[00:24] Hey everyone. I was wondering. I setup a samba share on a fresh Ubuntu 16.04 server but when I try to connect from windows it is slapping a domain on my user, what should I be putting there? [00:25] the share shows up but I can authenticate [00:26] I'm really new to using samba so I added the share like this: http://paste.ubuntu.com/16072871/ [00:27] I can't authenticate. Sorry. [02:47] Is there anything I need to know that possibly the interwebs isn't telling me about changing an ip on eth0? I have an Ubuntu-Server running LAMP on my network. It was on a dhcp and I've released the ip through ifconfig and put in the new static ip address, netmask and broadcast as well as adding those arguments to /etc/network/interfaces and restarted as well as a reboot to test if it [02:47] stuck... does anyone have any input on this? [02:50] your first message was cut off at "to test if it" [02:50] sorry [02:50] to test if it stuck, does anyone have any input on this?...... [02:50] oh! hah [02:50] :P [02:50] I expected to see a problem of some sort in there.. [02:50] so, uh, did it work? :) [02:51] I'm currently waiting on connectivity from a router/demarc. [02:51] but hypothetically speaking, should i have anything to worry about besides those? [02:51] so long as you've modified /etc/network/interfaces, that should take care of just about anything [02:52] (granted, sometimes people may put a dhcp-assigned address into a configuration file.. but those people are living in a state of sin) [02:52] I've got mapped ip's in this situation. [02:53] what is a *mapped* ip? [02:54] 1-1 ip mapping; a public address to a private address. [02:55] heh? [02:55] that is called nat, not mapping [02:55] ok heh. [02:56] but that still has nothing at all to do with this [03:01] patdk-lap: thanks for disparaging that. [03:02] I was about to elaborate, sarnold, I'm going to be dealing with nat/pat as well. [03:03] zorbs: that's alright, that's quite common :) [03:04] * zorbs chuckles [03:08] ya, interfaces file doesn't care about any of that [03:08] that is all firewall stuff [03:08] I've got the rules for it set so it should be all right. [04:50] hey all. I replaced a hdd in a mdadm raid1 array, and forgot to fail the old disk first [04:51] of course upon boot, I get an error but I can skip it since the raid array is just for data, not /boot or anything essential [04:51] what can I do now to get mdadm to come back up and start rebuilding the array with the new disk? [11:37] hia I ssh tunnel my mysql connection to the host and thus use “localhost” [11:37] however, since yesterday, on one host it tries to connect to 127.0.0.1 instead [11:37] and won’t connect, obviously :p wtf [12:42] I'm about to upgrade a virtual Ubuntu server from 14.04 to 16.04. I had to run do-release-upgrade -d" to get 16.04. While upgrading it hangs while showing this output: http://paste.ubuntu.com/16078429/ [12:44] upgrades arent offered until the first pointrelease, unless you use -d [12:44] So... there's always a risk === dorrek_ is now known as dorrek [13:01] @lousdisk: I did the same - testing - and after 'do-release-upgrade -d' the network interfaces could not start anymore ... [13:02] I guess we should wait until June/July when 16.04.1 is released and release-upgrade is possible without the -d switch [13:02] doesnt -d get you to 16.10 now ? [13:02] (since thats now open) [13:02] if you where runnong 16.04, yes [13:03] if your not running 16.04, nope [13:03] I didn't see 16.10 in the meta release file last I looked. [13:03] as patdk-wk said: from 14.04 you go to 16.04 [13:04] why would I want to switch from LTS to non-LTS on a server? [13:04] ansd i always though we imaplementd /etc/update-manager/release-upgrades for that :) [13:04] except the fact that LTS is a lie for itself ... [13:04] (so you dont need the -d ) [13:05] thats not true [13:05] you will need the -d option if you want to release upgrade from 14.04 to 16.04 NOW [13:06] the file mentioned has PROMPT_lts per default [13:06] well, then i understood it wrong back then :) for me -d always meant the latest development release :) [13:06] well if you consider 16.04 the release version and 16.04.1 sth. else then yes [13:06] and to do updated to lts before .1 releses you edit the prompt to normal ... do you upgrade and set it back to lts [13:07] .1 is the date wheer the lts offers the upgrade, never before [13:07] true - but if you want to test anything before June/July ... [13:07] so yes, i consider them something different, especially since .1 will have a different default kernel [13:08] not major differneces, or does it? I saw 4.4 after a release-upgrade to "devel" [13:08] usually the point releases use the hwe stack [13:08] but yes sounds true [13:09] when you install a fresh 14.04.4 you get kernel 4.2 [13:09] right, if you install a 14.04 image you get 3.13 [13:09] yes, and the HWE kernel pushes you to 3.19 [13:10] the last point release has the last wily kernel stack [13:11] heh? why would 16.04.1 have a different kernel? [13:12] my 14.04 installs still have the same old 14.04.0 kernel [13:12] HWE kernels are nothing but problems [13:12] cause they are not tested against anything other than themselfs [13:12] so any kernel modules you install, fail to work with them [13:13] cause those modules don't also have compatable backports from where the HWE kernel came from [13:13] patdk-wk_, well, if you install ioff a 14.04.4 media you get a 4.2 kernel from the wily-lts hwe stack [13:13] * patdk-wk_ looks at mainly xtables and vmware tools [13:13] but not a wily xtables module [13:13] so the kernel fails and panics and the system is unusable [13:14] you get the very same kernel package wily has ... [13:14] till it's but back only the orig 14.04 kernel [13:14] ogra_, yes, and the kernel modules in 14.04 are NOT compatable with it, from packages outside the kernel package that was backported [13:14] ? [13:14] the kernel modules come from the kernel package [13:14] oh? [13:15] so nvidia is compled into the kernel package now? [13:15] xtables is compiled in? [13:15] vmware drivers are compiled in? [13:15] dkms <3 [13:15] yes, and dkms fails, cause those modules are not compatable with those kernel headers [13:15] patdk-wk_, no, they are dkms built indeed ... [13:16] but thast the reason why the .0 images are always provided alongsice the .n ones [13:16] yes, and why I refuse to use anything newer than .0 [13:16] and .2+ is just evil [13:16] atleast for server installs [13:17] patdk-wk_: FYI, the .1 keeps the same kernel base version as the .0 [13:17] hwe is very limited scope, it solves an issue, by breaking so many others [13:17] sdeziel, currently, not sure there is anywhere it says it will always be that way [13:17] it used to be all .x releases had the same, but that changed [13:18] though, I do wish a newer iso image would come out [13:18] since .1 releases 3 months after the .0, no new Ubuntu release is released in that short time [13:18] so I wouldn't have to constantly upgrade a crapload of packages after an install [13:19] yeah, I could use fresh iso with the same kernel base version. Would be handy [13:19] ATM, when we install new nodes we pick the latest point release then downgrade the kernel [13:20] oh ya, I was testing openssh :) [14:13] I use the HWE kernel 3.19 in ubuntu 12.04 on like 50 virtual linux-servers hosted on wmware/esxi - I never had any issues with that. [14:13] or was it 14.04 ? [14:15] ronator: probably 14.04 because 12.04 most recent HWE kernel is 3.13 [14:15] yes [14:16] everywhere, where I can install "linux-image-generic-lts-vivid" and "linux-headers-generic-lts-vivid", I do so. [14:17] the issue was with vmxnet3 drver and that was in 12.04 I believe [14:17] not sure about 14.04 [14:17] I know xtables issue is in 14.04 [14:21] we use 12.04 on some legacy machines and did not have any issues with vmxnet3 driver - I checked that with our vmware-guy. [14:21] what was the issue like you experienced? [14:22] what driver are you using? [14:23] driver is vmxnet3, type of adapter is e1000 - or what driver oyu mean? [14:23] * patdk-wk_ wonders how you run vmxnet3 on e1000 [14:23] well, vmxnet3 is a nic [14:23] you can use many drivers on it [14:23] the vmware-tools package driver, the open-vm-tools driver [14:23] and each of those have many different versions [14:24] I do not know I am not the vmware guy. ahh, no i do not use open-vm-tools - we use the "legacy tools" [14:24] I think newer kernels also include it [14:24] ya, I'm using the open-vm-tools [14:24] okay good to know so I won't try to use them ;-) [14:25] thesedays, I don't use it for the pvscsi/vmxnet3 stuff [14:25] but still need it for the balooning driver [14:25] and not sure, if I need it for timesync or not [14:26] ignorance is bliss ;-) [14:29] I'm reminded of "If ignorance is bliss, why aren't more people happy?" [14:30] cause it doesn't pay the bills [14:31] saw ubuntu phone, must buy ubuntu phone, slave at job to earn money to buy phone [14:40] For some reason Xenial 32bit server won't let me change the home directory of user with UID 1000. Some systemd process is continuously in use even when that user is not logged in [14:43] reminds ME of: “We buy things we don't need with money we don't have to impress people we don't like.” [14:57] genii: can you tell what systemd proc? === deathpip is now known as ivyharriet [14:59] genii: did you try to add another user, who should get uid 1001 and see if the problem persists? [15:26] ronator: Works with testuser and UID 1001. Does not work with normal first user with UID 1000 [15:27] homedir encrypted? [15:27] No encryption [15:27] i have no idea but at least we know it has sth. to do with the very first user account created in installation process :/ [15:27] I made a second admin user, logged in with that and tried to change the home dir of the first admin user with that login, still no dice [15:28] If I find the time I will also setup a fresh xenial 32bit. If I see the same problems, I will tell you here the next days ... [15:29] I can still do it by manually editing the path in /etc/passwd , but sudo usermod -d /someother/directory firstuser still doesn't work [15:29] * patdk-wk_ watches ronator create time [15:29] ps aux| grep firstuser shows some systemd process that keeps respawning if it's killed [15:30] "patdk-wk_ watches ronator create time" [15:31] can you gain more info from syslog? [15:31] good think I live in a 5d world, and can see time :) [15:32] 'some systemd proc' is still quite undifferenciated :/ if you can find out which proc it is (to what servcie it belongs) you may find better help on the internet [15:32] lsof? [15:33] genii: if you'd like systemd to stop respawning the process you should stop the process/service instead of killing it [15:34] on my systems, I cannot find anything, but I am using 64bit, and used minimal-install [15:35] Hi can anyone help me with a mysql issue in here? [15:35] no idea [15:35] That's that cleared up then :P [15:36] Harry-GAA001, well, you didn't describe your issue, so how can anyone in here know if they can help you? [15:36] !ask [15:36] Please don't ask to ask a question, simply ask the question (all on ONE line and in the channel, so that others can read and follow it easily). If anyone knows the answer they will most likely reply. :-) See also !patience [15:37] ok will do, so my server host 123reg stuffed up all their servers, rebuilt them and now I get this issue: Warning: mysqli::mysqli(): (HY000/2002): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (111) [15:38] The hosts are not responding to support tickets or calls so I'm trying to fix the issue myself [15:38] is mysql running? [15:39] Yep my sql is running. I've also checked the mysqld.sock file is located in the right place and that the path in my.cnf matches it [15:39] Grrr can't replicate it now [15:39] what is the ownership and permissions of the /var/run/mysqld and /var/run/mysqld/mysqld.sock [15:39] I have also stopped and started the server and run updates to make sure it's running properly [15:39] ..which I guess is a good thing [15:40] Will just check that, 2 secs [15:41] srwxrwxrwx 1 mysql mysql 0 2016-04-27 16:40 mysqld.sock that is what comes up as the owner [15:41] The permissions are set as 755. Is that what you mean? [15:42] well, that is the second part [15:42] what is the directory set to? [15:42] though that is not 755 [15:43] var/run/mysqld/mysqld.sock [15:45] Do you mean read/write/execute permissions on the mysqld.sock file? [15:45] no [15:45] file!=directory [15:46] That's for the directory: drwxr-xr-x 2 mysql root 80 2016-04-27 16:45 mysqld [15:46] ok [15:47] anything in dmesg that looks related? like aparmor lines? [15:47] Ok you've lost me there with my limited knowledge. Is dmesg a file? [15:47] it's a program [15:49] I'm using Putty, is that the same sort of thing [15:56] patdk-wk, I'm not familiar with dmesg or aparmor. Would these already be installed, or something I'd have to download? [15:57] Harry-GAA001: might worth checking the mysql log and error files [15:57] I've checked the mysql log and the error is the same as the one I've posted from the website [15:58] hello i am using ubuntu 14.04 i have 2 domain names that are configured i need to have separate project in my server for every domain how can i do that ? [15:58] Guest18219: what do you mean by 'separate project'? [15:59] Guest18219: for websites, you can easily define different server names for each site configuration to work for - you then get to specify different directory paths for each domain to have as a 'document root' and therefore serve different sites [16:00] teward, i configured from name cheap website to point to my ip [16:01] teward, could you explain me in details a little please ? [16:02] teward, is just related to apache configuration ? [16:09] Can anyone else help me, desperately need to fix Warning: mysqli::mysqli(): (HY000/2002): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' [16:10] I have checked the path which looks correct; I've checked it matches in my.cnf; I've restarted and reinstalled mysql and still no change [16:22] Didn't realise the chat had timed out. So my question was, how can I fix this error: Warning: mysqli::mysqli(): (HY000/2002): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' [16:23] I have already checked mysql is running, checked msqld.sock is in the right location and matches with my.cnf. I've restart mysql a bunch of times and even reinstalled it and still the same error [16:26] anyone? [16:28] Didn't realise the chat had timed out. So my question was, how can I fix this error: Warning: mysqli::mysqli(): (HY000/2002): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' [16:28] Harry-GAA001: can you pastebin the mysql error log? [16:28] Does mysqld.sock belong to the mysql user? [16:29] Soz, I'm new to IRC and didn't realise it had disconnected me. Do you want me to paste the whole error? Some bits I don't think are relevant but I can if you think it useful [16:29] Harry-GAA001: pastebin as in https://wiki.ubuntu.com/Pastebin [16:30] I will see if I can do that. So you want the contents of mysql.log yes? [16:30] and/or the .err === ksx4system_ is now known as ksx4system [16:32] ok bear with me a sec and I'll sort that :) [16:37] Sorry, my guy who is connected to the server is awol from his keyboard. I don't have access to it at the moment. I'll get the pastebin asap. === rcj` is now known as rcj === rcj is now known as Guest12905 [17:19] Guest18219: My apologies I got called away [17:19] Guest18219: Yes, it's simply the requirement of making additional Apache site configurations - one for each differing site domain name. [17:19] teward, no problem [17:19] Guest18219: I can't help from the Apache side - I'm an nginx guy unfortunately. [17:19] but yes, it's just a case of updating the Apache site configurations to have VirtualHosts for each site [17:19] with different configurations for each site [17:19] (thereby different docroots) === Guest12905 is now known as rcj [17:20] teward, and no need for configuration in name cheap site that i have configured my domain names? [17:21] Guest18219: the only configuration you need to do from NameCheap's side is DNS changes to point the domains and any relevant subdomains you create to your server via A / CNAME records. [17:21] (I recommend A records) [17:21] Guest18219: the rest of the configuration is done at your server - it has to know what sites it serves with individual configuration records in the Apache configs for how to deal with each site [17:38] arges: smb: I trust noone will shed a tear if i try my hand at merging libvirt from debian later this week? [17:39] hallyn: have at it. i'd be happy to review [17:39] hallyn, not me [17:40] ok, thx [18:05] Hi, does the ubuntu-server cd install network-manager by default? Or it's just using ifupdown? [18:06] the latter ... might switch to systemd-networkd at some point [18:06] Thank you ogra_ :) [18:06] :) === ReScO_ is now known as ReScO