=== Lcawte is now known as Lcawte|Away [00:35] Hrmm. I'm a bit surprised there's no easy way to tell what files are missing chunks due to bad sectors from a ddrescue .img and .log. Guess I was hoping to be lazy! [00:36] keithzg: heh, that seem slike the whole point of having a .log file :) [00:36] keithzg: check for strings of 512 or more zero bytes perhaps? [00:40] sarnold: I didn't actually run ddrescue with any fill setting, so as far as I understand it just skips bad sectors, then. [00:41] The suggested way seems to be to md5sum every file and then run again with filling in missing data and then md5sum *that*, which seems . . . inelegant. [00:43] I just naively assumed these details would be in the log file, which was indeed naive of me since ddrescue itself uses the log file as a reference (for repeated passes on failed sectors). [00:55] The main problem is the literally millions of files in question here, so . . . this is going to take a while. Alas! [00:57] and if you have md5sum executions waiting on dead blocks in files that might take even longer === markthomas is now known as markthomas|away === G__ is now known as g === g is now known as G [01:30] Well, I'm running it on the image file created via ddrescue, so that *shouldn't* be a problem, unless I'm being braindead and missing something. [01:30] Actually yeah, when I think about it more I don't know why I'm so certain of that, heh. [01:31] But so far it's going through files quite quickly, there's just a huge ton of files. [01:38] How good is ubuntu server? Running on old Xeons, in a VM, with only 600MB of memory, it took me 2 years to finally need more memory. Started hitting swap last week, upped to 2000MB now. [01:40] it's my 'primary' linux machine, too. [01:40] hehe, I bet it reboots in a few seconds, too, thanks to caching in the host :) === CiPi is now known as cipi === wmp is now known as Guest5451 === Guest5451 is now known as wiuempe [07:20] gosh peta.. http://motherboard.vice.com/read/peta-sues-for-copyright-on-behalf-of-selfie-taking-monkey :-/ === athairus is now known as athairuzzz [08:45] hi, some1 got an idea how to configure squid3 to cache by filename? I want to build a deb/rpm proxy and ignore the hostname === RedHacK is now known as ASA === ASA is now known as RedHacK === Lcawte|Away is now known as Lcawte [12:03] Can HDD failure cause Ubuntu server to crash at startup? [12:07] Yes. [12:13] T3DDY: a disk failure can do most annoying this, so yes. better boot on something live and run smartctl -a /dev/whateveritsname [12:14] Cool, thank you [12:26] +1 [12:28] hello. I intstalled ubuntu-server on vbox. It is working on windows host (Connection by putty). But on linux host it is not working. I can not connect to server via ssh. It says that Network is unreachable. I set the network adapters with same options. Can u help me? [12:42] i'm installing libreoffice on a cli server, it's failing out on the install of gnome-integration. I can skip that yes, I shouldn't need anything from gnome [12:42] KervyN: I think apt-cacher-ng does that. it keeps a list of mirrors and considers paths to them to be the same. [12:47] hello. I intstalled ubuntu-server on vbox. It is working on windows host (Connection by putty). But on linux host it is not working. I can not connect to server via ssh. It says that Network is unreachable. I set the network adapters with same options. Can u help me? [12:48] i have eth0 and eth1 visible. eth1 ipv4 is 192.168.56.101 [12:49] and i can not log in there. i allowed ssh in ufw and i am out of ideas [12:49] If I mount to a folder that has files in it, can I access the files or are they unreachable until I umount the folder? [12:51] nevermind, I found the answer [12:53] kgirthofer: libreoffice doesn't care if its going on a headless box, it's a GUI application and has the associated dependencies. [12:53] the install script removes a few of the gui dependencies, I think I can remove them [12:54] I can't use the install script [13:33] Guys I'm using LDAP to have my user Database. How can I sync the system users also ? [13:33] My concern it's for the root/administrator user to have a fail-save mechanism to login on the server. [13:35] ParsectiX: my experience in this area is very dated, but afaik, there are no good solutions for this. [13:38] jrwren: thanks for the answer [13:40] I just want somehow to have my administrator password locally on each server and also I can change/sync it centrally so every machine can get the change. [13:52] no good solution to this? [13:53] I wouldn't recommend doing that at all [13:53] just sync your ssh keyfile [14:15] coreycb, ok to work through rc1? [14:16] jamespage, yep I can start today [14:52] how can i now what requires systes restart when i get msg on login "1 package can be updated. [14:52] 0 updates are security updates. [14:52] *** System restart required *** [14:52] sorry, should only put that system restart required [14:53] what requires my system to be restarted ? where can i find out it ? [14:53] 0 updates are security updates. [14:53] *** System restart required *** [14:53] damn [14:53] sry, accident paste [14:54] Usually kernel updates. [14:54] damn laptops, should disable the touchpad [14:54] lordievader: but if it isnt kernel ? [14:54] im gonna check ubuntu security msgs [14:54] There are a couple of other things that require reboots. Libc for example. [14:55] but is it "required" if just normal use and nothing special than ssh running on server ? [14:56] udev too ... upstart/systemd would ask for a reboot as well i think [14:56] Meh, required is a big word. Required to reboot if you want to run the new versions of the software, but if you don't want to reboot nothing is going to stop you. [14:56] i understand that if kernel update and for example remote root exploit in kernel but just a little update that doestn do anything special should not be a cause to boot a server [14:56] I am running a PPTPd server on a Ubuntu 12.04.5 LTS blade server to serve more than 1000 pptp clients. As the number of clients grow, the ubuntu machine keeps crashing. Any idea? [14:57] why remote exploit in kernel? [14:57] local exploit is horrible [14:57] if you run any kind of remote services, such as a webserver [14:57] i run apache and ssh [14:57] I have also set up a watchdog timer but it doesn't seem to help. When the server crashes, only a hard reboot will do. [14:57] apache2 with php5 and ssh [14:57] lordievader, well, indeed, but i think all of them switch that notification on [14:58] weather you listen to it is indeed up to you :) [14:58] Guest92591, so any exploit in the kernel, local, remote, or otherwise, your vaunerable to [14:58] unless you can guarrentee there is no exploit paths in your apache/php programs at all [14:58] and I have never known that to ever be the case [14:59] patdk-wk: there is nothing that would be for me in ubuntu security notices for ubuntu-server 14.04 LTS and my software-config === Guest92591 is now known as krsna [15:00] you only run php files from ubuntu? without modifications? === krsna is now known as Guest31451 [15:00] you have never installed a php file outside of apt-get [15:00] you have never been vaunerable to a 0day vaunerability === Guest31451 is now known as neurotus [15:14] Is there a limit on the number of pptp clients that I can accept? [15:14] on a ubuntu server [15:14] Tazmania, assuming nothing shows up in the logs before it crashes, I'd set up a netconsole to another machine, as the final gasps probably aren't being saved [15:14] (if things _do_ show up in the logs, you should pastebin your logs :p) [15:15] Tazmania: I think the max clients is limited by either a config option or your system's memory being consumed [15:15] a 'crash' is also not very useful, because it could mean it segfaulted, or your server rna out of memory, etc. [15:15] patdk-wk: but there's nothing in ubuntu security that would cause me a problem after last kernel reboot and apt-get upgrade [15:15] (neither of which should cause a hard lockup) [15:15] cwillu_at_work: except if all the swap is also consumed and something's not clearing it out [15:15] patdk-wk: update and kernel change and reboot [15:16] teward, in which case the oom killer gets involved, and the oom killer definitely will scream into the logs [15:16] mhm [15:16] (and if you've disabled it, well, stop hitting yourself :p) [15:16] patdk-wk: no security notices for apache or php [15:16] cwillu_at_work: oom-killer once hard-froze a Hardy system when it started killing things, so ehh [15:16] [citation needed]; lots of things get confused for "hard froze" [15:17] what is oom-killer? [15:17] Tazmania, probably not relevant, unless you have a log file that says otherwise :p [15:17] Out of memory killer. [15:17] ok [15:17] thanks [15:17] (have you looked at /var/log/kern.log after a crash?) [15:18] To prevent OOM you usually have a bit of swap. [15:18] erm [15:18] neurotus, you seem to be confused [15:18] you usually have a bit of swap so that you don't have to keep idle memory in physical ram [15:18] what does a security notice have to do with apache or php? [15:19] MemTotal is 16GB and SwapTotal is also 16GB [15:19] they can both be security free, and working as intented, and still allow people to run whatever they wish [15:19] Tazmania, as I said, probably not relevant unless you've seen something in a log that suggests this is the problem [15:19] I mean, you told it to run a php file [15:19] if that php file allows you to, well, exploit the kernel [15:19] well, your rooted [15:19] I am just checking the server detail and will look into the log [15:19] and if not, your just now part of a botnet [15:20] patdk-wk: but no kernel change after that or anything else im running [15:20] patdk-wk: so im confused why reboot required [15:20] patdk-wk: what log should i read ? [15:21] patdk-wk: or install some security logging app [15:25] I get a lot of input/output errors from PTY failed [15:25] patdk-wk: i read /var/log/apt/history.log and theres nothing security related after last reboot [15:26] patdk-wk: i have only installed rtorrent after last boot [15:26] so it should not require reboot [15:33] if you do an uptime -s and then pastebin that history.log I think you may find out that is not true [15:35] AvatarA: ive done apt-get autoremove which removed old kernel from bootlist, should that be the cause ? [15:35] AvatarA: a cloud-server im running [15:36] AvatarA: kvm virtualization === markthomas|away is now known as markthomas [17:05] smb: arges: thanks guys. [17:05] hallyn, yw === neurotus is now known as krsna [18:03] hi guys im trying to connect to my wireless at home. i have some issues. it works with windows. what can i do? === athairuzzz is now known as athairus [18:37] arcsky: need more details.. [18:39] arcsky: is the network broadcasting SSID ? [18:42] arcsky: going away. hope someone helps u [18:47] Okay I'm having a really frustrating issue getting SSL to work with dovecot :/ [18:59] krsna: yes === markthomas is now known as markthomas|away === pekkos is now known as krsna === krsna is now known as Guest3289 === Guest3289 is now known as neurotus [21:01] Hello folks! [21:03] Could someone help me figure how to unblock a virtual server runing ubuntu-server please? [21:03] hi [21:04] hi [21:04] my friend has recently upgraded his server to ubuntu trusty [21:04] well, what is a virtual server [21:04] but now he's getting errors when trying to upgrade [21:04] and how exactly did you block it [21:05] orfeo: What do you mean by "unblock"? [21:05] http://pastebin.com/Mk8jq8g3 these are the errors I'm getting [21:06] how can I resolve this? I'm pretty sure insserv is conflicting with upstart [21:07] well, I did install a bunch of applications on it and now when I try to stop it running to reinstall it it won't allow me. I figure that it is because there is a process running which is somehow blocking the machine. [21:08] "when I try to stop it" means what, and "it won't allow me" means what? [21:09] For example, when I try to shutdown the server via control panel I receive a message saying "An Error Occured! A process has already been instantiated for this virtual server. Please wait for the process to finish." [21:09] What control panel? [21:09] the VPS control panel [21:10] it is a virtual server. Not local. [21:10] What happens why you try to shut it down from within the machine itself? [21:10] do you mean via ssh? [21:11] what command should I use for that? [21:11] Could anybody help me try to resolve my errors with insserv? [21:11] SSH or virtual console - either way. [21:11] shutdown.... [21:11] Hold on, let me try it. [21:13] tonyyarusso: shutdown: time expected [21:13] Yup [21:13] Look at the man page for the syntax. [21:15] http://pastebin.com/Mk8jq8g3 can anybody please help me with this? [21:19] the worst happened. I did shutdown now, lot access and the VPS control panel still blocked saying the server is runnig. [21:19] how weird is that? [21:20] What arguments did you use? [21:20] shutdown now [21:21] So you didn't halt it. [21:21] is it a problem? [21:23] Well, that's why your server is still running. [21:23] Should I use a shutdown -q now ? === markthomas|away is now known as markthomas [21:26] -h is what you wanted [21:26] I don't think there is a -q... [21:27] * tarpman generally invokes poweroff(8) rather than try to remember the correct shutdown(8) incantation... [21:28] I thought I saw it. Well, never mind, now its too late. But thank you anyway. [21:29] I use shutdown -h now to halt, shutdown -r now to reboot [21:30] genii: cool, good to know. For the next time maybe. [21:31] Serus: do you need ppp stuff? I think first thing I'd try is apt-get purge ppp pppconfig pptpd ; if you need mdadm for raid on the system to work, leave that alone for now... === Lcawte is now known as Lcawte|Away [22:06] sarnold: some guys in #ubuntu helped me, thanks anyway! [22:06] Serus: oh cool, what'd they do? [22:09] sarnold: not a lot! Turns out Serus' server release upgrade crashes part way through and it was hard-rebooted [22:10] TJ-: ah, thanks :D [22:11] tarpman, I do that too [22:11] but I learned on some systems, the poweroff command means just that [22:11] But I dug into the issue with insserv; it *seems* there's a bug in the LSB headers for the /etc/init.d/mdadm script declaring it requires mdadm-raid, but no such service file exists [22:12] direct poweroff, WITHOUT safely unmounting or syncing the filesystems [22:12] patdk-lap: poweroff without shutting down first, you mean? which systems do that? [22:12] solaris [22:12] ah k [22:12] * tarpman mental note [22:13] gone oldschool on those, init 0 [22:13] tarpman: beware of the 'reboot' command there too [22:13] na, reboot is sane [22:13] oh? was it sunos that just insta-rebooted then? [22:13] ya, same deal, but htat is safe [22:13] .. if you wanted sync and umount, you would typed those too :) [22:13] no, that is different [22:14] reboot would do a normal safe reboot [22:14] but it would not do a bios reboot by default, but just a kernel reload, if all the drivers supported it [22:14] otherwise you needed to do a reboot -p, to force a full bios reboot