[06:01] Good morning === zyga_ is now known as zyga-mbp === tds5 is now known as tds [11:33] hello, after reboot, /run/php is often missing and then, php fpm services fail to start [11:33] then I have to do mkdir /run/php first [11:33] why is this directory not create automatically? [11:37] you should provide your ubuntu and php version so others can help you [11:39] ubuntu 20.04 [11:40] php5.6, 7.3 and 7.4 is installed [11:40] php-fpm [11:42] why do you have multiple php versions installed ? [11:43] because there are two websites, that require older php versions [11:43] Um...thats a mess AND security issue [11:43] you know 5.6 is WAY out of date [11:47] yes, I know, but I can't change it now. [11:51] Aison: are all 3 PHP versions affected the same? [11:53] yes [11:53] Aison: I cannot reproduce with php7.4-fpm from the official repo [11:54] all of them create their pid and socket file in /run/php [11:54] hmm [11:54] right but the dir /run/php/ get created by php7.4-fpm when it starts [11:55] Aison: actually, it is probably create due to /usr/lib/tmpfiles.d/php7.4-fpm.conf [11:56] is it created by php7.4-fpm.service? [11:57] Aison: no, the php7.4-fpm package creates the tmpfiles.d snippet from above and then systemd takes care of creating the dir [11:58] Aison: if you need to workaround, tmpfiles.d snippets can be created/overridden by the admin by putting them in /etc/tmpfiles.d/ [11:58] is there a systemd tmpfile service? [11:58] Aison: https://www.freedesktop.org/software/systemd/man/tmpfiles.d.html [11:58] maybe something is wrong there [11:59] Aison: which PPA are you using to get PHP 5.6 and 7.3? [12:00] https://launchpad.net/~ondrej/+archive/ubuntu/php [12:04] Ussat, I know it is a security problem and it is a bad solution. To gain at least a little bit of security, this server runs in a virtual machine with no access to something else [12:07] sdeziel, as workaround, I create a config file in /etc/tmpfiles.d [12:07] at least this is working [12:10] but I can remove 7.3 now ;) [12:17] Aison: OK, Ondrej's PPA is at least reputable (for a PPA) [12:18] it is the only ppa I use [12:19] (ok I use some for testing, but not on this server) [12:19] Aison: that said, since you have the PPA on a Ubuntu 20.04, the php7.4 stuff from the PPA probably "shadows" what's officially coming from Ubuntu's repo [12:25] yes that may be the problem [12:58] hi, i just booted ubuntu-20.04.1-live-server-amd64.iso on a dl360 gen9 and after selecting the language in the text installer I was dropped to a shell (ubuntu-server@ubuntu-server:~$) [12:58] how can I restart the installer? [12:59] oh … some modifier key was enabled, I was just switching ttys apparently … my bad [12:59] haha [12:59] alt-left/right, or alt-function keys, or ctrl+alt-function keys [12:59] that's what you get with these shitty html5 remote consoles :< [12:59] stuff happens and you don't really notice [13:01] fwiw: I can just press left+right without modifiers to switch ttys :< [13:02] the html console sent an Alt down but no Alt up? fun! [13:03] you can do the installation over SSH, AFAIU [13:03] the installer welcome screen tells you how, I think [13:12] the installation over SSH works well (after you've typed the very long randomly generated password without typos) ;) [13:13] maybe the html5 console allows copy/paste [13:13] argh … it offered to update the installer to a newer version [13:13] I feel stupid for following that offer [13:13] oh? [13:13] it crashes with a python exception in subiquity now [13:13] there are bugs fixed in the newer installer [13:13] and, looks like, bugs introduced :( [13:14] what's the exception? [13:14] https://ubuntuforums.org/showthread.php?t=2448414 [13:14] this one [13:17] I kinda don't want to reboot, since I'm remote and the installer wanted to verify integrity during startup, which took >1h [13:17] you can skip that, I believe [13:17] I have network going fwiw [13:17] https://bugs.launchpad.net/subiquity is the bug tracker for subiquity [13:18] https://bugs.launchpad.net/subiquity/+bug/1892763 [13:18] Launchpad bug 1892763 in subiquity "Installer crashed in the middle of the installation and the installation restart attempts are blocked" [Undecided,New] [13:18] it's similar, and as you see the devs couldn't reproduce it and didn't get the requested logs [13:19] it's been a while since I had a launchpad account, well. here I go [13:22] ERROR root:39 finish: subiquity/Refresh/configure_snapd/switching: FAIL: must be str, not NoneType [13:22] so yeah, I'll be posting logs then [13:29] so … how would I go about downgrading the installer? [13:32] `snap info subiquity` only offers 20.09.1, which is broken for me [13:32] with snap refresh snap-name --channel=something, I think? I don't use snaps much [13:33] same [13:38] and the installer stops crashing after apt fullupgrade in the live system huh [13:39] that... how... but... [13:39] yeah, I don't even [14:03] the installer ultimately crashed again [14:03] restarting Hi guys I have KVM installed on ubuntu server 18.04.I am using Virt manager to manage the KVM. These are my results for ip (i have removed static ip with **1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00inet 127.0.0.1/8 scope host lovalid_lft [14:44] forever preferred_lft foreverinet6 ::1/128 scope hostvalid_lft forever preferred_lft forever2: enp38s0: mtu 1500 qdisc mq state UP group default qlen 1000link/ether 00:d8:61:a6:30:49 brd ff:ff:ff:ff:ff:ffinet 110.145.42.250/32 scope global enp38s0valid_lft forever preferred_lft foreverinet .../32 scope [14:44] global enp38s0valid_lft forever preferred_lft foreverinet 192.168.1.219/24 brd 192.168.1.255 scope global dynamic enp38s0valid_lft 53468sec preferred_lft 53468secinet6 fe80::2d8:61ff:fea6:3049/64 scope linkvalid_lft forever preferred_lft forever3: virbr0: mtu 1500 qdisc noqueue state UP group default qlen [14:44] 1000link/ether 52:54:00:51:2d:28 brd ff:ff:ff:ff:ff:ffinet 192.168.1.219/24 brd 192.168.1.255 scope global virbr0valid_lft forever preferred_lft foreverinet6 fe80::5054:ff:fe51:2d28/64 scope linkvalid_lft forever preferred_lft forever4: virbr0-nic: mtu 1500 qdisc fq_codel master virbr0 state DOWN group default [14:44] qlen 1000link/ether 52:54:00:51:2d:28 brd ff:ff:ff:ff:ff:ff8: vnet0: mtu 1500 qdisc fq_codel master virbr0 state UNKNOWN group default qlen 1000link/ether fe:54:00:3b:03:3b brd ff:ff:ff:ff:ff:ffinet6 fe80::fc54:ff:fe3b:33b/64 scope linkvalid_lft forever preferred_lft foreverThis is what is contained in sudo nano [14:44] /etc/network/interfacesauto lo virbr0iface enp38s0 inet manualiface virbr0 inet dhcpbridge_ports enp38s0I can not seem to get the internet to work on the virtual machines i create.Any help would be appreciated. [14:44] !paste | WobblyBob [14:44] WobblyBob: For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [14:45] https://paste.ubuntu.com/p/WF453Z3VCr/ [14:45] !paste [14:45] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [14:48] Hi guys sorry for repeat posting - I need help with a ubuntu server 18.04 kvm internet issue please see https://paste.ubuntu.com/p/WF453Z3VCr/ [14:51] WobblyBob: your paste is hard to read but it seems like enp38s0 is not "enslaved" by virbr0 [14:51] also, you have an IP on enp38s0 but that should be on virbr0 itself [14:52] im running plesk on this server also so have to be careful i dont ammend it incorrectly. [14:52] is there any other commands i can run to give you a better insight [14:54] WobblyBob: I think I've deciphered it, thanks ;) [14:54] WobblyBob: and provided what I believe would fix your issue [14:55] ok brilliant i appreciate it, how do i enslave enp38s0 to vibr0 do i do this using the document /etc/network/interfaces? if so what changes should i make im pretty new to linux and i am still learning [14:57] WobblyBob: first question is probably to ask if you have direct access to the host (and not depend on the network to get to it) [14:57] I personally dont have direct access to the host as in physical sense. I do have ssh root access from home. [15:00] WobblyBob: that's not super cool then because what you'd need to do risks breaking the connectivity [15:01] WobblyBob: is that a new bridge? As in it never worked/existed before, right? [15:02] yes its a new bridge its currently working fine for plesk and all other services. I broke it once before applying netplan so im aware of the potential breakage '=D [15:02] seeking some alternative help before potentially breaking it again ha [15:02] never existed or worked before that is correct [15:03] WobblyBob: is plex running on the host or that new (?) VM? [15:03] no buddy plesk only and all services included with that. Oh an I ntopng runs on it also [15:04] they run directly on the server without any vm. the vms will run the employees windows machines. [15:04] k [15:05] oh and vncserver [15:07] so I have a project that still depends on gcc 4.8 (ancient), is it still possible to find a build for that for 20.04? [15:07] last I can find is 16.04 [15:08] WobblyBob: I think that all you need is to change "auto lo virbr0" to "auto lo enp38s0 virbr0" in /etc/network/interfaces and then reboot. But I'm not 100% sure on this one [15:09] sdeziel: ok ill give it a whirl now :) [15:10] WobblyBob: that will have enp38s0 use virbr0 as "master" and will also ensure the DHCP assigned IP is on the right NIC [15:12] !paste [15:12] sdeziel: https://paste.ubuntu.com/p/yGjq77C7PW/ [15:12] hows that looking? [15:16] WobblyBob: yeah, looks good [15:22] sdeziel: i think its broken lol [15:23] WobblyBob: sigh [15:24] I just realized that the IP 192.168.1.219 was configured on both enp38s0 and virbr0, which is weird [15:24] '=D ill see if i can get into a local machine on same network and undo the changes [15:25] WobblyBob: was virbr0 managed by libvirt? [15:26] i have no idea buddy how could i find out? [15:28] WobblyBob: "virsh net-dumpxml default" I think [15:30] seems I can still use xenial repositories [15:32] jiffe: you'd probably better served with a Xenial lxd container or VM [15:33] jiffe: looks like it's also available in Bionic (https://packages.ubuntu.com/bionic/gcc-4.8) BTW [15:34] sdeziel: ill save that command for later once i can get into the network again. thanks for your help so far [15:34] WobblyBob: sorry it didn't work :/ [15:35] sdeziel: its ok ill get it back to how it was and try again later [15:37] WobblyBob: good. Keep in mind that when a NIC is hooked to a bridge, the IP needs to be moved from the NIC to the bridge [15:37] ok, ive posted this issue else where so hopefully later on this evening ill have a few answers to try [15:39] sdeziel: im going to relax till later once i can get back into network. thanks again speak soon [19:53] trying to get ntopng to send alerts via ssmtp mta...anyone else have to do something similar? [19:53] * mgedmin prefers postfix [19:54] ya, but ssmtp is like 4 lines to configure. Ive got it to work with everything else.. [19:55] It works on the box...i just cant figure out how ntopng wants me to configure put into there gui tool to use it [20:00] grendal-prime: for a daemon-less MTA, msmtp-mta never failed me [20:00] normally programs that send mail just use the /usr/sbin/sendmail interface, and programs usually supply an executable there that works [20:00] and ssmtp is unfortunately abandonware at this time [20:00] I also use msmtp-mta, works fine [20:01] it's no Real Server, but works for my laptop :) === tds9 is now known as tds [20:02] sarnold: nowadays, msmtp-mta ships with an optional daemon listening on 127.0.0.1:25 for the rare programs who wants to inject using SMTP [20:02] sensible default of to not running this daemon [20:03] what I want more is a queue for outgoing mail if I happen to send some while not online [20:03] anywway AFAIU grendal-prime's problem is not sending the mail, but in getting ntopng to talk to the local MTA [20:03] I don't even know what ntopng is [20:05] I always read it as something to convert 'n' to 'png'. whatever 'n' is.. [20:06] could be an n(etwork?)-top-new-generation [20:07] yes, very network [20:07] when and how does it send mail? [20:08] the package dependencies do not mention mail-transfer-agent === vlm_ is now known as vlm [20:51] !paste [20:51] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [20:53] yes ntopng is network monitoring tool. You can run it on a linux software router. It does very good job of logging werid stuff..and really small footpring [20:53] print that is [20:54] Can anyone help me setup my networking on ubuntu server 18.04 so the internet will work in VM instances my current set is here https://paste.ubuntu.com/p/YcjMchm4ZX/ [20:57] id yank out that werid netplan, install ifudown to start with [20:58] that will get you using the /etc/network/interfaces file like we have been doing for the last 20 years [21:02] what hypervizor you using? [21:05] KVM QEMU [21:05] i will install ifudown soon, will this affect my current setup on server as have plesk and other bits running on it [21:06] right on...i have 10 installations like that [21:06] grendal-prime: sorry new to this chat and forgot to tag your username [21:06] grendal-prime: ok two mins :) [21:07] umm so long as you statically assign the ip address that currently have [21:07] i would do that before you yank out netplan [21:08] grendal-prime: how do i do this? sorry im fairly new to the linux scene [21:08] id install ifupdown first, then.....wait there is a walkthrough i use on this [21:10] https://itectec.com/ubuntu/ubuntu-ubuntu-18-04-switch-back-to-etc-network-interfaces/ [21:10] there ya go [21:14] grendal-prime: theres a seperate file for plesk in netplan folder ill install ifupdown now and go from there [21:15] grendal-prime: i do have vnc installed so as a newbie is there another way to bridge using gui to save you time? [21:16] grendal-prime: seems i already have ifupdown installed [21:17] was this an upgrade or a new install? [21:17] if you have virtmantager installed there is an entire gui section just for creating bridges [21:17] grendal-prime: i may have edit the interfaces file back to original state to get my self back online after breaking it earlier today [21:21] do you have a linux workstation you can install virtmanager on? [21:21] grendal-prime: i have virt manager installed already on this server is that any good [21:22] and do you have ssh-server installed on the server [21:22] yes i am currently using them [21:22] ok so you have a desktop installed on the server? [21:22] grendal-prime: yes i do for remote access vnc4server [21:23] ok...well later you will probably want to get yourself a laptop with linux workstation on it...it really makes this sort of stuff much easyer [21:23] anyway [21:23] grendal-prime: im on linux now using ubuntu desktop? [21:24] if you double click on the host itself you should get a ui that has the network interfaces tab on it [21:24] im talking about withen virtmanager [21:26] http://resume.booksnmore.com/videos.php?selected_video=virt-example.webm [21:26] that might be a bit basic for where your at but its kinda helpfull if your not familar with virtmanager [21:28] its an intro i use to bring interns up to speed [21:31] could probably make up a quick vid on how to create a bridge device [21:31] it would be on a vm. [21:32] but would get the right idea [21:33] ok yes please that would be great as the tutorials im looking at dont seem to look anything like what im looking at [21:33] does the one i posted look close? [21:34] grendal-prime: very similar yes, im just curious do i need to setup the vm fully before being able to see the networking options properly within virt manager [21:34] no the bridging happens at the virtual host [21:34] hmmm ill screenshot the software im looking at now [21:35] !paste [21:35] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [21:35] basically, if you bridge a physical device, your guests then are bridged to the network the host is connected to...so for instance if you had a dhcp server running on your physical network, your vm-guest would be able to get an ip from that dhcp server. [21:36] https://imgur.com/a/18S7E5F [21:37] please advise once seen so i can delete post as has ip on it lol [21:37] ya your good [21:37] you can kill it..and i would worry about the ip [21:37] ill leave you enough cash to get back home to mom and dad [21:37] hahahaha [21:38] haha [21:39] just so i know...you are running the libvirt server on the same box there ..so you have a "localhost" as your sever identifier right? [21:40] because with ssh you can connect to any kvm host on the network and view it through that interface...its pretty fkn amazing piece of software [21:40] and you can even set up failovers with another image on the other servers. [21:40] anyway [21:42] grendal-prime: sounds complex ive yet to learn most of this stuff. im not too sure what you mean so basically i use this for plesk and hosting websites and emails. [21:43] am i on the right path to getting it bridge though? [21:43] well once you get your network straitend out the rest is pretty easy [21:44] have to move to office brb [21:53] sorry [21:53] i have strep throat.. [21:53] in quarnteen...but like not...you know like covid...just strep..its confusing and the meds dont help [21:53] ugh [21:54] ANYBODY WANT A STEP KISS!!!? [21:54] STREP THAT IS! [21:55] grendal-prime: think ive bridged it installing vm now to see if its working [21:56] its nice if you have more than one nic, then the guests can share one and the host use the other [21:57] ? [22:00] I second interface is usually used to connect to another network. different subnet, some physical network seperate from your primary that you need vms to have access to ither as well or exclusively. Like a lab switch or something like that [22:00] or like an industrial lan [22:00] scada systems and whatnot that are typically...well they are riddled with legacy ethernet devices designed by plumbers and electritians. [22:01] grendal-prime: was that to me? also the bridge worked thank you for your help [22:01] RIGHT ON BOB!! === tds8 is now known as tds [22:02] grendal-prime: quick one, my keyboard wont work in the vnc instance into the vm [22:02] any ideas [22:02] check your vnc client make sure you have the right keyboard selected [22:05] now all you need is a vpn concentrator and you will be off to the races....fixn stuff back at home with your phone, spying on your kids while you are on vacation....oh wait...well in two years when there are no more bugs on everything... [22:07] We will have to do the concentrator on another day though. [22:07] grendal-prime: please expand further on the vpn concentrator [22:08] my meds are kicking in and...im all wobbly and the only thing i can eat is jellow and soft frut [22:09] vpn concenrator. Basically a server that allows youto connect from your phone...laptop...from anywhere in the world. Puts you right on the same network..you push a route to your remote device. Encrypeted tunnel that ends at your home. You can work from anywhere on your virtual cluster, turn on your equipment, watch movies off your plex server....from mcdonalds. All secure. [22:09] no limit to the number of clients you can have. [22:10] grendal-prime: i see what you did there about the meds ha, where do you originate from grendal prime [22:10] and you can use it to connect to other devices that are connected to your concentrator. Like if your parrents need help with their computers, you set up a client on their machine and you can then vnc into there machine from anywhere and help out. [22:11] an egg. [22:11] i mean more than likely [22:12] grendal-prime: ha, and sounds good - do you use comcast as your supplier the 1gig line? [22:12] dont have gig here. [22:12] im in the sticks in a fairaday cage hiding from "the man" [22:13] i do have a pretty good connection though [22:13] grendal-prime: ha your funny, thank you for helping me anyway [22:13] is the video above yours? [22:14] ya, my resume is on there...although its about 6 years old now [22:14] do you teach people this stuff? [22:16] use to ...alot, now I do network engineering for water authoritys, scada support mostly. Make sure everything keeps talking. I get to climb towers and ride motorcycles into places where there is no other way in to maintain equipment. [22:20] sounds very interesting - sounds like a cool job tbh [22:43] grendal-prime: going to sleep buddy ill reconnect tomorrow if your online about the vpn concentrator [22:49] oh ya...i had a problem to! [22:50] so ssmtp you guys are all like "no way" msmtp thats the game now? [22:50] ssmtp...i mean it still works with gmail and everything and it sets itself up to be the mta..if i do a sendmail it utilizes it to send. [22:52] in the little box though it asks for an smtp server...and i just figured id put sendmail in there but that does not work [22:52] smpt.localhost nogo smtp.127.0.0.1 also notgood [22:53] and the gmail entry it suggests it wont even save to the config. [23:10] grendal-prime: with msmtp-mta you get all that (relay through Gmail, sendmail command, etc) [23:12] grendal-prime: for better control, you probably want to edit /etc/msmtprc directly [23:13] so what would i put then in as a smtp server on the application? [23:13] grendal-prime: none, use the sendmail command, like you did before with ssmtp [23:13] be careful with that system-wide configuration file: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=944188 [23:13] Debian bug 944188 in msmtp "/etc/msmtprc password disclosure" [Normal,Open] [23:14] see thats the thing...if i could feed it sendmail that should work with ssmtp..other apps are using just fine. just not ntop...hmm [23:15] sarnold: it's an unfortunate debugging (mis)feature of msmtp. I'm curious how you workaround it? [23:15] grendal-prime: so you need a smtp listener (TCP/25)? [23:16] sdeziel: in my case, I'm on a single-user system, and use ~/.msmtprc instead of the system-wide configuration, so it's not something I'm too worried about [23:16] grendal-prime: if yes: `sudo systemctl enable --now msmtpd` [23:16] sarnold: right but then what about cron email/spam ;) [23:17] grendal-prime: after that, you should have a SMTP listener on 127.0.0.1:25 (or whatever you tuned it to in /etc/default/msmtpd) [23:17] see thats the thing, cron has no problem sending with ssmtp. [23:18] sdeziel: I'd rather not get cron mail :) heh [23:18] grendal-prime: yeah, cron will just execute /usr/sbin/sendmail directly [23:18] grendal-prime: my cron spams me using msmtp-mta, I can tell you that ;)