wouldn't trust them as far with encyclopedic information, but I'll believe in their internal Gregorian calendar. [03:16] i am failing to start apache2 [03:16] [Tue Mar 23 23:15:58 2010] [error] Unable to configure RSA server private key [03:16] [Tue Mar 23 23:15:58 2010] [error] SSL Library Error: 185073780 error:0B080074:x509 certificate routines:X509_check_private_key:key values mismatch [03:17] any help would be appreciated [03:36] it was working until thursday [03:51] Hi guys. I've had some filesystem inconsistency which fsck has fixed, but now when booting the process is taking very long at 'Checkign qoutas..' [03:51] Any idea how I can either skip Quota checking, or is it normal to take that long? [03:53] I am getting ready to run "euca-run-instance -k keyname " where can I find the keyname, or create one? [03:54] euca-describe-keypairs [03:55] "EC2_ACCESS_KEY environment variable must be set" [03:57] I just ran "euca-add-keypaid mykey > mykey.priv" [03:58] ran a couple more commands and get the same environment variable must be set [04:00] well, do you have it set ? [04:00] I am not sure what to set it to [04:02] you generally set it by sourcing the eucarc that the euca admin web pages give you in a zip file [04:02] follow the general getting started stuff if you don't know what I'm talking about [04:03] Thanks, probably should've done that in the first place. Thought it would be as easy as downloading from the web interface [04:07] i just mounted a ext hdd to my ubuntu server i need help because it gives me the total and I formated the 320gig hd and without the ext hd it was about 280gig and now its 586gigs but with 568gigs used how can i get it to give up more space thats its not even using [04:08] i am doing all this in the command line and if there is any way i can get some help by leading me to a website [04:10] baffle: can you plz help me [04:11] Xodiac13: Did you add a file system to the ext hd? [04:11] rap424: no i just want to add extra space to the server its on a laptop and i want to add a 329gig ext hdd [04:12] i did follow this https://help.ubuntu.com/9.04/serverguide/C/httpd.html [04:12] still failing to start apache2 ssl [04:12] rap424: i mounted it but it shows that 560 gigs total and 530 gigs used i just formatted the hd [04:12] is the cloud the same as google os? [04:13] i meant similar [04:13] run "mkfs.ext3 /dev/sd*" [04:13] rap424: okay [04:14] Xodiac13: make sure to subsitute sd* for the ext hd, and unmount it first [04:14] rap424: okay [04:14] lifeless: Would the keypair end with .pem? [04:15] no, you'll nee dto create it [04:15] rap424: i just unmounted and did what you said and it said it needed to be unmounted? [04:15] the command you ran before errored ;) - check the file it created [04:16] rap424: somehow its still mounted with the command [04:17] Xodiac13: When you ran fdisk did you just erase all the partitions? [04:17] rap424: no i get this /dev/sda1 is mounted; will not make a filesystem here! [04:18] Xodiac13: I don't think the ext hd will be sda1, that is generally your first hd [04:19] rap424: thats wierd cause hda1 is the filesystem [04:19] He might have PATA disks and a pre-libata release [04:19] rap424: thats how i got it mounted by using sda1 [04:19] rap424: i also renamed the mnt extdisk [04:20] Xoadiac13: Run umount extdisk [04:20] rap424: k [04:21] rap424: it says its not mounted [04:21] rap424: or not found sorry [04:21] Xodiac13: When you run "ls /dev/sd" what comes up? [04:22] lifeless: cat eucarc >> .profile ?? [04:22] rap424: i get this ls /dev/sda1 [04:23] Xodiac13: That is all that is listed? [04:23] rap424: yeah [04:23] rap424: im runing ubuntu server 8.04 [04:23] Xodiac13: I am guessing twb might be right [04:23] rap424: twb? [04:24] Xodiac13: Are your disks PATA? [04:24] rap424: im thinking they are [04:24] Xodiac13: There should be more partitions on your drives, unless you manually created partitions during startup and didn't make a swap partition [04:25] Xodiac13: Try "ls /dev/hd*" [04:25] rap424: i just used the auto install [04:25] rap424: it says its not found [04:25] rap424: or add '. path_to_eucarc' to .profile [04:26] rap424: im using webmin if that makes a difference from actually being on the server [04:27] Xodiac13: I have never used webmin, so not sure [04:27] rap424: is there anything i can do to get my ext hdd working for extra space [04:28] Xodiac13: I am sure there is, but I am not sure why there is only one partition showing up in /dev [04:28] rap424: uh [04:29] Xodiac13: On a manual install Ubuntu should create a primary partition and a swap partition, usually /dev/sda1 and /dev/sda2 or /dev/hda1 and 2 [04:30] rap424: uh thats wierd [04:30] rap424: is there anyother way to format it and then after that it should work just fine? [04:31] Xodiac13: All I know of is fdisk [04:31] rap424: okay i will try [04:31] Xodiac13: If you have a windows box you can double check there [04:32] rap424: what program can i use for windows to format it to ext3 [04:32] Xodiac13: I don't think there is one, but you can format it to FAT. Linux will read FAT [04:33] Xodiac13: or just format it to FAT, and when Linux recognizes it format it to ext3 [04:33] rap424: okay but with FAT i dont think it can use 320gigs though [04:34] Xodiac13: but at least you will make sure you formatted the right drive with fdisk, it will be labeled as a FAT drive [04:35] rap424: true [04:45] lifeless: Not enough resource :( [04:45] lifeless: ./eucarc didn't even work, had to copy and paste [05:10] what's the channel for ubuntu sysadmins? [05:10] the planet RSS fails to validate [05:13] yeah, it's coming back empty [05:14] been that way all day [05:15] qman__: its not empty [05:15] qman__: it just doesn't validate [05:16] if someone cleared the planet cache it would probably pick right back up [05:17] ah === swift_ is now known as swift === dendrobates is now known as dendro-afk === swift_ is now known as swift === dendro-afk is now known as dendrobates === alvin_ is now known as alvin [08:25] is this the right channel for UEC issues, or is there a better place for that? [08:30] rcsheets: yes, this is the right place to ask for help on Ubuntu Enterprise Cloud (Eucalyptus) if that is what you are talking about [08:31] cool, thanks. sorry that was kinda asking to ask :) [08:31] lol [08:37] i've done the cloud controller install, which seemed to work fine, and i rebooted. i was under the impression that the first node i install would get its network configuration via dhcp, with the cloud controller being the dhcp server, but there doesn't seem to be a dhcp server running. [08:38] in fact, when asked which interface would be used to communicate with nodes, i said eth0, and eth0 didn't have an ip address at all [08:38] rcsheets: did you select a DHCP server to be installed while installing Ubuntu Server on the cloud controller? [08:38] no, was that somethingg i needed to do manually? [08:38] rcsheets: no [08:38] i just accepted defaults for virtually everything [08:39] rcsheets: eth0's configuration shoud have been configured during install too [08:39] should* [08:39] i said eth1 was my primary network connection, and that the nodes would be on eth0 [08:39] ah [08:39] eth1 is using dhcp (served by my router) and that is working fine [08:39] rcsheets: you will have to configure eth0 manually by editing the /etc/network/interfaces file as superuser [08:40] ok. did that. [08:40] rcsheets: did you ifdown eth0 && ifup eth0? [08:40] do i need to tell something to run a dhcp server there, beyond simply telling the installer that the nodes will be on eth0? [08:40] i rebooted (the machine needed to be moved off the floor :) [08:40] ah lol [08:41] so everything is working but you need a DHCP server to assign IP address to the nodes? [08:41] eth0 now shows both 169.254.169.254/32 and 10.10.10.1/24 [08:42] i am only assuming that a dhcp server should be running to assign IP addresses to nodes. i haven't begun the install on the first node yet. i was trying to sanity check the cloud controller configuration before i went down that road. [08:43] i can install dhcpd if i need to ... i was just kind of surprised it (or something similar but more UEC-themed) wasn't already there, so i thought maybe something was wrong [08:43] rcsheets: obviously not =) sorry, I'm not familiar with cloud computing configuration because I don't use it on any of my servers, but I'll assume installing a DHCP daemon on that server and configuring it will do the trick [08:43] rcsheets: it needs to be specified to be installed at install-time [08:43] ok [08:44] with the "Install Ubuntu Enterprise Cloud" option i had been hopeful that it would install all the necessary software by itself [08:44] but i will go get dhcpd :) [08:44] rcsheets: you can select more than one option [08:44] rcsheets: using the space bar and then pressing enter once you have selected all you need [08:45] oh, dhcp3-server is installed [08:45] Callum__: on which menu do you mean? [08:45] rcsheets: when you go through the Ubuntu Server installer, I mean.. [08:45] the UEC install process doesn't present the task selection menu [08:46] (or i missed it) [08:46] Hmmm [08:46] it seems to be a bit single-purpose [08:46] i think i might try it again with eth0 as the primary interface, in case something is assuming the first interface will be primary [08:46] okay [08:46] heh, my SCSI cables keep getting damaged or something (probably because of the way it is in the case...) so that my array becomes degraded and I have to get a new cable and then rebuild the array... I'm pretty sure its the way I have the cables sitting in the case, I'm going to rearrange my drives so that the SCSI drives are in a much better place where they can't get damaged [08:47] haha [08:47] s/drives/cable/ [08:47] that's an unusual reason for an array to be degraded :) [08:48] rcsheets: I know lol, but this is kinda old equipment so I'm waiting on some 1TB SATA drives to be ordered so I can put this hardware SATA RAID controller I also have in the machine to good use [08:48] cool [08:48] oh one other oddity with UEC i'm noticing ... any idea why some uids wouldn't be resolved to usernames by ps? [08:49] sometime I may be able to "borrow" an external SCSI enclosure to put some SCSI drives in so I can still use my SCSI RAID controller (since that is still fine) [08:49] rcsheets: not sure about that I'm afraid [08:50] if you're curious... http://pastebin.com/vi1jVmSJ [08:52] wow, that is weird... [08:52] rcsheets: maybe its because they are remote users? [08:52] well, are they? [08:52] nope [08:52] hmmm [08:52] 'getent passwd 102' and 'getent passwd 107' do exactly what you'd expect [08:52] 107 is eucalyptus and 102 is messagebus [08:52] because I notice that happens when foreign users put content onto my home server... [08:53] right [08:53] maybe from expanded tarballs with the uids preserved [08:53] yeah [08:53] yeah but these are running processes, and the users are in /etc/passwd [08:53] rcsheets: it that a problem, or it just nitpicking? =P [08:53] is* [08:53] well ... [08:53] it's weird. i wonder if it's a symptom of some larger problem. [08:53] if you know what UID 107 and 102 are, its not a big problem [08:54] well, sure. it's not like a showstopper. [08:54] oh i guess i never mentioned this is on lucid beta1 [08:54] Hah, I see [08:54] it just seems like some polish that would be nice :) [08:55] rcsheets: it might be a good idea to report the bug, it can and might be fixed by the final release... [08:55] unless its been reported already [08:55] ok [08:55] i'll look around [08:56] I defy the laws of the man... I'm multitasking, talking here, fixing a server and doing my Japanese homework [08:56] yeah the closest thing i'm seeing to the task selection menu (which would have the option for manual package selection, i think) is [!!] Select cloud installation mode [08:57] which asks me if i'd like this to be a cloud controller, walrus storage service, cluster controller, storage controller, and/or node controller [08:57] with the first four selected by default [08:57] Hmm [08:57] In which case, DHCP server should be installed by at least one of those, not sure about it being actually being configured [08:57] then i tell it which interface to use to talk to the nodes [08:57] (in this case eth1) [08:58] then i wait forever for mkfs.ext4 [09:02] yeah, then it starts installing packages. there's no opportunity to select extra packages. [09:02] which is probably fine for the UEC installer [09:12] wow, 9999% cpu usage [09:12] reported by top [09:12] taht's a lot [09:17] heh [09:27] hmm yeah [09:28] hmm, whats the plain past form of む verbs... [09:28] i specified eth1 as the interface for the controller to talk to the nodes, but 'sudo grep -ir eth1 /etc' shows nothing even vaguely related to eucalyptus [09:28] that definitely doesn't seem right [09:29] i'll file a bug tomorrow after i investigate a bit more. for now, some sleep. [09:29] thanks for the help Callum__, and good luck with homework [09:29] rcsheets: there is probably some information about setting up Eucalyptus at the Ubuntu and/or Eucalyptus website... [09:29] rcsheets: thanks lol [09:30] yeah it looks like i need to set VNET_PRIVINTERFACE="eth1" [09:30] but it's set to eth0 [09:33] later [09:45] does anyone know how I can view what the current version of a package is from the command line before I install it? [09:49] nevermind, got the answer from someone else [09:49] okay [09:49] Callum__: my ps username resolution bug turns out to be a wontfix. https://bugs.launchpad.net/ubuntu/+source/procps/+bug/14027 [09:51] hi, I've been looking into virtualization, but I'm still confused about it - what are the immediate benefits? [09:52] rcsheets: I see. Well, that's too bad... [09:54] maxo: you can run many different "machines" on a single computer provided you have the hardware, and you can backup, restore and run VMs on different computers easily, as well as do some other fancy things [09:55] maxo: virtualisation is really only advantageous if your machine is actually capable of running more than one or two [09:55] as it can be difficult to set up for users who don't know what they are doing [09:55] and VMs are slower than running on native hardware [09:57] Callum__, I'm considering getting a new server where I work and put Ubuntu on it. But there are 2 issues: the first, the organisation uses Microsoft's ISA server firewall, and I'm worried that configuring Ubuntu to get through that firewall could be a problem. The second issue, it would be good to be able to run ASP.NET applications, and Mono didn't work that well for me. So I'm considering whether it might be worth running Ub [09:59] maxo: ISA can be configured to allow Ubuntu through it (although technically it definitely should give it access without any configuration as all the protocols are operating system independent) [09:59] maxo: I'm afraid ASP.NET requires IIS, which as I'm sure you know is Windows only and always will be [10:00] oh wait [10:00] Mono can run ASP.NET? [10:00] heh, never knew... [10:00] Callum__: yes it does, but I had problems testing it out. I got some things working, but not quite how I wanted. It's not ideal [10:00] but anyway, an IIS server is always best if you are locked into a Microsoft technology such as ASP.NET [10:01] maxo: you could always virtualise Windows Server lol [10:01] Callum__: as for the ISA server, I think it grants permission on a user basis? But on Ubuntu, I'd have local users trying to connect to the internet, like apache's user account, and then apt does something different I think [10:02] I think connecting through to the ISA server is my main worry. If I can't even use apt - that would be terrible! [10:03] maxo: APT uses direct HTTP downloading... [10:03] Callum__: yes but ISA servers use some strange different protocol to normal proxies [10:03] unless you have some kind of NAC going on which isn't willing to let Ubuntu onto the network, I'd expect it would work fine [10:03] you could always try a live cd [10:04] maxo: well you can always try it out and see what you come up wioth [10:04] with* [10:05] the ISA server definitely requires some kind of authentication. Even on my windows machine logged on, because Google Chrome doesn't fully pick up proxy settings, it prompts for a login when I go to any external website [10:05] maxo: hmmm, so when you try to go to a website it instead asks you to login... [10:06] I think I might need to set up something like this: http://ntlmaps.sourceforge.net/ - but then I would need to configure every account to go through the proxy, and I'm not sure how I'd do that. [10:06] maxo: yeah, Ubuntu would have problems with something like that [10:06] well, APT would [10:15] can i stop dhcp client from setting the default route? [10:25] yes incorrect [10:26] i don't see an option like "no route" in the dhclient.conf file [10:27] look in /etc/network/interfaces [10:27] wait, you're talking about routes, not adresses? [10:28] when you say dhcp as the config method you then defer configuration to dhclient [10:28] that has its own config [10:28] i've stopped it from setting my resolv.conf, now i want to bring up multiple interfaces using dhcp, just need to stop the default route from getting set by dhclient [10:28] do you have multiples interfaces with dhcp? [10:29] i will do once i can bring them up without over riding the default route [10:30] but how do you set the routes then? [10:31] with dhclient, just for only one interface [10:34] looks like this is interesting https://help.ubuntu.com/community/InternetAndNetworking/DualHomedGatewayDHCP [10:37] sort of [10:37] the guy only has one dhcp configured interface [10:37] i am going to have 4 [10:37] Callum__: ok so I looked some more into it, and it seems it's possible to configure an ISA server to allow computers through by IP address. So that would probably solve things. And then perhaps I could even run windows server virtually on ubuntu lol [10:39] he's creating a hook in /etc/dhcp3/dhclient-exit-hooks.d [10:39] maxo: yeah, that would do the trick =) [10:39] well, /etc/dhcp3/dhclient-enter-hooks.d too [10:39] * Callum__ facepalms [10:40] I hate these SCSI drives [10:42] They are the bane of my existence in my home servers... [10:42] why can't they be like my SATA drives, aka "just work" [10:42] there, good push in of everything made it bloody work... [10:43] * Callum__ closes up the case and tidies up everything [11:08] Good morning everyone. Anyone know how to run rsync so that it copies the results to a file but still maintaining verbose on the screen (-v) ? [11:09] geneticx_rs: man tee [11:09] tee - read from standard input and write to standard output and files [11:10] geneticx_rs: http://en.wikipedia.org/wiki/Tee_(command) [11:12] henkjan: thanks [12:04] anyone that knows a tool on how to automatically create a network map, layer two? I guess some switch autodiscovery protocol would do [12:04] RoyK, there are a couple but they are made of fail of and disappointment [12:05] RoyK, you got solarwinds, cisco, 10strike === dendrobates is now known as dendro-afk [13:25] Hey guys, I am making a dd_rescue copy of my failing hard disk's ubuntu server install currently but I was wondering if there is a way to do a mysqldump (to export the databases) without actually having to boot the OS.. aka can I dump them from the image? [13:25] or does mysql actually have to be running to get them [13:27] axisys_, http://www.enterprisessl.com/ssl-certificate-support/server_faq/ssl-server-certificate-apache.html [13:28] djveer: you could collect the data files [13:29] there are KB available on restoring MySQL from data files. [13:29] ninjix: interestnig.. because even dd_rescue is giving me 'bad block' errors tryign to create this image.. so this hdd is probably pretty buggered. [13:30] however the databases aren't very big.. maybe I can get them off there without getting I/O errosr [13:30] I've done it from a failing Win2k box to a new Ubuntu server a year ago [13:30] *interesting, *errors [13:31] djveer: I'm a big fan of Steve Gibson's SpinRite for HD recovery (even if it isn't OSS) it just works. [13:31] Oh yeah .... dang I think I had a copy of that... isn't it a paid for thing now? [13:33] IMHO it's cheap from a cost / benefit analysis of your time and money. [13:34] yeah true [13:34] if you can get your disk to a "readable" enough state, other recovery options may become available [13:40] yeah, thanks for your recommendations ninjix [13:40] this hard drive is really buggered. getting I/O errors like 10% into the dd_rescue transfer [13:40] lol === dendro-afk is now known as dendrobates [13:44] ouch === djveer is now known as djveer_work [14:01] How can I jump to the EOF in nano ? === ChrisRut_ is now known as ChrisRut === robbiew_ is now known as robbiew [14:03] I'm trying to do a little planning for the upcoming 10.04 release. It looks like there's no native Xen Dom0 support, and thus no native 8.04 -> 10.04 upgrade path, for Dom0's. Is that the case? [14:08] bpgoldsb: you are own your own [14:08] I guess the suggestion is to use KVM instead? [14:16] bpgoldsb: Yes, I think so [14:17] bpgoldsb: maybe switch the dom0 to debian [14:18] hi can anyone help me.. i'm stuck configurating FW and mysql access [14:19] i've mysql user test set to allow login from % .... and via UFW i've allowed Anywhere ALLOW 192.168.2.0/24 [14:20] if i'm right this should be enough right? [14:20] to allow connections from local lan to that machine wtih that user [14:23] BillyBoy^B: by default mysql-server is not listening on external interface but only on localhost [14:28] hmm ok.. and where can i allow it.. do you know? [14:32] BillyBoy^B: /etc/mysql/my.cnf [14:32] bind-address = [14:35] huh shit mysql is binded by default... omg === dendrobates is now known as dendro-afk [14:38] BillyBoy^B: To the local address right? [14:38] now it works [14:38] henkjan thx === dendro-afk is now known as dendrobates [14:42] henkjan: Do newer Xen kernels talk back to older hypervisors without issue? [14:47] Hi, I'm running a system-wide pulseaudio server to receive sound from computers around the house to play on my speaker system. I'm encountering two issues, one of which is after the server has run for around 24 hours, the network interface on the server stops listening to traffic of any sort(incl. ssh). Restarting networking on it solves it, but it's still fairly annoying. [14:47] bpgoldsb: i'm running the lucid 2.6.32-302-ec2 kernel on my domu vms without any problems [14:48] anyone encountered something similar? [14:48] henkjan: Running Lenny or Etch as the Dom0? [14:48] The server does not have the same problem when I'm not running the pulseaudio server, thus leading me to believe pulseaudio is the culprit [14:49] bpgoldsb: lenny [14:51] henkjan: Did you have to set the clocksource, or was that not an issue for you? === dendrobates is now known as dendro-afk [15:24] zul: thanks === dendro-afk is now known as dendrobates [16:04] how do i make my postfix respond to sender address verification requests? [16:19] anyone know of a updated amazon AMI tool .deb that may be in a PPA somewhere? [16:21] Maybe smoser. [16:24] ok /etc/apache2/sites-enabled/default-ssl is pointing to the following [16:24] SSLCertificateFile /etc/ssl/certs/puppet.crt [16:24] SSLCertificateKeyFile /etc/ssl/private/puppet.key [16:24] zoopster, hold on [16:24] the modulus are matching on key and crt file [16:25] zoopster, https://launchpad.net/~ubuntu-on-ec2/+archive/ec2-tools [16:25] but i am still getting this error [16:25] [Wed Mar 24 00:17:06 2010] [error] SSL Library Error: 185073780 error:0B080074:x509 certificate routines:X509_check_private_key:key values mismatch [16:25] the karmic ones are probably fine for intrepid or jaunty [16:25] any idea why? [16:25] smoser: you rock thanks [16:28] hggdh: hi! [16:28] hggdh: what's up with the UEC test rig? Are you using it? [16:33] mathiaz: yes, I am running the tests [16:33] mathiaz: you need it? [16:33] hggdh: not for noow [16:34] mathiaz: OK. Then we may have a problem. The maximum instances test failed [16:34] hggdh: how? [16:35] mathiaz: all 64 sessions go into pending, then are terminated [16:35] mathiaz: same with 32 sessions [16:36] hggdh: is starting one instance working? [16:37] mathiaz: yes, just did it, then started 7 more, which worked, now I am adding 8 more [16:37] mathiaz: 16 sessions (starting as I described above) working. Will stop all, and start 16 in one single sweep [16:41] hggdh: seems like a good plan [16:43] mathiaz: -n 16 worked. Adding 8 more [16:44] mathiaz: ah, got it: Not enough addresses left in the network subnet assigned to requested group: default [16:44] hggdh: ha - makes sense then [16:45] mathiaz: so we either increase the address allocation, or decrease the stress test [16:45] hggdh: increase the address allocation [16:45] hggdh: hm - I though there was more than 16 addresses allocated thogh [16:45] mathiaz: roj. Should it be documented somewhere? [16:46] hggdh: we should the preseed files [16:46] hggdh: to configure an infrastructure that supports maximum number of instances the hardware can run [16:47] hggdh: the public ips are: 10.55.55.100-10.55.55.240 [16:48] hggdh: that should be more than enough for the testing [16:48] so the CC /etc/eucalyptus/eucalyptus.conf is seeded from there? [16:49] mathiaz: (I see a VNET_ADDRSPERNET="32" there) [16:50] hggdh: right [16:50] hggdh: is 32 the limit you've hit? [16:50] mathiaz: not sure, will check now [16:52] mathiaz: 24 seems to be the limit\ [16:53] hggdh: -n 31 fails? [16:53] hggdh: make sure that there aren't instance running [16:53] hggdh: another way to test it is to create a new security group [16:53] hggdh: and try to start instances there [16:55]