[02:23] can anyone explain me how update-ca-certificates works? [02:24] Does it grab all the files that are not commented out in /etc/ca-certificates.conf? [02:43] Hello. What is the smallest server OS available? [02:43] that supposes a precise answer to a vague question i think [02:46] smallest in MB size, mwhudson [02:46] Joe_knock: but what is a 'server os' ? [02:46] i don't really have the time to talk about this anyway, so ignore my trolling :-) [02:46] mwhudson: The equivalent of ubuntu server [02:58] Joe_knock: a server provides services. If you build a bare host and then only add the serivices you need you should be able to get it rather small [02:58] What is the hardware? [02:58] and define small [02:59] you see, I have a ubuntu 12.04 vm that takes 2GB [02:59] raub: This is what I am trying to do. I want to install owncloud on a 4GB USB flashdrive. it requires the AMP stack and I only have linux desktops. so I want a very thin LAMP stack installed to run owncloud. The reason why (if you ask why) is that I want to learn about building portable web apps. [03:02] Joe_knock: have you seen what people have done to run ubuntu in the raspberry pi or the (lesser known) cubox? [03:03] raub: that is the problem I am seeing. They're all Pi-related scenarios. I don't want to use a Pi. I just want the USB device to act as the way I communicate with owncloud. [03:03] You are thinking I am recommending the pi. I am not. My point is see what they did to make it fit [03:04] How much memory do you have? [03:04] 4GB on the flash disk. raub: I actually think I understand what you're saying now. Hacking their solutions to my scenario [03:04] Yes [03:05] And actually I was meaning how much RAM do you have [03:06] on my PCs, it is generally 1GB or higher [03:06] You see, you need to figure out how much hacking you are willing to do [03:07] The question regarding ram is you want to forget about swap and put /tmp in ram [03:08] This is not even hacking yet since you are still with a standard ubuntu server [03:08] raub that is why I want the thinnest server possible. Perhaps I need to think of a way of building a USB server [03:09] I have not finished yet [03:09] then you need to remove drivers for all the devices you do not use [03:10] But that can be fun since next kernel upgrade they might be back [03:10] But, you can go over all the packages [03:10] find out which ones you need [03:10] and of the ones you think you can get rid of, the ones that are not needed by something you do [03:10] I think you can get down to a few hundred packages if you take the time [03:11] raub: Wouldn't it be easier to go the other way? I determine the packages I need, let the system pickup dependencies and install only what I need (which is basically LAMP) [03:11] Now, if and only if you had loads of RAM, you could upload your webserver (the stuff in /var) to a ramdisk upon boot [03:13] Joe_knock: in principle yes, but even the default is a bit bloated [03:13] Joe_knock: How about this: https://help.ubuntu.com/community/Installation/MinimalCD [03:13] http://www.maketecheasier.com/install-a-minimal-ubuntu-on-old-laptop/ [03:14] i.e. do the manual package selection [03:14] 30MB or less. damn that is small. This could actually work! thanks for that raub! [03:16] Joe_knock: you might want to use nginx instead of apache though [03:16] good point again. If I'm going minimal, I might as well play with nginx too (been meaning too). [03:17] And, for the love of all that is evil and putrid, do not install network manager [03:17] neither it or gdm have any business in a server [03:17] ok, maybe gdm in very specific conditions ;) [03:18] would network manager apply to a LAN server? [03:19] you can do static and dynamic crap without needing it [03:19] and vlans and vpns and all that exciting stuff [03:19] so basically it is bloatware :P [03:19] For a laptop I can see its use [03:20] for a server, you can do better editing a couple of files while making hot dogs === justizin_ is now known as justizin [03:42] raub: 1 last question. Is the core of ubuntu desktop and ubuntu server based on the same thing? Or are these 2 different uses of linux vastly different? [03:46] Joe_knock: pretty much the same thing. To demonstrate this, you could install Ubuntu Server on a machine, and turn it into a desktop with like 3 commands and a reboot. [03:46] sajan: That makes a lot of sense, considering how easy it is to setup a local server environment on desktop. [03:48] Joe_knock: yeah. The only real difference is the packages included in each .iso; The server version excludes unnecessary desktop packages, and the installer. Desktop you get a graphical, GUI installer, on the server version, it's more of a ncruses like installer. [03:49] Joe_knock: In the past, I believe they were two different kernels. However, I'm positive even that is not that case anymore. [03:50] sajan I think they may possibly have abandoned by original desktop kernel and simply built the GUI features on top of the server one (my theory). === markthomas_away is now known as markthomas === markthomas is now known as markthomas_away === markthomas_away is now known as markthomas === markthomas is now known as markthomas_away === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha === alexmoldovan1 is now known as alexmoldovan [09:37] Good morning. [09:51] jamespage: please could you subscribe ~ubuntu-server to bugs in python-jujuclient, urwid and websocket-client? [09:52] rbasak, yes [09:52] Thank you! [09:52] * rbasak carries on with the MIR [09:53] rbasak, done [09:57] rbasak, thanks for picking this up btw - appreciated :-) [10:23] jamespage: np. I've almost done as much as I can I think. MIR reports next. [10:23] rbasak, good-oh [10:29] What's the easiest console only bootable rescue cd? [11:31] b [11:31] ls === chuck_ is now known as zul [11:45] zul, special treat for you when you start [11:45] figure out the failing test in libvirt in the CA :-) === pgraner` is now known as pgraner [12:02] jamespage: that all? :) [12:03] zul, well ceilometer needs a few bits fixing but aside from that :-) [12:03] jamespage: good to be back :P [12:03] zul, most things are working; I've not exercised the cloud-archive that much - mainly focussed on 14.04 [12:03] zul, nice to have you back - how was the beach? [12:03] ok ill take a look today [12:03] it was good...im starting to peel [12:05] jamespage: no problem getting home either..although its -20C outside today :( [12:05] zul, lol [12:05] it was nearly 19C here yesterday [12:05] almost like spring! [12:06] same here [12:06] like the whole winter [12:06] (about 10-15°C to warm since Nov in germany) [12:07] zul, I flushed everything through in the CA - http://reqorts.qa.ubuntu.com/reports/ubuntu-server/cloud-archive/icehouse_versions.html [12:07] (and fixed that report) [12:10] whats up with openvswitch? [12:10] biab i need to take liam to school [12:28] jamespage: sorry -27C with the windchill [12:32] zul, ignore ovs - it ftbfs in virtual ppa [12:33] jamespage: ack [13:00] zul, oh also - for stable/havana [13:00] https://review.openstack.org/#/c/80687/ [13:01] and https://review.openstack.org/#/c/70750/ [13:01] zul, adam_g pinged me last week about that and whether it would foobar us for the SRU [13:01] thoughts? [13:02] jamespage: it shouldnt since we have oauthlib in the archive [13:03] zul, yeah - and its in main - that was my thinking [13:03] but wanted your opinion on it as well [13:03] jamespage: yeah i was the one who asked them to use oauthlib originally ;) [13:03] zul, I remember! [13:04] jamespage: we should just add the build-depends when we are doing stable/havana though [13:04] otherwise the tests will fail [13:04] yes - agreed [13:05] adam_g, OK - so we don't think the oauthlib change in stable/havana will cause issues; we'll focus specifcally on that during testing [13:06] jamespage: i dont think anyone is using oauth on havana in Ubuntu === lool- is now known as lool [13:18] coreycb, https://bugs.launchpad.net/ubuntu/+source/python-neutronclient/+bug/1292661 if you fancy it [13:23] jamespage, sure thing, I'll work on it [13:23] coreycb, ta [14:07] how to override global .bashrc over user .bashrc ? [14:08] raj__: just told you in #ubuntu - please don't cross-post [14:12] whats up guys. [14:13] i feel like an idiot, but i was wondering if someone knew how to make ubuntu boot into the xen kernel.. [14:14] sleepee, Activate grub menu and select the "with Xen entry" or change the grub default to the string of that submenu [14:14] i edit the GRUB_DEFAULT line in /etc/default/grub but ubuntu still boots into the regular kernel. [14:15] well, i really want to change the default so i don' t have to choose the xen kernel manually every time i boot [14:16] sleepee: did you run "sudo update-grub" after editing /etc/default/grub? [14:16] yes. [14:16] GRUB_DEFAULT="Xen 4.3-amd64" [14:17] that's what i have in my /etc/default/grub. [14:17] but i must be missing something else. [14:17] sleepee, Make sure that is what is in /boot/grub/grub.cfg and then re-run update-grup [14:17] update-grub [14:18] not exactly sure what i'm looking for in /boot/grub/grub.cfg but i do see this line: [14:18] set default="Xen 4.3-amd64" [14:19] sleepee, Sorry I meant later in that file where the submenu is defined [14:20] So 'submenu "Xen...' or so [14:20] i see: menuentry 'Ubuntu GNU/Linux, with Xen 4.3-amd64 and Linux 3.11.0-12-generic' [14:20] and: submenu 'Xen hypervisor, version 4.3-amd64' [14:20] So you need that string in default [14:21] ahhhhh... ok. i've been following other guides and how-to's and they just give me the string to put in /etc/default/grub. i guess it would've been more helpful to know where that string actually came from. thanks smb! [14:22] i'm going to edit it and reboot and then come back and let you know. [14:22] sleepee, Yes, unfortunately it changed between releases. Which guide (maybe something I can change) were you looking at? [14:23] here's one i looked at [14:23] https://help.ubuntu.com/community/Xen [14:25] so that means any time a new xen kernel comes out, i'd have to manually change that line in /etc/default/grub right? [14:25] sleepee, No, only if there would be a new version of Xen [14:26] The setup now always picks the kernel with the highest version from that submenu [14:26] right. got it. thanks smb. [14:26] i see. === psivaa_ is now known as psivaa [14:27] well, i'm going to restart and see if this works. be right back [14:34] jamespage: i think i have libvirt fixed [14:35] zul, what was it? I suspected hardening flags in 12.04 but was not 100% sure [14:35] jamespage: the failing test doesnt like Bsymblic-function [14:35] ah [14:35] yeah - that's what I though [14:35] smb, i guess it didn't work. :( [14:35] jamespage: anyways doing a test build and then ill upload [14:35] i'm still running the regular kernel [14:35] zul, good-oh [14:36] but i'm not sure if it's something i'm doing wrong as far as xen goes. [14:37] sleepee, It would be the same kernel actually but just as a dom0 under the Xen hypervisor. But just to make sure, where did you change the string and did you run update-grub after (if it was in /etc/default/grub)? [14:41] i changed GRUB_DEFAULT="Xen hypervisor, version 4.3-amd64" in /etc/default/grub and then sudo update-grub [14:41] and then i rebooted. [14:42] so there's not supposed to be a different kernel when i run 'uname -r'? [14:42] sleepee, And something like "sudo xen info" still does not work? [14:42] sleepee, No, its the exact same kernel version as without [14:42] i get ERROR: Can't find hypervisor information in sysfs! [14:42] ahhh.. ok. [14:42] but still. the xen utilities don't work so i must be doing something wrong. [14:43] sleepee, probably. maybe you can pastebin your /boot/grub/grub.cfg for me [14:44] http://pastebin.com/NCqhtLQX === TREllis_ is now known as TREllis [14:46] am i supposed to put the string that comes after "submenu" or "menuentry"? [14:46] sleepee, Oh drat, fooled myself... Yeah it should be "Ubuntu GNU/Linux, with Xen hypervisor" from the menuentry [14:47] ahh.. ok. so basically, in this case, it should be 'Ubuntu GNU/Linux, with Xen 4.3-amd64 and Linux 3.11.0-18-generic' [14:47] right? [14:49] ok. so i edited the /etc/default/grub file and ran update-grub. i'll restart again and come back in a minute. [14:49] sleepee, No, I think just the string I wrote. Basically the top level menu of the Xen hypervisor entries [14:50] i must not be seeing where that comes from. what line is that from the pastebin? [14:50] sleepee, 220 [14:51] ok. i'm going to change right now. [14:52] ok. reboot time. i'll let you know if it works. [14:52] ok [14:58] anybody using the ec2 package mirror getting signature errors on precise-updates at the moment? I don't *think* it's at my end (have run apt-get clean, cleared /var/lib/apt/lists, etc, with no changes) [14:58] smb, it worked!!! [14:58] cool. And I added a note to the guide (hopefully correct and understandable) [15:00] you just edited the xen documentation? [15:00] Well the community doc, yes [15:02] ok. i see the change. it makes sense now. thanks for your help smb! [15:03] sleepee, your welcome :) [15:03] i've got to go now, but you were awesome! [15:04] jamespage, zul: can one of you review this when you have a chance? https://code.launchpad.net/~corey.bryant/ubuntu/trusty/python-neutronclient/1292661/+merge/211333 [15:05] zul, Just to check, you are not uploading any libvirt for Saucy any soon? [15:05] smb: nope [15:05] smb: unless hallyn has something [15:06] zul, Well I got something we waited for the previous upload to go to updates and I just recently modified the proposed fix for that [15:06] jamespage, zul: ps: do I need to ping you for reviews or is submitting a merge review enough? [15:06] coreycb, done [15:06] coreycb, ping is good at this point in cycle [15:07] jamespage, thanks! Ok [15:07] coreycb, I think zul and I both acked you - merged! [15:07] jamespage: yep [15:08] zul, hallyn kind of has it on his todo. Just wanted to avoid doing too many small uploads if you were about to do one for S, too [15:08] smb: zul: nless i have something for what? [15:08] hallyn: libvirt saucy [15:08] there's an existing upload, [15:08] hallyn, I just saw zul upload a version for T [15:08] other than taht i only had smb's right now [15:09] jamespage: huzzah https://launchpadlibrarian.net/169822752/buildlog_ubuntu-precise-amd64.libvirt_1.2.2-0ubuntu3~cloud0_UPLOADING.txt.gz [15:09] jamespage, zul: thanks! you guys are quick. :) [15:09] hallyn, And note that I just updated the fix for S in bug 1248025 [15:10] smb: yup, saw that, thanks. when the other fix clears, i'll look at that bug for the latest diff :) [15:12] hallyn, Ok, cool. Yeah, just tried to avoid multiple uploads as the previous upload seems to be in updates right now and zul was doing something for T which may have been something to go into S, too [15:12] smb: its not going into S [15:13] zul, /me is always trying to play safe. :) [15:14] smb: i know :) [15:14] zul plays by a more freewheeling set of rules :) [15:14] hallyn: the cloud awaits you :) [15:15] zul, what was the fix? [15:15] jamespage: LDFLAGS = $(shell dpkg-buildflags --get LDFLAGS|sed -e 's/-Wl,-Bsymbolic-functions//g') in the debian/rules [15:15] zul, smb: is it possible to get whatever fix we need for Xen in the CA into trusty as well? [15:16] jamespage: im working on xen now as well [15:16] jamespage, That is already there [15:16] smb, urh - that was for libvirt [15:16] smb: its gone in Xen [15:16] smb, zul has other magic for xen [15:16] jamespage, Just minus a little update to avoid an unnecessary error in libvirts log [15:18] zul, The LDFLAGS magic was gone in S (Debian made some fix to which flags are pulled when) [15:18] smb: hmm...ok [15:18] suck [15:21] id anyone got a copy of The Official Ubuntu Server Book, 3nd Edition, July 2013 [15:22] bought it a while ago, but I'm travelling now and would like to have a PDF of it [15:26] Can someone point me to the maintainer of the precise-updates apt repo that is hosted on Amazon S3? I am getting a signing issue running apt-get update and think there may be an issue with the repository [15:32] Why is lynx/curl not working to outside internet when I am logged in with su - username .. but it's working when I use sudo -u username -s [15:35] seems it's working when I have proper root enviroment only. [15:37] dylanl: I've been trying to find out too, but no luck yet [15:38] Sentynel: thanks. I asked a question on the Ubuntu forums (https://answers.launchpad.net/ubuntu/+question/245627) and AWS forums (https://forums.aws.amazon.com/thread.jspa?threadID=148233&tstart=0). Do you mind commenting so people realize it isn't an isolated issue? [15:39] Am I right in guessing that 14.04 will use a strange mix of sysv+upstart+systemd? [15:40] dylanl: commented [15:40] http://paste.ubuntu.com/7108777/ <-- "System information disabled due to load higher than 1.0 [15:41] but the load is <1 [15:41] Sentynel: thanks. I figured an apt repository used by AWS customers would be a carefully watched thing but I guess not... === markthomas_away is now known as markthomas [16:20] Sentynel: it looks like this might be isolated to us-east-1. If I create a brand new instance in us-west-2 I can run 'sudo apt-get update' successfully. Not sure if that helps you at all [16:20] dylanl: yeah, there's separate mirrors for each ec2 region [16:21] it's not causing me any problems just yet; it's just it's refusing updates on my established ec2 server [16:22] Sentynel: Yeah, wasn't sure if this was impacting all mirrors or just us-east-1. We provision a new box each time we deploy so this is blocking us from deploying at the moment. [16:52] Sentynel: dylanl: are the other references you found current, or are they historical? [16:52] rbasak: somebody else asked in #ubuntu a couple of hours back [16:52] when I first noticed it [16:52] There's a race condition which can get apt stuck into that situation I think (it'll cache the bad or old file). [16:53] OTOH, it could be a problem with the mirror. [16:53] utlemming: ^^ including backscroll going back ~90 minutes. [16:54] rbasak: ugh [16:57] rbasak: Pretty sure this is an issue with the us-east-1 repository mirror. us-west-2 works fine. And sudo rm /var/lib/apt/lists/* -vf followed by a sudo apt-get update doesn't fix the issue [16:58] dylanl: thank you for the confirmation. I think utlemming is looking into it. [16:59] rbasak, dylanl: yes, indeed we are on it :) [16:59] rbasak, dylanl: sorry, and thank you for your patience [16:59] utlemming: Great, thank you! [17:08] Sentynel: fyi - utlemming is looking at this [17:08] dylanl: yup, I saw, thanks [17:09] Sentynel: np [17:09] dylanl, Sentynel: DNS has been updated. As soon as the DNS cache times out things should start working [17:10] utlemming: Fantastic, thanks. Also, for my info, are you with Amazon/AWS or Ubuntu? It would be helpful for me to know where to report issues like this in the future [17:11] dylanl: I work for Canonical/Ubuntu. [17:11] utlemming: thanks. [17:12] dylanl: generally AWS lets us know pretty fast about mirror issues, but in the future #ubuntu-mirrors is a great place. Our IS team watches that IRC sub. [17:12] utlemming: I asked in there [17:13] Sentynel: hrm, well normally that works. [17:14] #utlemming: got it, thanks. I opened a question on Launchpad as well (https://answers.launchpad.net/ubuntu/+question/245627). I didn't see a way that I could close that. Not sure if you can. [17:26] I added a startup script to /etc/rc2.d It has the executable bit but isn't executed when the server boots. Where can I see a log of why it wasn't executed ? [17:26] ubuntu 12.04 LTS [17:27] toyotapie: did you add the corresponding S??name symlinks too? [17:27] yeppers [17:27] S99myscript [17:27] I added the script to /etc/init.d/ and the symlinks in rc2.d [17:27] runlevel says I am in 2 [17:28] if I call it myself, /etc/init.d/myscript start, it stats immediately [17:29] README says I might need the LSB style header [17:34] So the question is where is init logging too ? [17:34] to* [17:36] toyotapie: there's a /var/log/boot.log -- I suspect it isn't as useful as you'd like though [17:36] Just my luck, I don't have boot.log [17:37] toyotapie: and the 'upstart jobs' are logged din /var/lor/upstart/* -- but that doesn't appear to have the sysv compat jobs [17:37] yea, for some reason my ubuntu 12.04 doesn't seem to have upstart [17:37] I am looking at syslog right now. fingers crossed... [17:38] toyotapie: 12.04 uses upstart [17:38] something is horribly wrong with this machine. All my other 12.04 machines have upstart. THis one doesn't... [17:38] lol [17:38] it's not ubuntu [17:39] hehehe [17:39] All my servers except 2 run ubuntu [17:39] one runs centos and the other debian [17:39] this one is debian [17:39] I feel like a complete idiot [17:39] sorry to have wasted your time. [17:39] debian just uses sysv [17:39] so does centos [17:40] I only use centos because of proprietairy software [17:40] and debian for t38modem [17:40] oh [17:40] t.38 [17:41] toyotapie: well now, that just raises new questions :) whyu isn't that machine's sysv-init working? :) [17:41] I've had nightmares about that [17:41] sarnold I wish I knew. I am checking now. [17:45] when I logged out of a ssh connection on remote server, I got message : Received SIGHUP or SIGTERM, what does this mean ? [17:47] I deleted all the /etc/rc?.d/S99myscript files. I did this instead "update-rc.d t38 start 2 3 4". Now when I reboot, the t38 service starts fine. thanks for your help even though it wasn't debian [17:47] ubuntu* [18:04] I created a new motd script and distributed it to all my servers, now it says which O/S is on the machine when I log in :) [18:09] hehe [18:09] toyotapie: haha, nice :) [18:09] I usually know, or I can run a uname -a or lsb_release -a or something to find out if I'm curious ;) [18:10] In my case, it's easy. It's always Ubuntu unless the ssh takes more than 10 seconds to connect, at which point it's CentOs. [18:10] raj__: SIGHUP is sent to processes when their controlling terminal is disconnected -- e.g., closing an ssh connection. you might have had a background task running or suspended or similar. [18:10] I always forget that I have a machine running on debian [18:11] toyotapie: if so, it's probably a dns reverse issue, or there's ipv6 in dns and the server isn't configure correctly [18:11] http://pastebin.com/MvfPjEJn [18:12] I know why centos is slow, it activates GSSAPIAuthentication by default. Once I disable it and 'UseDNS no', it's as fast as ubuntu [18:12] sarnold: so that should be normal when i exit an ssh conn to remote server .. right ? [18:12] toyotapie: then fix your dns [18:13] raj__: I wouldn't say 'normal' but 'not unexpected' anyway [18:13] toyotapie: I have rather a few rhel/centos machines running at work, and it takes a fraction of a second to login [18:13] okay.. [18:13] once GSSAPIAuthentication is off, it takes less than 1 second to login. [18:14] toyotapie: never disabled that on any of my machines [18:14] do you have a dns resolver on each machine or do you have one for all of your servers or do you use ISP provided DNS servers ? [18:15] we have our own dns servers === _KaszpiR___ is now known as _KaszpiR_ [19:10] I am looking for a command that will handle a line of input similar to PHP's preg_match. I want to give a regular expression, and the values that are in brackets are each assigned to a bash variable. [19:10] Is there any command that can do this ? [19:12] I think I may have found my answer [19:14] n/m [19:14] http://stackoverflow.com/questions/2624300/save-part-of-matching-pattern-to-variable === markthomas is now known as markthomas_away === markthomas_away is now known as markthomas === mjohnson15_2 is now known as mjohnson15 [20:56] hallyn: ping [20:56] . [20:56] hallyn: I have an intersting lxc issue [20:56] I'm following bug 1293549 [20:56] trying to skip past the unknowns [20:57] I have an lxc container [20:57] cloned from a precise ubuntu cloud [20:57] I have done the following: [20:57] sudo chown ubuntu:ubuntu /etc/ssl/private [20:58] $ ll /etc/ssl # output snipped [20:58] drwx------ 2 ubuntu ubuntu 4096 Jun 4 2013 private/ [20:58] $ ll /etc/ssl/private [20:58] ls: cannot open directory /etc/ssl/private: Permission denied [20:58] any idea why? [20:58] for a service although a pid file exists but still on attempt to start service, it fails saying "* could not access pidfile for Cassandra"... checked folder permissions under /var/run(owner: root) , the subfolder cassandra is owned by cassandra user but still serivice cannot access pid file.. & the file exists as well..(as pid is shown in cat)..so why says it can't service access pid file? [20:58] if you can put a short set of lxc commands to reproduce it in the bug, that'd be useful [20:58] logged in as ubuntu user [20:58] & running the service as a standalone process just works.. but not using "service start"..! [20:58] hallyn: ok, let me see [20:59] thumper: it could be apparmor messing up (has happened before), overlayfs (has happened before esp with apparmor), [20:59] hallyn: my gut tells me it is aufs [21:00] d'oh [21:00] I just have a call now [21:01] but will try to get a short list of commands after that [21:05] thx [21:19] Some how the enviroment variable $http_proxy is set for normal users(not root).. I've checked with bash -lx and figured out bash didnt set it.. How do I figure out where it's set? (I want to remove it). [21:20] sander__: grep -R http_proxy /etc [21:20] should point you in the right direction [21:21] qman__, thanks alot.. It was /etc/environment === Guest25821 is now known as jrgifford [21:38] Hey [21:38] anybody here? [21:39] I have a question. We have a school project where we have to setup servers and all that so I thought about using Ubuntu Server. [21:39] And I want to use my old laptop or desktop as my server [21:40] how fine will that work? [21:40] and is it easy to stress test? does Ubuntu Server have those tools? === markthomas is now known as markthomas_away === markthomas_away is now known as markthomas [21:53] ilhami: there are many ways to stress a machine; if you run a full compile cycle for something huge, like the linux kernel, you'll have some amount of faith in the hardware, kernel, and toolchain [21:54] ilhami: one machine that I thought was rock-solid for a few years showed trouble when I started asking it to run hour-long compile and package tasks (that was qt4-x11) -- after some way into the process, the machine would either log a non-fatal machine check exception or it would instantly reboot, having suffered from a fatal machine check exception === markthomas is now known as markthomas_away [21:59] dude sarnold: I just have a database and a web service I am going to host === markthomas_away is now known as markthomas [22:04] ilhami: if you want to stress-test a web service, probably the 'ab' tool can be made to help you out [22:05] I think it will be a hard enough to setup the server hehe :D [22:05] sarnold: if the server can't handle the load, just use varnish [22:05] ilhami: its very easy [22:06] yeah we will use varnish I think [22:06] RoyK: I think he wants to stress test the machine a bit to make sure it's sane before relying on it [22:06] we would like it to be able to handle 50 requests per second? Is that fine? [22:07] has anyone here used byobu? [22:08] ilhami: depends upon what those requests do... [22:09] like buying some products and also entering website [22:11] buying products from the website that is [22:12] ah, okay, nothing too fancy, it feels like the kind of thing that should work fine [22:12] i hope so. [22:13] Can I write the ISO to my USB? [22:13] .iso file [22:13] in FAT format? [22:14] and I should go for the LTS version right? [22:14] ilhami: 50 requests per second would be easy for most static work. if it's active pages, use caching like varnish [22:14] ilhami: depends upon what you're trying to do; if you want the USB stick to install ubuntu, you should use dd to -overwrite- the filesystem on the USB stick with the iso [22:14] I am on Windows now sarnold. :) [22:15] ilhami: oh.. [22:15] I can't use Unetbooting now? [22:15] oops I mean for this [22:15] ilhami: maybe you can? I've not looked into it. a pal wrote this, it may also be able to help http://sourceforge.net/projects/win32diskimager/ [22:16] yeah I have that tool actually :D [22:16] We are yet to find the hardware we want the server to be installed [22:16] we are thinking about an old laptop or desktop. [22:17] is there a huge difference between the LTS and the new 13.04 version? [22:18] 13.04 isnt new, it is EOL. 14.04 will be released in April. [22:18] ilhami: 13.04 has already reached end-of-life; 13.10 release is still supported. The newer release is using apache 2.4 [22:19] oh yeah I meant 13.10 sorry [22:19] ilhami: sticking with 12.04 LTS would be a fine choice, it'll give you the opportunity to upgrade to 14.04 whenever you'd find it convenient, rather than being forced into upgrading soon [22:19] sarnold but I guess I can just update to 2.4 from the LTS version? [22:20] ilhami: if you choose to compile your own apache, yes, but that's an annoying hassle :) [22:20] ilhami: No. There are no version bumps within a release of Ubuntu. [22:20] sarnold: when upgrading do I have to do a fresh install or is it now possible to upgrade without breaking packages? [22:21] * hallyn tears out his hear and throw half of it at libdbus and half at pthreads [22:21] hallyn: oh? [22:21] * sarnold throws rocks at libdbus [22:21] * thumper gets popcorn [22:22] ilhami: we expect that you can upgrade from release to release with a minimum of fuss; I understand some apache 2.2 configuration options have changed and are no longer available in apache 2.4; this is of course a possibility with most programs, but in general it works quite well [22:23] ilhami: (I've upgraded machines through seven or eight ubuntu releases, and upgraded through a decade of debian releases before switching to ubuntu; the history of safe upgrades is strong :) [22:24] sarnold I am really excited about this. :) [22:24] if it will work as we expect. [22:24]