[00:20] VBoxManage guestcontrol $guest stat /bin/sh # hangs for me using 5.2.18_Ubuntur123745 on host and guest additions 5.2.18_KernelUbuntu r123745 [00:28] VBoxManage guestcontrol $guest stat /bin/sh # does not hang when using 5.2.22 direct from Oracle. [00:34] anyone about? [00:34] I added new nvidia drivers to 18.04 and reset. after resetting the pc, I tried to log in... [00:34] The screen blinks and then goes back to login. [00:42] is automatic upgrade of packages turned ON in this configuration: [00:43] https://pastebin.com/0SEgvsBj [00:46] upgrade =/= update [00:47] 20auto-upgrades and 50unattended-upgrades are responcible for that [00:47] k, I am wondering if packages like, apache2/mariadb/redis will get upgraded [00:48] cognition___: no, it's downloading the lastest list of packages and even downloads the packages, but it stops short of installing them it looks. seems reasonable if somebody wants to keep control _when_ it happens but automate it partially [00:50] OerHeks: theres no 50unattended-upgrades [00:51] hmm maybe part of livepatch, not sure [00:52] https://pastebin.com/Jt2XieCK [00:52] mutante: TY [00:56] OerHeks: It is zenial 16.04.3 LTS [00:56] xenial [01:00] so yes, "1" means enabled [01:13] `curl -v fake.url` is routing back to our network router. any ideas how to fix? this happens for any unknown hosts. so when I would normally get "server not found" errors in Chrome/Firefox, I just get our local web server trying to serve the request. [01:16] CoreyW: dont make your resolver respond (incorrectly) with the routers' ip address instead of NX, or switch to a resolver which works correctly. [01:16] tomreyn: i dont know what any of that means, aha. is that something i do on my local machine or at router level? [01:17] i assume my machine bc the windows machines dont have this issue [01:19] CoreyW: it's not a default configuration, so it is something the administrator of your computer, your network or your ISP setup. [01:20] tomreyn: no one has touched the computer except myself, and i just did a normal installation of ubuntu [01:21] `systemd-resolve --status` shows on last line `DNS Domain: ourdomain.com`, so that must be the culprit [01:21] not sure how to fix though [01:22] How can I get git to track symbolic links? It seems to be hit and miss. Thanks. [01:23] CoreyW: "ourdomain.com" sounds like you may have copied and pasted instructions from a tutorial, and missed to replace this by your actuual domain name. [01:23] tomreyn: that was just an example. it actually is our domain there [01:23] sorry for confusion [01:24] resolv.conf also has `search ourdomain.com` [01:25] CoreyW: this search domain is probably sent by your router upon dhcp and inherited by your ubuntu system. [01:25] tomreyn: ah okay. so is it okay to override that on my machine? or should i modify the router setup? [01:26] not sure if dhcp will overwrite my changes or not [01:26] CoreyW: in other words: if you don't like those lookups you should either reconfigure the router not to send it to this or all of your systems or reconfugre your ubuntu system to ignore this information received as part of the dhcp response. [01:27] is reconfiguring my ubuntu simple enough? i dont want to mess with router if possible. [01:28] if it is plain wrong and you manage the network / router then you should 'mess with the router'. [01:28] CoreyW: That sounds like a common ISP move where they'll intercept DNS requests for invalid domains and direct them to an ISP landing page, either on the router, or elsewhere. [01:29] okay i will have my boss contact our ISP and see if they can get it sorted out [01:29] iosecure: sounds very likely [01:29] iosecure: in this case though its just sending it back to us, which is weird [01:32] thanks tomreyn [01:32] CoreyW: you can have the isp change it, thats good, but if this is a company environment you should also configure these values on the router yourself properly, and (beforehand) understand what they do. if you dont want to do it on the router but on the clients, it's a matter of configuring 'supersede domain-name ""' and 'supersede domain-search ""' in /etc/dhcp/dhclient.conf [01:33] okay. im not a network guy, but i will look into that. appreciate the information! [01:36] if you dont have a network guy / gal, then either have $boss hire one, teach it yourself (have your $boss have someone teach you) or find an isp which doesn't provide you with such intentionally-misconfigured-for-profit-maximization gear / network configurations. [01:36] tomreyn: dang, my boss just said we cant configure our router because its managed by our ISP [01:37] lol [01:37] how to calc the disk the ubuntu got when we do apt autoremove? https://bpaste.net/show/9ae148c73d0e afaik we only know only disk space needed [01:40] CoreyW: If it were me, I'd set up a DHCP and DNS server internally, and bypass the router for those functions. [01:40] But that's me. [01:40] CoreyW: so that's a small office / home office setup, not unusual. have the isp fix it and tell them you dont approve their default configurations which cause mistyped domain names and non found domain names to be sent over the (Internet) wire, in what is a non default configuration. [01:40] CoreyW: i'm the same as iosecure there [01:42] fishcooker: it will say once you actually do the autoremove === daniel is now known as Guest29493 === capella|away is now known as capella [01:47] iosecure: after talking more, we do have DHCP server setup here. [01:47] CoreyW: Is it on that router? [01:47] apt-get dist-upgrade --dry-run #might tell? [01:47] brandonkal: try asking in #git if you haven't. this matches my observations: https://stackoverflow.com/questions/954560/how-does-git-handle-symbolic-links [01:48] tomreyn: there is line `option domain-name "ourdomain.com";` that i just commented out to see if it works. restarting dhcp service, but not sure if that resets active leases or not [01:48] iosecure: on that router? all i know is our router is setup to go to our local server, which has dhcp server running [01:48] CoreyW: it won't, since this is initiated by the clients. [01:49] ok. is there way to reset my machine? [01:49] CoreyW: actually, it will reset the leases, most likely, but the clients will have to request new ones to catch up on the new configuration [01:49] new ones /renewal [01:49] which they will do after half the lease time [01:50] Thanks tomreyn [01:50] sorry, not sure what that means [01:50] boss mentioned i could just take a static IP and then it prob won't be issue. but still would be nice to fix while im here [01:52] CoreyW: either wait or reboot the router (temporarily disconnecting all users and their connections) or restart the dhcp clients on all computers, or a combination of these. [01:54] ok thanks [01:55] CoreyW: if you use network-manager (default on ubuntu desktops), you can "systemctl restart network-manager", this shoould also restart the dhcp client or get you a new lease anyways. [01:55] or just use the gui to disconnect from, then reconnect to the network. [01:57] tomreyn, that helps. I also found: `ln -s new-file ./gitignored/file` works with git but `ln -s new-file gitignored/file` only works on the system but not in git [01:58] that's interesting, news to me. [02:01] tomreyn: i did restart network-manager after `systemctl restart dhcpd` and `system restart network` on the server for good measure. didn't seem to help commenting out the "option domain-name" line. [02:01] maybe its something else in our config [02:02] Changing that domain name will not affect the function of the router in intercepting DNS queries. [02:02] what should i change then? [02:02] The router's configuration, or stand up your own forwarding resolver that refers to an external DNS service that doesn't do that crap. [02:03] As long as you're going to that router or your ISP for DNS, this will continue to occur. [02:03] iosecure: so youre saying it is something above our dhcp server and actually a router configuration then? [02:03] because i see DNS information on our DHCP server settings, it has a couple of our IPs and then 8.8.8.8 and 8.8.4.4 [02:04] Are any of those "couple of our IPs" the router itself? [02:04] And handing an external DNS server, like 8.8.8.8/8.8.4.4, directly to clients is a bad idea. [02:04] Clients should -always- use internal DNS, or -always- use external DNS. [02:04] "our IPs" may be wrong actually, they look similar to ours, but they are probably just our ISPs ips. but no, one of them is not the router [02:05] so the DHCP server you run is separate from the (potential, and likely existing) DHCP server on the router? [02:05] at least as far as i know its not the same ip [02:05] Mixing the two will almost always lead to inconsistent behavior. [02:05] tomreyn: possibly. [02:05] best i do just consult with our ISP then. [02:05] CoreyW: which single system works as a DHCP server for your network is what you need out first. if it's two, kill one. [02:06] * nneed to find out [02:06] you almost never want more than one [02:07] tomreyn: i can view the active leases for our local DHCP server, and it shows my machine there. if that means anything [02:10] CoreyW: you said you (company) run your own DHCP server. whoever stated this should know where it runs. alternatively you could analyze DHCP reponses sent on the network (as a result of DHCP requests), or, simpler, switch off the router for a while (cutting everyone off the network) and see if you can still get a dhcp lease (i.e. an IP address assigned via DHCP) at the time. if that happens, you have a separate (and probably two, one in addition [02:10] to the router, DHCP servers which is a very bad configuration for a small company network). [02:12] cognition___: 16.04.3 is much outdated, 16.04.5 is current. [02:13] (so apparently updates ar eindeed not applied automatically, and even then you'd still need to reboot for kernel patches to be applied) [02:15] Jerbot: you should probably diuscuss what you mean by "I added new nvidia drivers", i.e. how you added those. [02:15] tomreyn I already reformatted. [02:15] Is the Deja Dup app automatically backing up data to a remote server by default? [02:16] tomreyn What I'm trying to do is verify that my os uses the dedicated gpu when necessary. [02:16] Hi! I'm trying to install Ubuntu and Elementary side by side on different partitions of the same SSD, and I would like to have a different UEFI entry for each of the systems, however both get named "ubuntu" on the UEFI bootloader. I've tried several ways to rename them but no luck. Does anybody know the right way to rename the UEFI label of an Ubuntu based distro? Thanks!! [02:18] tomreyn installing the driver was difficult. I added the nvidia graphics repository with some ppa command then updated then did apt install nvidia-driver-default and was shown the error E: dependencies required but not met. The nvidia-415 driver was the dependency, and it also had a dependency. So, traced the dependencies down installed them all and finally nvidia-driver-default worked.... [02:18] I think I figured it out, looks like the answer is No [02:18] Then I restarted my computer and got the "login loop" issue. Typing the correct password sends you back to the login screen. [02:18] * Jerbot breathes [02:18] Jerbot: i have no current first hand experience but i think the proprietary nvidia drivers provide some kind of gui to configure which applications (based on executable name) to run through the dedicated gpu. one of them could be glxinfo. [02:19] tomreyn I'm going to try that. Ijust don't understand why nvidia's driver-managing gui doesn't allow me to force dedicated. [02:19] tomreyn: i just talked to ISP, they said they have no DHCP servers running or anything special. [02:20] so it must be our server. [02:20] erts: efibootmgr probably [02:20] thats all the help they could offer [02:20] Jerbot: i'm really the wrong person to give advice regarding nvidia, i don't use it - sorry for responding at all. i hope someone else can help. [02:21] tomreyn Chatting about stuff helps, man. I'd rather talk to someone unknowing than nobody at all. Thanks for contributing. [02:21] epicmetal: I tried using that to re-create the entry with a different name, but it doesn't work..still looks for an ubuntu directory [02:22] looking over our dhcpd.conf, only place i see "ourdomain.com" is on that one line. so idk where else it would be coming from [02:23] it has "option routers 192.168.2.1" which is our router. but i assume that is normal [02:23] which apparently our server *is* our router though, according to said boss [02:24] CoreyW: a Linux server can also be a router by flipping one switch basically [02:24] so that's possible [02:25] makes sense then. so something on our server/router is causing our ubuntu machines connected to it to pass unresolved requests back to itself [02:27] resolution is set to Hosts file > DNS > Local hostname. might that be the issue? [02:27] oh shoot, nvm. i see Search domains now. [02:28] CoreyW: is there "search" in /etc/resolv.conf ? [02:29] what is the issue again? i did not follow this from the start [02:29] yes there is [02:29] issue was our ubuntu machines, going to `fake.url` in browser/curl results in request made to our server/router [02:29] instead of 'server not found' error [02:30] CoreyW: the second part to that is /etc/nsswitch.conf it decides where names are looked up, can be "files" /etc/hosts/ or "dns" or both [02:30] resolve.conf has a `search ourdomain.com` line, and it says `Generated by NetworkManager` at top [02:31] CoreyW: when you say "request", do you mean the DNS lookup ? [02:31] i mean it first has to try before it knows that it is "not found" [02:32] and probably you have your DNS server set as your router [02:32] for caching [02:32] yes sounds right. i just mean though if i make an http request to a nonexistant domain, it ends up back at our router IP [02:33] or do you mean sometihng else, like a captive portal [02:33] where you get an actual HTML page that is on the router [02:33] yeah [02:33] CoreyW: if you want to check whether you have multiple DHCP servers on your LAN and using nmap is an option and you have it installed already, you can run this and see if you receive more than one response (i.e. if it says something other than "Response 1 of 1" you have a rogue DHCP server and need to fix this first): sudo nmap --script broadcast-dhcp-discover -e eth0 [02:33] oh yea, i think this brings us back to "open a ticket with the ISP" [02:34] yea, i wanted to say that too. see here http://ask.xmodulo.com/monitor-dhcp-traffic-command-line-linux.html [02:34] how to check for DHCP traffic , where it really comes from [02:34] mutante: i just called them and they made it sound like they had no control over anything like that [02:34] CoreyW: ask them how _you_ can get access to it then. everything needs some admin, heh [02:35] since the boss claimed they manage it [02:35] CoreyW: oops, had a typo there, here's the right commands: MAINIF=$(ip -o route list default | cut -d' ' -f5); echo "Main interface: $MAINIF"; sudo nmap --script broadcast-dhcp-discover -e $MAINIF [02:35] mutante: he said they manage the router, but they said all the router was doing is passing to our server. idk. this stuff is over my head mostly. [02:35] tomreyn: ok i'll try that.... after editing resolv.conf, it is overwritten by networkmanager and adds the search domain back in [02:36] CoreyW: ok, try to run the "sniffing" tool (tcpdump) to find out if they are saying the truth or not [02:36] how is that happening? [02:36] One should ask permission before running tools like nmap or tcpdump [02:36] What could be adding things to the path of a non login bash shell except ~/.bashrc and /etc/bash.bashrc? === DrWatson_ is now known as DrWatson [02:37] CoreyW: not to insult you, maybe you should question why you are tasked with network management stuff if your boss seem to know well about it and you don't. [02:37] CoreyW: i dont think it's the search domain, unless you get "fake.url.yourdomain.com" in the browser [02:37] i'm trying to install ubuntu server 18.04.1 and i'm on the screen that shows which interface i want to use as the default. if i switch to another tty how can i get the interfaces/ip addresses? ifconfig doesn't work [02:38] mutante: okay. no the url remains fake.url in browser. [02:38] forgotmynick: "ip" replaced "ifconfig" [02:38] tomreyn: youre right, hes just not concerned as i am [02:38] forgotmynick: try "ip a s" [02:39] tomreyn: nmap shows "Nmap done: 0 IP addresses (O hosts up) [02:39] thank you [02:39] it showed on IP offered [02:39] what tcpdump command do i run? [02:40] CoreyW: it was in the link above [02:40] in method 1 [02:41] CoreyW: I would ask permission prior to running tools like nmap or tcpdump [02:41] CoreyW: is there no output about "broadcast-dhcp-discover" from nmap? [02:42] tomreyn: yes it showed a few lines. what line has meaning? [02:42] cryptodan_mobile: you havbe a point, but this very command i provided only does what a dhcp client does, send a broadcast request and see whats returned. [02:42] erts: maybe just copy the EFI app to the default path [02:43] erts: EFI\BOOT\BOOTX64.EFI [02:43] tomreyn: https://pastebin.com/5JsW8wQA [02:43] CoreyW: quopting myself: if it says something other than "Response 1 of 1" you have a rogue DHCP server and need to fix this first [02:43] erts: where "EFI\" is a directory in the root of your ESP [02:44] CoreyW: oh you run an older nmap than myself, it doesn't say it in this version, but you seem to only have this one dhcp server there. [02:44] tomreyn: okay cool [02:45] so if search domain isnt the issue, i wonder what would be [02:46] CoreyW: so no rogue dhcp servers. line 12 lists the resolvers (Domain Name Servers) returened by your DHCP server. i don't think the google ones (#3 and #4) would return those fake responses., so it's probably either #1 or #2 [02:46] tomreyn: those look like our ISPs IP addresses. but they claim to not be doing it. [02:47] CoreyW: you maybe appending your domain to URLs maybe [02:47] so basically its not even trying googles dns youre saying? [02:47] cryptodan_mobile: huh? [02:48] will the tcpdump be useful or was that just to check for double dhcp server? [02:51] CoreyW: https://askubuntu.com/a/473932 [02:52] cryptodan_mobile: on no, i dont have that line in there [02:52] CoreyW: your computers will use the resolvers provided in the order they are provided, so usually the first of them and only failover to the next if the first one fails to respond entirely. [02:53] tomreyn: i see. so you think that the first DNS server can't find the URL, and responds by sending back our IP? [02:56] CoreyW: i think it's likely that the first DNS server can't find the host name you asked for ("fake.url"), then your client asks it about fake.url.$YOURDOMAIN instead and the first resolver either correctly or incorrectly responds with your routers WAN ip address. === mnemonic is now known as semeion [02:56] tomreyn: hmm.. lookin in nsswitch.conf i see `hosts: files dns myhostname` [02:56] literally `myhostname` [02:58] and the hosts file has several entries for the local and public IP of our router/server [02:58] CoreyW: you want to configure your DHCP server to send the dns domain-name [02:58] CoreyW: you want to configure your DHCP server NOT to send the dns domain-name [02:58] ^ this [03:00] or maybe you want it to send a diferent domain name, one which you manage for internal DNS lookups only. but then you wouldn't know how and explaining this now takes more time than i'm willing to spend. [03:00] tomreyn: haha i know, and youve been super patient. [03:00] so the easier option is to not have it send the domain name [03:00] where do i set that option? [03:00] on the dhcp server [03:01] right but how? [03:01] the *much* less manageable option is to se all clients to supersede the domain-name option by the empty string [03:01] is there a web browser anitivirus [03:01] for website protection [03:02] CoreyW: which dhcp server do you run there? [03:03] pi0: https://en.wikipedia.org/wiki/NoScript [03:03] CoreyW: do you want to fix this just for yourself or for the entire organization? [03:03] ISC DHCPd version 4.2.5 [03:04] entire org would be great [03:04] CoreyW: do folks at work usually enter things like "www" in their web browsers and expect to end up on your website? [03:05] tomreyn: just www by itself? no [03:05] or "ftp" to end up on your ftp server or similar [03:05] no [03:05] thank you [03:05] installing it now [03:06] how do i tweak my laptop monitor to set the settings higher [03:06] not enough to squint lol but just to have the apps navigatable [03:07] CoreyW: ok, then check your dhcp server's configuration files (main configuration file and zone files) for "option domain-name" [03:07] tomreyn: yes, its set to "emrl.com" [03:08] CoreyW: so comment it out or set it to the empty string, then reload the dhcp server. then clear your local dns cache and try fake.url again [03:09] tomreyn: i did that earlier with no luck. can try again though to make sure all is reset [03:09] pi0: for me it's System -> Preferences -> Hardware -> Displays but it depends what desktop environment you are using [03:10] tomreyn: restart dhcp server with `systemctl restart dhcpd` should do it yes? [03:11] CoreyW: be sure to know how to revert this configuration change, since this can have side effects on how your collecgues need to / expect to be able to work. as indicated above ("short names to access your website, ftp server etc. weill no longer work, people and systems need to always use the fully qualified domain name to access resources hosted on your WAN IP address). [03:11] CoreyW: yes, this should do it if this is a system using systemd [03:11] thats fine. we dont do anyting like that hear... and by local DNS cache, do you mean at my local ubuntu machine level? or browser level? DHCP server level? [03:12] do i need to renew lease? [03:12] on your client you need to: systemd-resolve --flush-caches [03:12] clent = your local ubuntu machine [03:12] ...that's if it uses systemd-resolved [03:13] its standard 18.04 ubuntu so i assume [03:13] yes it does then [03:13] hey it worked this time! last time I just hadnt cleared the dns cache [03:14] ok, problem solved. it's indeed not your ISPs fault, just your org's [03:14] perfect! i don't know who set that before or why but i'm glad its gone now [03:18] Is there a super simple GUI analog to 'notepad' in ubuntu? [03:19] something that pops open in 1 second, just to scrawl a note real quick? [03:19] tomreyn: thanks very much for sticking with me. appreciate it. [03:19] and others as well! [03:20] Geo: try "pluma" [03:20] Geo: There is namo for quickies . [03:20] I have trouble using the RPM command on Ubuntu 18.04. Is the command only suitable for Redhat? [03:20] Geo: also "atom" is cool but that's much more powerful [03:21] mutante: did you seriously just suggest atom as a simple "notepad" clone? [03:21] sam_wong: yes, rpm is for redhat only, the r is for redhat [03:22] sam_wong: well and anything based on redhat, like centos, fedora, mandrake.. [03:22] Bashing-om: is that namo or nano? [03:22] mutante: ok, thanks [03:22] CoreyW: for the general approach in an organization environment: it's common to have the company LAN's DHCP server return your company LAN's resolvers as DNS servers to DHCP clients, as well as a domain-name or domain-search option value of "emrl.corp" (or similar), which can't be resolved on the internet's and ISP's resolvers. then you have your own LAN-only DNS server at the location the DHCP server points your clients to, which manages the [03:22] "emrl.corp" zone, and allows you to point anyone on your office LAN requesting resources such as "mailserver" to "mailserver.emrl.corp", and that to your actual LAN mailserver. [03:22] sam_wong: apt is the command you're looking for on ubuntu [03:22] sam_wong: there is a package called "alien" to convert rpm to deb but first try other solutions [03:22] Geo: Sorry ,, typo . nano . [03:22] mutante: ok [03:23] Geo: leafpad is by far the simplest text editor/notepad looking clone [03:23] Bashing-om: no worries, wasn't sure if that was some new fork or something :P [03:23] tomreyn: i see. still, never seeing a "server not found" message for anything is weird isn't it? [03:24] ok, thanks all. I'll take a look at thiese [03:24] sam_wong: apt-get install is the oldschool way on Debian-based distributions. Ubuntu is one of them. The more modern one is apt or aptitude [03:24] CoreyW: it should happen now, does it not? [03:24] nano is not GUI Bashing-om [03:24] Geo: :) .. yeah never can tell .. thinks progress so quickly. [03:24] sam_wong: or if you have a local file, dpkg -i package.deb [03:24] tomreyn: it does. but you were saying some orgs can type "mailserver" as shortcut, but if "mailserver" does not exist, shouldn't they still see a "server not found" page? [03:24] or would they expect to see their orgs page? [03:25] Kumool: I do stand corrected . [03:26] mutante: I have to download another file with a "deb" ending, right? [03:27] sam_wong: the best optioon is you use the "apt" commands to do both for you, download AND install [03:27] sam_wong: only if that is not possible, go to option b to download a .deb manually [03:27] CoreyW: depends. some companies have internal-only resources, just as a local file sharing server which is to be only available form your lan. and then you don'T want to make everyone type "files.mydomain.com" in their web browsers (or whatever client application) every time they want to access it, but just "files". that's what the 'dns-search' and 'dns-domain' option are usually used for. [03:28] tomreyn: right i get that. but with that setup, what if someone typed "filse" (wrong spelling), what would happen? [03:28] mutante: the first option seems easier [03:29] sam_wong: what are you looking for specifically? you can search with apt-cache search or look on https://packages.ubuntu.com/ then all you need to know is the package name itself, not an URL [03:30] Bashing-om: I think your full stops are not stopping well . [03:31] mutante: I am looking for Arduino IDE [03:31] CoreyW: first, clients will make a dns lookup against "filse.", which won't exist (resolver returns "NX domain"), then against "filse.emrl.corp." which your local DNS server will then repond to with what you configured it to respond to, if anything (if nothing is configued it will also respond with "NX domain" and the client software will return a DNS error) [03:32] sam_wong: "arduino - AVR development board IDE and built-in libraries [03:32] sam_wong: so that means try "apt-install arduino" and that should do it all [03:32] okay so i could have left the domain setting and configured something else to respond better basically? [03:32] im fine with the empty domain setting, but just for sake of understanding [03:33] sam_wong: sorry, either "apt-get install arduino" or "apt install arduino" , my bad [03:33] i gotta take off now anyways though. thanks again. [03:33] mutante: I am trying "sudo apt-get install arduino" and it seems to work fine [03:33] sam_wong: cool, looks good [03:33] CoreyW: if you would like to drive your company's network to this next level you really should have your boss send you to training though. [03:34] mutante: the program has been successfully installed. thanks a lot [03:35] sam_wong: you're welcome. i recommend you to always use apt-get instead of downloading manual software unless there really is none [03:35] sam_wong: that way it will stay more stable and should have no issues with dependencies [03:35] CoreyW: the way things were was not a good configuration, it effrectively leaked your employees DNS requests for non existing hostnames to the ISP, which is not needed and should not happen. as a company, you should always run a local DNS server for all requests from your employees (something you do not do, yet). [03:36] mutante: yes, you are very much helpful [03:36] tomreyn: yeah that would be cool. [03:36] or at least hire someone who knows what theyre doing more so [05:04] anyone here run a multihead display and use the nvidia driver? i'm having some issues apparently that xorg doesn't like... [05:05] noted OerHeks + tomreyn [05:29] having issue when installing ubuntu server on hardware raid Intel® C246 controller. When it reaches the partition/disk stage it show 2 disk instead of hardware raid 1 then if i continue [05:30] ubuntu 18.10 server [05:42] jmazaredo, i have no idea, but I'd assume it was your intelc246 setup (in error), whenever I've used hardware-raid the OS only sees single-disk (without play to get past raid for smartctl diags purposes etc) [05:47] I tried using other distro like centos7 only shows 1 disk [05:48] actually i removed the centos7 to install ubuntu [05:48] i will test the ubuntu desktop to see if same issue occurs [06:00] curtain command install preparing for installation configuring storage running 'curtain block-meta custom' curtain command block-meta removing previous storage devices [06:10] jmazaredo, you could ask on #ubuntu-server ; I recall there are multiple images of server, some/one I believe have less hardware-modules (drivers) on it but I don't recall details.. [06:17] ok im trying now the old iso anyway thanks! [06:23] guiverc_d, i tried 16.04 and works just fine :D [07:13] Hey. Is there any point trying to upgrade lte release [07:14] BrickedInTheHead: wich is your current release? [07:16] 16.04 [07:16] !ltsupgrade | BrickedInTheHead [07:16] BrickedInTheHead: Upgrade from 16.04 to 18.04.1 is now available, if you do not receive the upgrade window try update-manager -c [07:16] BrickedInTheHead: you can choose yourself what to do, stay on 16.04 or move to 18.04 [07:20] BrickedInTheHead: is that what you were hoping to hear? [07:26] Yes. Next trying for bluetooth audio [07:28] !details | BrickedInTheHead if you need more help [07:28] BrickedInTheHead if you need more help: Please elaborate; your question or issue may not seem clear or detailed enough for people to help you. Please give more detailed information; for example, we might need errors, steps, relevant configuration files, Ubuntu version, and hardware information. Use a !pastebin to avoid flooding the channel. === SP6NKW is now known as nikow [08:28] Dieser Kanal wurde in einen privaten Kanal umgewandelt. [08:31] !de | Schnabeltierchen [08:31] Schnabeltierchen: In den meisten Ubuntu-Kanälen wird nur Englisch gesprochen. Für deutschsprachige Hilfe besuche bitte #ubuntu-de, #kubuntu-de, #edubuntu-de oder #ubuntu-at. Einfach "/join #ubuntu-de" eingeben. Danke für Dein Verständnis! [08:32] sorry lotus, it was an automated message from my irc-relay... === capella is now known as capella|away [08:45] hello [08:46] So I finally reinstalled 18.04 using a new image [08:46] everything looks good except the mp3s [08:46] stil can't get it to work [08:46] sentiment: whats wrong [08:46] I have installed restricted extras and gstreamer stuff too [08:47] neither banshee nor rhythmbox play mp3 [08:47] both raise an error [08:47] I am freakin at my wits end with this [08:47] sentiment: did you try vlc or smplayer? [08:47] it's a new installation and still doesn't work [08:47] lotuspsychje: vlc worked in the previous installation [08:48] sentiment: wich error do you get? [08:48] this is the exact same problem that happened in the prev installation [08:48] vlc works, others don [08:48] t [08:48] GStreamer-CRITICAL **: 12:18:44.455: gst_object_unref: assertion 'object != NULL' failed [08:49] I reinstalled 18.04 into the root directory [08:49] but did't touch the /home directory [08:49] so I suppose I did the right thing [08:49] yet, this kind of crap makes me doubt that [08:50] GStreamer-CRITICAL **: 12:18:44.111: gst_bin_add: assertion 'GST_IS_ELEMENT (element)' failed [08:50] sentiment: if you reinstall ontop of your /home you also keep all the previous configs [08:50] I know, but what kind of 'personal' config can mess up such a basic functionality? [08:51] is there no separation of concern in this OS? [08:51] I formatted the root partition, and that is what should matter as far playing mp3 is concerned, because it's a system function [08:52] what do you think I can do to make it work besides booting into Windows as I have done in the last two weeks? [08:53] sentiment: to test this, you could try booting a liveusb and see if you can play mp3 by default to compare with your current install [08:54] sentiment: is your system up to date also? [08:55] lotuspsychje: how can I install coreelements? [08:55] it seems to be related [08:55] https://forums.freebsd.org/threads/gstreamer-broken.12158/ [08:56] it's almost fuckin 2019 and this distro still got issues with mp3 [08:56] And no, /home should not be related to such things at all... for gods sake [08:57] I'm sorry but I have been struggling with this for almost 2 weeks now [08:57] sentiment: please stop complaining, you cant generalize such statements, you can only ask for your own system [08:57] MP3 is proprietary I thought? [08:57] tried various crap since the last botched upgrade and wasted freakin time over that [08:57] until it was suggested here that I should reinstall the whole thing [08:57] !language | sentiment [08:58] sentiment: Please avoid any language that may be considered offensive, including acronyms and obfuscation of such - also see https://wiki.ubuntu.com/IRC/Guidelines || The main channels are English only, for other languages, see https://wiki.ubuntu.com/IRC/ChannelList [08:58] so I did that and most of issues went away except this one so far [08:59] sentiment: uname -a please? [09:00] TheMesquito: As far as reality is concerned, ogg is more 'properietary' [09:00] Linux Ubuntu-Desktop 4.15.0-29-generic #31-Ubuntu SMP Tue Jul 17 15:39:52 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux [09:00] sentiment: so...you come here and complain with an outdated system.. [09:01] what? mp3 needs updated software? are you serious? [09:01] Windows 98 played it out of the box for gods sake [09:01] sentiment: in the future please only seek for help with a fully updated system [09:01] even previous versions of Ubuntu [09:02] sentiment: if you feel to discuss it more join #ubuntu-discuss ok [09:02] Why don't you just pruge gstreamer to remove any messed up cfg files and reinstall it? [09:03] I've had success with doing that when various libs break [09:03] messed up config files? I reinstalled, formatted / [09:15] cpaelzer, Hi ! Is there a chance to see a QEMU 2.11.2 update for 18.04 LTS at some point ? [09:18] so I fixed it finally [09:18] rm -r ~/.cache/gstreamer-1.0 [09:18] you were right [09:18] !info qemu bionic [09:18] qemu (source: qemu): fast processor emulator. In component universe, is optional. Version 1:2.11+dfsg-1ubuntu7.9 (bionic), package size 204 kB, installed size 687 kB [09:18] but I doubt that I was wrong [09:19] groug: if you like a newer version test, try a higher ubuntu version or look into !backports or snaps [09:19] what was the discussion channel? [09:20] groug: we dont reccomend mixing package versions on ubuntu [09:20] !discuss | sentiment [09:20] sentiment: Want to talk about Ubuntu, but don't have a support question? /join #ubuntu-discuss for non-support Ubuntu discussion, or try #ubuntu-offtopic for general chat. Thanks! [09:26] lotuspsychje, it isn't a newer version actually, it is the latest stable release of QEMU 2.11.x... It doesn't bring new features, just bug fixes. I'd like to know if there's a chance to provide those fixes to LTS users [09:28] !latest | groug [09:28] groug: Packages in Ubuntu may not be the latest. Ubuntu aims for stability, so "latest" may not be a good idea. Post-release updates are only considered if they are fixes for security vulnerabilities, high impact bug fixes, or unintrusive bug fixes with substantial benefit. See also !backports, !sru, and !ppa. [09:28] groug: when security flaws are found in package versions, updates will handle those [09:28] !sru | groug [09:28] groug: Stable Release Update information is at https://wiki.ubuntu.com/StableReleaseUpdates [09:30] groug: for example, ubuntu 18.10 uses Version 1:2.12+dfsg-3ubuntu8.2 [09:31] lotuspsychje, I'm talking about LTS... not about interim releases === davor_ is now known as davor [09:42] lotuspsychje, the fixes I'm interested in affect Live Migration of VMs between 16.04.x and 18.04.x, which is an important feature for our (IBM) customers. I'll follow the SRU procedure and see how it goes. Thanks. [09:50] groug: see also snap find qemu [09:59] Guys! [10:00] Anyone know how to route traffic from ports into my lxd container? [10:01] how to get pip for python.36 working on 19.04? [10:01] python3.6 [10:03] 19.04 isn't close to out yet [10:04] Ben64: well, I have it installed on my computer [10:06] ok, but it is supported in #ubuntu+1 until release, which is in April [10:06] Ben64: thank you! [10:09] Anyone know if this ticket does what im trying: https://github.com/lxc/lxd/issues/4512 [10:10] Hello. Is it possible to show a DVD-movie via some video player (vlc for example) on a TV using a hdmi-cable? Are there any things I should consider or would that work out of the box? [10:10] I mean, playing the DVD on Ubuntu and showing the signal on a connected TV, if that wasn't clear. [10:12] What id like is to create a lxd container that is restricted to 2 core 8GB of ram. I will have a steam game running as a server. Now this game need inbound traffic to be open at ports 7777,7778,27015, 27017 and TCP via 27020. For the life of me i cannot figure out how to get the firewall open at all. I have tried disabling the ufw on both the host [10:12] and the container to no avail. When I finally gave up and ran the server on the host boom everything worked out of the box. This mean either 1) the filesystem in lxd is creating issues or 2) Firewall is not allowing the direction of incoming traffic through the nat btw host<->container [10:13] Mrokii, for info on playing dvds - https://help.ubuntu.com/stable/ubuntu-help/video-dvd-restricted.html.en i haven't tried it in ages, but I've connected fine via hdmi to tv [10:13] guiverc: Playing DVDs works fine, that's not the problem. I was just wondering about getting the signal to the TV. [10:16] I've connected to tv with no problems, and with problems (some where the edges aren't shown, the tv overscans I think it's called) - on some tv's it's perfect, other tv's have provided me hassle (where the tv seems to have trouble with any computer/os so it's tv & not Ubuntu in my experience) [10:16] guiverc: Okay, thanks, so I guess I have to give it a try. [10:17] guiverc: Just in general, do you know if signals are transferred automatically to a connected TV or is there usually some option I have to tweak, like redirecting the signal or something? [10:19] no sorry I can't help there Mrokii - i have no idea how it works, just talking from my experiences.. [10:20] okay, thanks. [10:20] I'll try and see what I can achieve. [10:26] Is it possible to access the files on an Ubuntu PC if I have forgotten the password? [10:28] ddoobb: As long as the files aren't encrypted, there should be no problem. [10:29] Mrokii how can I do this? I am seeing some tutorials that do it with "recovery mode" prompt. [10:30] hello [10:31] ddoobb: As long as the files aren't encrypted they are accessible at least as a superuser (using sudo for example). I have never had that problem, so I'm not sure if there's a specific way to reset the password. [10:31] I don't need to reset the password, just copy them to an external drive [10:34] ddoobb: The easiest way to me seems to be to boot from another medium, then use "sudo" from the terminal to become superuser and then to use "cp" to copy the files from one drive to another. [10:34] Thanks I'm going to try this Mrokii [10:35] ddevault: You're welcome. Hope it helps. [10:36] "Don't reinvent the wheel", but I can't just find a simple way to mirror a part of ubuntu repository without it relying too much on sources configuration of my PC or requiring httpd or ftpd. Say I want gedit, latest version for amd64 (http://archive.ubuntu.com/ubuntu/pool/main/g/gedit/gedit_3.30.2-1_amd64.deb) along with all its dependendies. Then let apt take packages from there. Does apt support filesystem paths or file:// at all? [10:36] I think I'll end up writing something myself. === SimonNL is now known as _SimonNL [11:06] geirha, are you here? [11:07] hey anyone knows of a dlna server that can push content to the receiver? [11:07] i.e the the player becomes the server [11:07] in Windows WMP could do this using PlayTo [11:41] afx_: yes-ish === astra` is now known as astra === astra is now known as ASTRA === ASTRA is now known as astra` === astra` is now known as astra [11:58] I've given up on lxd to do what i wanted it to do [11:59] Anyone know how to route incoming traffic from internet (0.0.0.0) -> (localhost:port) -> (kvm:port) [12:00] thagabe: perhaps a question more for ##networking ? [12:02] geirha, sorry to interrupt! I am going to add my user to the input group but I am wondering if this is a secure approach to resolve the permission issues I have with python-evdev and /dev/input/xxx [12:06] lotuspsychje I'll give that a shot! Thanks [12:08] afx_: well, the user will get access to all the input devices [12:09] geirha, can I give access to a specific /dev/input somehow ? Are there any keylogging concerns to think about? [12:10] if you don't trust the user, it may be better to force the user to use sudo instead [12:11] e.g. In sudoers you can specify that user X are allowed to run (only) your python-script as the group input [12:12] and the user would then execute your script with sudo -g input yourpythonscript [12:13] I am the only user and I am a sudoer , so in case someone steals my password it doesnt really matter what group I am in right ? [12:13] afx_: maybe you could use an ACL or something on the specific /dev/input? [12:14] EriC^^, let me read about ACL [12:15] if you have sudo and someone gets your password then he can do whatever he wants [12:16] EriC^^, exactly! But what about a malicious keylogging program for example that is run under my user account? [12:17] afx_: sure that could be done [12:17] there's actually a keylogger in the repos i believe [12:19] did I Ident proper? [12:19] BrickedInTheHead: if you were able to join, then yeah [12:19] EriC^^, :O ! I didn't know that is a ready made solution for this :) [12:20] some idiot made password of 10x same digit and fat fingered one in the middle. Anyway Ill be bothering you here setting up this new environment [12:20] EriC^^, So atm without my user account being part of input group, can a keylogging program get my keyboard input ? [12:20] afx_: i'd think so, but anyways they could always add your user to the input group [12:21] EriC^^, let's say you don't have my passwrd ,and somehow you managed to install a keylogger .. By having the account out of the input group will the keylogger get my input ? [12:22] I am trying to understand why a user account gets restricted from accessing /dev/input/ and this is the only thing that comes to mind [12:22] i have no idea, i'd think so, as i dont have my user in the input group but 'xev' shows key strokes [12:24] ah but xev only shows the strokes if it's selected as the terminal [12:24] afx_: try it and see i guess, try installing the keylogger, you'll know at least if it's easily possible, or possibly just harder to do [12:25] !info logkeys | afx_ [12:25] afx_: Package logkeys does not exist in bionic [12:25] !info logkeys xenial | afx_ [12:25] afx_: logkeys (source: logkeys): keylogger for GNU/Linux systems. In component universe, is optional. Version 0.1.1a+git5ef6b0dcb9e3-2 (xenial), package size 32 kB, installed size 120 kB [12:26] I see, so its not in bionic [12:26] yeah seems so [12:26] xev as the terminal shows strokes , but I guess you can have a software emulating a terminal [12:27] so it can forward any input you do [12:27] afx_: yeah but xev is taking the input you're giving it i think so maybe it's different like that [12:27] let me try installing logkeys i'm on xenial [12:28] ok [12:29] afx_: it works, even though i'm not in the input user group [12:29] i guess since it uses sudo it can access /dev/input anyways no need for group chananigans [12:30] it does mention in the man page that it uses /dev/input to get the key logs [12:30] EriC^^,logkeys needs sudo ? [12:30] hi all, trouble with skype 8.34.0.78, my mic does not seem to work, it looks ok on gnome device settings though, any tips? [12:30] afx_: yes, it asks for sudo [12:30] EriC^^, this makes sense, as with sudo I can run the python script in question [12:31] I would like to use this but stay on the safe side too .. [12:32] use what? i wasnt following earlier [12:32] EriC^^ yes sorry . This is what I am trying to use https://gist.github.com/t184256/f4994037a2a204774ef3b9a2b38736dc [12:33] it uses evdev which needs access to /dev/input [12:33] EriC^^, also found this https://askubuntu.com/questions/1057120/is-adding-a-user-to-input-group-secure if that helps in any way [12:35] afx_: if they dont have your password i think you'd be good [12:35] just add to sudoers to run the python script only and set the script non writable and in a dir it cant be deleted and recreated in [12:36] EriC^^, the SE answer suggests removing the w/r access from users not in the group (???) [12:37] Hey. (asked in #quassel, anyone know where to get not so glitched one). And bit laggy with nvidia 560ti switching tabs. Just went with default drivers, but dont think messing with nvidia is worth it. How about disabling some of the 3d effects? [12:37] EriC^^, can you give me any instructions for this please ? [12:37] I want to install postgres server but *not* start the service right now. Is there a way or just use Docker (new toy in office ;))? [12:38] TEST [12:38] (oh, Thunderbird screwed up) [12:38] afx_: put the script in like /opt/pythonscript.py and chown root: then chmod 770 [12:39] afx_: and in sudoers add the line ALL= (ALL) NOPASSWD: /opt/pythonscript.py [12:40] afx_: then run the script as "sudo /opt/pythonscript.py" via some startup file or something i guess [12:41] EriC^^,thank you ! [12:41] Will try that out and let you know [12:41] afx_: no problem, ok [12:42] make sure you use ''sudo visudo'' to edit /etc/sudoers. Don't edit it directly [12:43] Hello community, is it possible to encrypt my /home on ubuntu 18.04, i dont need full disk encryption only /home, which has all my files, what software can i use for this? and i need to decrypt upon boot as /home will be read upon successfull login [12:44] with TMP chip you should be able to get reasonable security without any upkeep. Likely requires uefi boot [12:44] tpm [12:44] pagios: ubuntu has a built in home encryption that uses ecryptfs [12:45] pagios: sudo adduser --encrypt-home creates it, it'll wrap the passphrase (be sure to write it down somewhere) using your login password, so once you log in it'll automatically decrypt it for you, obviously use a good login password as well [12:46] BrickedInTheHead: The level of configuration and mucking around required to get a TPM working under Linux makes it not a great recommendation for someone that has to ask how to encrypt stuff in an IRC channel. [12:47] pagios: create another temp user, move your files there, then delete and recreate your old user using adduser and move the stuff back after logging into the account [12:48] pagios: So you know, only encrypting /home gives very little -real- security. It leaves innumerable places for remnant data to exist: temporary files, swap space, etc. [12:48] one second [12:49] how come its not secure if i use home encryption? [12:49] I just said that. [12:49] unless the attacker guesses the pass [12:49] swap will die on system shutdown [12:49] is there a suite to full drive(s) crypt that doesnt break everything? [12:49] temp files too [12:49] No, it won't. [12:49] well, temp files get erased upon next booting, not while shutting down [12:49] On boot, yes, but not on shutdown. [12:50] but temp files are on /home no? [12:50] in theory if the attacker has access to the pc, he could easily get the password next time you log in [12:50] pagios: No. [12:50] on drives ssd/hdd the data just gets de-adressed instead of format [12:50] he can boot using resuce disk, reset the user pass no? [12:50] There's also the reality that remnant data does not get removed upon deletion. Only unlinked. [12:50] and login as decrupted [12:51] pagios: You are strongly advised to use LUKS or something similar to encrypt the entire system. [12:51] pagios: even if he resets the password, it wont decrypt it, the actual literal login you used will only work, cause it was used with the passphrase to encrypt it [12:51] but my system is already running i dont want to format [12:51] pagios: Do you want the security that encryption provides or not? [12:51] yea [12:52] LUKS. I'll type that down, hope it can do chip + pass crypt [12:52] i see EriC^^ [12:52] pagios: Then you'll need a rebuild. [12:52] his only option is to bruteforce then [12:52] Because /home-only encryption is basically crap. [12:52] pagios: i think it depends what you're worried of, if the laptop getting stolen is an issue, then i think home encryption would be enough [12:53] pagios: if you're worried of meddling then LUKS would be better [12:53] pagios: If someone's going to steal the system with the intent of stealing data, digging through remnant data on the disk is task #1. [12:53] If they don't care about data, they won't even notice its encrypted. [12:53] i am gonna put a desktop in a shared environment [12:53] and worried about someone stealing the hardisks [12:54] pagios: Then you need to encrypt the entire disk, not just /home. Easiest way to do it is to create a LUKS container on the disk, use that as an LVM physical volume, and create your LVs inside that. That way you only need to unlock one disk, instead of every filesystem. [12:55] Too lazy to rm old assets under NDA so locked for just that. also hackmaster, typing that 24 digit passkey [12:55] i dont have lvm [12:56] pagios: You can configure it when you rebuild the system. Yes, you need to rebuild the system. [12:56] pagios: ubuntu will do everything for you, just reinstall using 'encrypt disk with LUKS + LVM' [12:56] ^ [12:56] pagios: then copy your data back and set things up again [12:56] trying to avoid a reinstall here [12:56] pagios: Too bad. I thought I made that clear. [12:56] so a reinstall is needed [12:57] 07:52 pagios: Then you'll need a rebuild. [12:57] Yes. [12:57] just got things running "smooth". might as well upgrade to newer LTS while at it then [12:57] well not necessarily, it'd just be more difficult to set it up manually, not so much i image though === tsglove2 is now known as tsglove [12:57] pagios: you dont need a reinstall, you can convert it to LUKS but you'd have to do the legwork yourself and not rely on the installer [12:57] EriC^^: Migrating the entire system to new filesystems and hoping that the initramfs and bootloader don't vomit on you later? [12:57] No thank you. [12:58] iosecure: um, he can edit the UUID and everything and update-initramfs etc [12:58] UEFI hehe [12:58] EriC^^: It is NOT that simple. Sorry. [12:58] yes it is [12:58] I've done that kind of conversation. It sucked. [12:58] No, it isn't. [12:59] how did it suck? [12:59] initramfs refused to recognize the LUKS container, just sat there screaming about how it couldn't find the root volume. === andyhuzhill_ is now known as andyhuzhill [13:00] No amount of triple-checking configs and rebuilding the initramfs would change its mind. === _SimonNL is now known as SimonNL [13:00] i will buy a new hardisk , install on it the OS and then add those hardsisk to the lvm [13:00] well you missed the initramfs parts of the conversion i'd guess [13:00] EriC^^: You missed the "triple-checking configs and rebuilding the initramfs" part of what I said, I'd guess. [13:01] i will install a fresh new OS on a new hardisk and add those disks at a later stage, wjats wrong with that [13:01] EriC^^: So apparently neither sarcasm nor reading comprehension are your forte. [13:01] iosecure: it's a computer, if everything is right it'll work, if not it wont, so yeah you missed something [13:02] iosecure: no need to get nasty, i did understand what you said, i just dont agree [13:02] so? [13:02] pagios: Honestly, the easiest thing you could do is rebuild the system on a new disk, with LUKS/LVM in place from the installer. Copy /home to the new system. [13:02] anyways this is all going offtopic very quickly, so let's just stop shall we [13:03] iosecure, ok [13:03] EriC^^: Nice walk-back. If you're going to pull out condescending crap like "well you missed the initramfs parts of the conversion i'd guess," don't be surprised when someone comes back in kind. [13:03] EriC^^: And I wasn't asking if you agreed to begin with. [13:04] iosecure: how is that condescending? [13:05] pagios: I'd advise against coping anything other than /home. That can cause weirdness, depending on what you copy. [13:05] iosecure: you're reading into stuff a lot it seems, i just dont want pagios to think it's so difficult and 'just wont work' you said it didnt for you, i said you must have missed something for it not to, geez [13:06] thank you both, and apologies if that created a clash between you guyus === TheHonorableKitt is now known as ThKitten [13:07] No clash at all. [13:07] no worries [13:08] Is the guest account similar to live boot? does it have any/configurable storage? [13:08] BrickedInTheHead: it's very different and limited, i think you could add storage manually [13:09] There used to be. I'm not sure if it exists, still, as various transitions somewhat broke the functionality. [13:10] I did set 26GB swap. that probably is included [13:10] 26GB swap? Is that a typo? [13:11] more is less here? [13:11] How much RAM do you have? [13:11] 16GB [13:11] Ouch, yeah. Overkill. [13:12] Couple questions: Is this a desktop, and do you want hibernation functionality? [13:12] (or laptop) [13:12] Desktop yes. And probably, ubuntu takes forever to boot. and the bios time is at least 25s on this motherboard [13:13] Gotcha. Okay, then it isn't THAT much overkill. I'd bring it down to 20GB, save a few gigs. [13:13] But you probably have enough storage that 6GB isn't a big deal. [13:14] hey guys [13:14] This HDD will die any day. 6 years running. Seagate quality +1 [13:14] does gnome-tweaks not work on ubuntu 18.10 ? [13:15] BrickedInTheHead: Oh, then totally screw with the swap sizing. Live life on the edge. ;) [13:16] I have recently migrated from Ubuntu 14.04 to 18.04. In 18.04, very few applications show in the System Tray. Dropbox is consistently there, CopyQ (the clipboard manager) too shows up whenever it is on. But Pidgin, to provide an example, does not show an icon. Is this the expected behaviour in 18.04? [13:17] The motherboard error checks blinking like disco lights already for years. Anyway offtopic.... [13:23] rasmalabar: how did you migrate, which desktop are you using? this is not the default behaviour on 18.04. [13:33] @tomreyn: I re-installed, from scratch. I am Ubuntu with GNOME 3.28... [13:34] Check install media for faults [13:35] rasmalabar, you probably need an extension ; many tasks have been moved to that [13:36] rasmalabar, and, btw, has been reported as a bug (!): https://bugs.launchpad.net/ubuntu/+source/pidgin/+bug/1786084 [13:36] Launchpad bug 1786084 in pidgin (Ubuntu) "Indicator does not show as tray icon on 18.04" [Undecided,Confirmed] [13:39] ioria, You mentioned about the extension, which one is it? [13:39] rasmalabar, go here https://extensions.gnome.org/# ; type 'pidgin' and see if there's something useful for you === TheHonorableKitt is now known as ThKitten [13:48] ioria, ubottu: Thanks for the tips and will check further after going offline as it might take a while to decide on which option to choose... Thanks! [13:48] rasmalabar, ok [14:06] hi there, is it normal to have "Ubuntu 16.04.5 LTS (4.15.0-42-generic #45~16.04.1-Ubuntu SMP Mon Nov 19 13:02:27 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux)" showing unbalanced CPU usage across cores? i launched gnome-system-monitor just for fun only to encounter cpu4 is constantly peaking between 80/100% and the other 3 are almost 0/10% ? [14:06] shouldn't cpu usage be balanced between the N cores avaiable? how can i debug this? [14:08] A process can only reside on one core. [14:08] whallz: no that's normal [14:09] what software is it that only runs on one? [14:09] it should be optimized to run on more (eventually) [14:09] So if a process starts, then later starts to consume large amounts of CPU resources, it's going to max out the core where it is. [14:09] tarzeau: Only if the application is spawning multiple processes. [14:09] Plenty of applications are single-process. OpenVPN is a good example. [14:10] i have only google-chrome and guake with tmux running [14:10] iosecure: or threads? [14:10] running "top" (better "htop") should tell which application is keeping this core busy [14:11] if i close google chrome they all balance out around 10% for a while, then cpu4 goes back up to 80% constant [14:11] lets see [14:11] iosecure: plenty are multi [14:11] iosecure: objc2 also uses multiple cores (relevant for gnustep) [14:11] tarzeau: And plenty are not. [14:12] tarzeau: My statement is correct. A single process is resident on a single core. [14:12] unless it's a single process supporting multithreading [14:12] Multithreading is not multiprocessing. [14:12] A single process that spawns multiple threads is still single-core resident. [14:12] i know, that's why your statement is not correct [14:13] Yes, it is. [14:13] iosecure: no it's not [14:13] Yes, it is. [14:13] By definition, yes, it is. [14:13] ok depends on multithreading, but how do i find in top/htop which one runs on just 1 core? [14:14] whallz: Any process shown in top/htop/ps is running on a single core. [14:14] https://imgs.xkcd.com/comics/duty_calls.png [14:14] Xorg seems busy too xD [14:14] whallz: Any application that shows multiple copies in those programs is potentially running on multiple cores. [14:14] iosecure: i mean, i a specific core [14:14] whallz: Ah. Not sure how to show which core a process resides on. [14:14] iosecure: right, that what i need to find out [14:15] something is killing cpu4 while cpu1, 2 and 3 are drinking mojito [14:15] need to find out what that process is [14:15] Likely a single runaway process. [14:16] right, but is it grabbing always the same core? is there an un-even degradation of whole cpu? [14:16] No idea. Processes generally do not have an assigned affinity. [14:16] first, i still need to know which process is it top and htop dont help [14:16] Whichever core it starts on, it stays on. [14:17] Guys I am running Compuwiz found here https://code.mradford.com/post/the-ubuntu-compiz-desktop/ How can I use gnome-tweaks with this? [14:17] whallz: you can configure htop to show the "PROCESSOR" column [14:18] herald85: that's what i need [14:19] open htop, press F2 , go down to 'columns' [14:20] awesome, thanks herald85 [14:20] i am also on ubuntu 18.10 [14:22] ok, seems to be /usr/bin/Xorg wtf [14:22] sorry /usr/lib/xorg/Xorg [14:25] hello all can anyone tell me why there are two of every application one is small in size one is large in filesize...eg vlc from the sofwware center has two choices of the same version but different file size [14:25] why does the software center have two of every application now days? [14:26] maybe each is an architecture specific? [14:26] cpu4 always seems busier that the other 3 [14:26] i only have 1 vlc in software center, ubuntu 16.04 unity here [14:27] :( [14:27] I am running Ubuntu 18.10 and re architechture do you mean different versions of ubuntu eg 8.0418.10 [14:27] conjo: no i mean 32 bit and 64bit [14:27] 1810 vs 18.04 [14:27] okay [14:27] Hey folks [14:28] BluesKaj, hey guy [14:28] i see no mention of architecture under details how do i tell the difference [14:29] BluesKaj, I keep saying hi back, since you are one of the few I remember from way back ") [14:29] hi coz_ [14:29] where in 18.04 is the network interface names defined? [14:30] yeah, coz_ guess we've been around a while [14:30] whallz: that sounds familiar, gnome-shell is horrible with software drawn windows. Do you have a slow cpu ? atom? old celeron? [14:30] BluesKaj, ") for sure [14:33] griphook: i would not recommend using compiz on 18.04 and newer. it's mostly considered to have failed now. https://www.reddit.com/r/linux/comments/75hxp6/why_did_compiz_die/ - better stick to gnome-shell and, if needed, customize it with extensions. [14:34] ...and gnmoe-tweak-tool [14:34] zetheroo, which "network names" ? [14:34] BluesKaj: eth0 eth1 etc ... [14:35] zetheroo, /etc/network/interfaces [14:41] zetheroo: you may want to read this if you haven't https://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ [14:46] we have a system with 18.04 which hasn't got ifup installed and there is a network interface installed as eno1p (at least it seems to be that in the /etc/netplan/*.yaml file) but how do we get the interface up? [14:47] we cannot install ifup because it's not online [14:47] also it seems that when we do 'service networking restart' there is no 'networking' service :/ [14:47] ip link set dev eno1p up [14:48] pleasse read the !releasenotes [14:49] cannot find device eno1 [14:49] Hi, I am on 18.10 and I closed my laptop and opened it [14:49] and now my screen is all black [14:50] i only see the title bar up at the top [14:50] zetheroo: "ip link" shows available devices [14:51] any ideas how to get the gui to pop up again? [14:52] zetheroo: as a result of reading the release notes, you should use use 'systemctl', not 'service', and use ip (from iproute utils), not ifup/ifdown (from ifupdown utils) [14:52] nvm I figured it out! I just did ctrl alt f1 and logged in and everything reloaded fine [14:52] phew [14:53] thanks for listening lol [14:54] tomreyn: ok, thanks. The netplan file had the wrong interface name. [15:01] heya [15:02] hey everyone [15:02] best linux native games? [15:02] share your opinion [15:02] 16.04->18.04 failed yesterday due to "cannot calculate upgrades" on do-release-upgrade. cannot figure what's offending, so I think of "upgrade from ISO", but is there an option to install over ex. system and keep custom /etc etc. or is it "raze root or bust"? [15:03] meph1st0: try #ubuntu-offtopic. This is a support channel [15:03] leftyfb, OK, THANKS :) [15:03] sorry for caps) [15:05] my ubuntus keep failing to download http://be.archive.ubuntu.com/ubuntu/pool/main/l/linux/linux-headers-4.4.0-140_4.4.0-140.166_all.deb , but after some experimenting I've noticed that using ftp:// works. Whats the best way to notify the repo maintainers? General ubuntu bugzilla? [15:08] deadrom, I have to ask , didyou update/upgrade existing packages first? [15:08] BluesKaj, yup [15:11] deadrom, ok, is the package manager set to upgrade LTS only? [15:12] BluesKaj, yes [15:14] any ppas? [15:14] BluesKaj, truckloads. upgrade-core tells me it disables those, though [15:15] or changes to the sources.list [15:15] deadrom, yeah, it's supposed to [15:16] BluesKaj, http://paste.ubuntu.com/p/b5GkD49Ddb/ <- current sources.list === [1]MrMobius is now known as MrMobius [15:20] btw I cannot paste proper outputs becuase LC_ALL=C gets me this: https://pastebin.com/Xdr1qpVL [15:32] deadrom, use 'LC_ALL=en_US.UTF-8' [15:33] You need UTF-8 support [15:48] can ubuntu lts support 12 cores? [15:48] koffel: ... yes [15:48] koffel: unless you mean some specific hardware version or something, but 12 cores in general, yes. It can support 100s of cores. [15:48] some reason i can only get it to see 6 [15:49] koffel, what CPU? [15:49] dual xeons [15:49] Since most processors use some sort of threading these days, make sure you have HT/SMT on [15:49] yes i do [15:49] it shows in bios 24 cores [15:49] koffel, ... more specific. What CPU model? Also, which Ubuntu LTS? 18.04? [15:50] e5656 16.04lts [15:50] koffel: 12 or 24 cores? and do you mean cores or hardware threads [15:50] my unit has 4 e5656 xeons 6 cores each [15:51] koffel: `dmesg`, `lscpu` in a pastebin [15:51] I don't see the "e5656" in Intel Ark. Sure that's right? [15:52] sorry i mean e5670 [15:52] Intel® Xeon® Processor X5670 [15:52] Ok, so each one is 6 cores, 12 threads [15:53] yes [15:53] koffel, and what tool are you using, which is showing 6 cores? [15:53] lscpu [15:53] and top [15:54] top doesn't show cores, it shows logical cpus [15:54] koffel: and again, can you pastebin the output asked for? [15:54] some reason in lscpu model is blank [15:55] koffel: did you buy it after-market? [15:57] hi! is there a lxc 2.0.9 package for ubuntu 18? can't find one online :( [15:57] no nacc i bought it new in my dell poweredge r720 [15:57] https://pastebin.com/3DpEBcpr [15:58] koffel: uh, you installed 32-bit [15:58] koffel: that seems wrong [15:58] and actually, that looks likea VM [15:58] oh shit i thought it was 64bit [15:59] "linux-Standard-PC-i440FX-PIIX-1996" [15:59] not vm [15:59] just to santiy check, are you trolling? :) [15:59] oh did you accept the random hostname it provides at install time? [15:59] yes [15:59] in any case, i'd install 64-bit and see if it sees your processor correctly [15:59] it a server [15:59] will do [16:00] koffel: i would suggest giving a real hostname as well [16:00] will do [16:00] thank ya [16:00] styx__: why would you want 2.0.9 in 18.04? it's at 3.0.3 [16:00] nacc: project reasons :/ [16:01] styx__: afaict, 2.0.9 is not in any ubuntu [16:01] nacc: hmmm and any 2.* version? [16:13] styx__: of LXC? 2.0.7 in Xenial, 2.0.8 in Xenial Backports, but that's it. (1.x in Trusty) [16:13] bionic and newer have 3.x [16:18] styx__: sorry, what teward said, there is the snap option as well, but that's for lxd [16:19] styx__: if your project *must* have 2.x of LXC (and NOT lxd) then you must use Xenial 16.04 - the older LXC versions are not available in 18.04 or newer. [16:24] Has anyone worked with NS3 on ubuntu here? [16:31] !details | Pl [16:31] Pl: Please elaborate; your question or issue may not seem clear or detailed enough for people to help you. Please give more detailed information; for example, we might need errors, steps, relevant configuration files, Ubuntu version, and hardware information. Use a !pastebin to avoid flooding the channel. [16:32] Pl, also, NS3 could mean name server, nonstructural protein 3, network simulator, and many more... please don't abbrviate [16:33] *abbreviate [16:33] pragmaticenigma: Hey sorry. I meant the network simulator. The C++ version. Not OTcl. [16:33] Got a semi ubuntu related question [16:33] Pl, what version of ubuntu are you running [16:33] !ask | CheggyCheeps [16:33] CheggyCheeps: 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 [16:33] ah ok [16:34] Sorry guys. This might be off-topic. [16:34] pragmaticenigma: 16.04 [16:34] I used to use a peice of software years back pre Ubuntu, for debian for displaying disk/cpu/mem statistics down the side of the screen. Any idea what it might be called or similar packages [16:35] Pl, yeah, sorry, network simulator doesn't appear to be a package or snap provided by Ubuntu or it's official channels. You might want to try in #networking or reaching out directly to network simulator community [16:35] !info ns3 xenial | Pl [16:35] Pl: ns3 (source: ns3): discrete-event network simulator for Internet systems. In component universe, is optional. Version 3.22+dfsg-2build1 (xenial), package size 105 kB, installed size 338 kB [16:36] CheggyCheeps: try conky? [16:36] That'll do it! [16:36] Thanks, [16:36] pragmaticenigma: Sorry. I did try #networking but did not get a response so I thought someone here can help. But I understand that it is completely off-topic. [16:36] pragmaticenigma: ns3 might be the specific one they're talking about (See ioria's call to ubottu) [16:37] Pl: are you talking about the 'ns3' package or something else? [16:37] teward: Yes, I'm talking about the discrete-event network simulator. [16:39] I'll post my script anyway. If anyone has any thoughts, well appreciated! :) Link: https://paste.ubuntu.com/p/G4vGkXcq5F/ [16:41] !info libns3-dev xenial [16:41] libns3-dev (source: ns3): Static and header files of ns-3. In component universe, is optional. Version 3.22+dfsg-2build1 (xenial), package size 6417 kB, installed size 55989 kB [16:42] Pl, try to install that ^ [16:43] does anyone know the correct incantation to make apt-get fully accept anything thrown at it? [16:43] ioria: Hey thanks! I'll try that. [16:43] apt upgrade -y isn't it; neither is apt-get dist-upgrade -qq --force-yes\ [16:43] got_me_an_irc: what are you trying to do, and blindly telling apt to accept all prompts is not recommended [16:43] it seems to get stuck (if unattended) on stuff like /etc/default/grub. I've tried removing /etc/default/grub before issuing the upgrade, doesn't work either [16:44] pragmaticenigma: i'm trying to create an up-to-date ec2 AMI [16:44] got_me_an_irc: If you have customized those files, apt will not be able to automatically update them if they are included in an update. It requires user intervention [16:44] so basically I want an unattended script I can run that gets me from their image to "now" [16:44] pragmaticenigma: right, but there are options. I want it to take the "just do it" option. [16:45] got_me_an_irc: It can't [16:45] the first option is "just do it" e.g. overwrite it and don't care. [16:45] so it's impossible to do an unattended upgrade [16:45] How do you install perl modules ? i.e. Automake::Config [16:46] there are results online for people who have had this issue on 12.x, 14.x, and so on, I just need the incantation that works on 18.04 [16:46] got_me_an_irc: No, what you should do is make sure you're not updating package maintained files and leverage the appropriate means of adding customized preferences [16:46] such as many programs offer a conf.d folder where you place system specific configurations to override the defaults [16:46] pragmaticenigma: i'm going from the base image on EC2 with no customizations [16:47] sure, it's their issue for having modified it, i assume someone else must have had this problem though [16:47] If you have seen those incantations they should still be working. without an example to work from, i'm shooting in the dark to help [16:48] How do you install perl modules, system wide ? i.e. Automake::Config [16:48] !patience | WoC [16:48] WoC: Don't feel ignored and repeat your question quickly; if nobody knows your answer, nobody will answer you. While you wait, try searching https://help.ubuntu.com or https://ubuntuforums.org or https://askubuntu.com/ [16:48] @got_me_an_irc: can you please re elaborate the issue .. [16:48] Was clarifying ;P [16:49] nisankhindia: take the base ubuntu 18.04 image on amazon EC2 and try to apt update and apt upgrade. [16:49] you can't do it unattended. i want the incantation to do that upgrade unattended. [16:49] specifically the issue is that it appears amazon have modified /etc/default/grub or /boot/grub/menu.lst or something [16:50] @got_me_an_irc: ok so in your case have you asked or dropped any query on amazon support [16:50] okay. just imagine I asked about not being able to find the X button in the corner or something, nevermind. [16:52] @got_me_an_irc: I do suggest to discuss this with Amazon EC support [16:54] nisankhindia: They already left [16:57] I've installed ubuntu server two times in the last month on 2 different machine. Both times I had to revert from the new installer to the old alterative one due to lack of previously available features [16:58] eg: install from serial terminal and boot from raid1 partition [16:58] i have an ubuntu 18.04 on a super new computer (nvme on pci), however the network interface doesn't come up (systemd-networkd), i have to unplug/replug and re-run restart systemd-networkd; autofs every time. known problem? [16:58] hi, does anyone know what could make grub boot into slic loader? [16:59] i don't even know what the heck it is. I just upgraded to 18.4 and all of a sudden I can't boot into Windows anymore [16:59] the very same machine also doesn't boot with 18.04 on usb stick, hangs at: [ ok ] started session c79 of user gdm. [16:59] (many of these starting from c41) [16:59] some creepy shit pops on the screen about memory addresses [17:01] it's got a graphics card like GeForce RTX 2080 Rev. A in. [17:01] @tarzeau: system details like network interface make , log details e.g system log etc are helpful to identify ... if you can provide these we can check about the issue [17:02] nisankhindia: 0000:06:00.0 Ethernet controller: Intel Corporation I210 Gigabit Network Connection (rev 03) [17:02] @tarzeau: boot log , system logs etc also required [17:03] nisankhindia: boot log: https://people.phys.ethz.ch/~myczko/dmesg.txt [17:03] nisankhindia: can you be more specific what exactly of system log you want? for the usb stick 18.04 i don't have any logs since systemd doesn't give me a chance to get a shell by default [17:04] without saying init=/bin/bash, and i'm not at the machine now [17:06] @tarzeau: alright .. let me go through the dmesg output [17:06] also the preseeded install would fail (bios doesn't boot the nvme on pci), manually partitioning the preseeded stuff, worked [17:06] it's 12 cores, 256 gb memory [17:07] we're using nvidia 410 driver with cuda 10 with tensorflow (py3) [17:09] tarzeau: if you've got very new hardware, bear in mind the 18.04 kernel is relatively old (v4.15 released in late 2017) - you might need to use a recent kernel from the HardWare Enablement (HWE) stack (v4.18), or even a mainline build of a later version [17:09] !info hwe | tarzeau [17:09] tarzeau: Package hwe does not exist in bionic [17:09] !hwe | tarzeau [17:09] tarzeau: The Ubuntu LTS enablement stacks provide newer kernel and X support for existing LTS releases, see https://wiki.ubuntu.com/Kernel/LTSEnablementStack [17:10] TJ-: good idea, i can try that! [17:10] it's very new hardware, got it 2 days ago, bios dated a few weeks old [17:10] linux-image-4.18.0-1004-gcp - Signed kernel image gcp [17:10] linux-image-4.18.0-1006-azure - Signed kernel image azure [17:10] linux-image-4.18.0-1006-aws - Linux kernel image for version 4.18.0 on 64 bit x86 SMP [17:10] is one of these what i want? [17:11] tarzeau: see the link provided by ubottu [17:13] tarzeau: I think you'll need to enable the bionic-proposed archive to get the 4.18 kernel (which comes from Ubuntu 18.10) [17:14] * tarzeau will try, thanks! [17:15] @tarzeau: bionic proposed must be enabled to get 4.18 ... [17:15] i see, so just add bionic proposed to sources.list [17:16] Anyone about? [17:17] maeud, many of us. What do you need help with? [17:18] @maeud: ask anything related to ubuntu issue or discussion === capella|away is now known as capella [17:24] tarzeau: uh, don't just add bionic proposed. [17:24] tarzeau: that will make your system install all -proposed packages [17:25] Hi nisankhindia, I'm trying to preseed an Ubuntu Bionic installation. I've created an offline repo using aptly which I can use without issue on an installed system, but trying to use it to actually install I get an error: "anna[3950]: WARNING **: bad d-i Packages file" [17:26] so it works fine when adding it as a source in apt, I can install other packages, update etc [17:26] but when I use that mirror to do an installation it throws that error [17:26] maeud: did you mirror the installer udebs? [17:27] maeud: they aren't ussed by anything except the installer, so your installed system wouldn't care [17:28] Which package provides autotools, i only see the autotools-dev [17:29] That could be it nacc, I included packages and sources but not the .udeb option, but I have lots of .udeb files [17:29] /pool/main/o/openssl1.0/libcrypto1.0.0-udeb_1.0.2n-1ubuntu5.1_amd64.udeb [17:29] etc [17:29] maeud: i'm not sure, that was just a guess [17:29] I'll try re-creating with that option enabled and see [17:29] thanks :) [17:30] WoC: autotools isn't a command [17:31] nacc, i thought it was the combo of; autoconf, automake, libtool [17:32] Hi all! I have a problem with fstrim taking over 7 minutes to perform a TRIM on my 120 gb ssd, this is happens once a week with Ubuntu 18.10 never had this issue with Ubuntu 18.04 [17:33] any ideas about what could be wrong here? [17:33] do you issue a command, or how do you se it takes that long? [17:33] see [17:36] compdoc, I used systemd-analyze blame ton find out why my system is taking that long to boot, and it reported that fstrim took that long to perform, after that I manually tried sudo fstrim / and vailidated it [17:37] I should also add that I have an encrypted lvm on this machine as well, but that has existed long before I upgraded Ubuntu too 18.10 too [17:39] Zteam: I'd suggest reviewing the device's SMART report for indications it is wearing out [17:39] Zteam, any IO errors in dmesg or syslog? [17:40] TJ-, device has 2 bad sectors, but those has existed for over 3 years, and the disk passes the self-tests just fine :) [17:41] WoC: autoconf, automake, libtool [17:41] WoC: respectively. use apt-file :) [17:41] Zteam: Smart is still reporting only 2 bad sectors? [17:42] when the number of sectors grow, its time for a new one [17:43] pragmaticenigma, yes, those arent growing [17:45] Hi, I have a problem building GNURADIO: No package 'gmp' found No package 'mpir' found. I have search and try to install many packages but I don't arrive to found the good one .. Did someone know witch package is needed ? How to know it ? [17:45] Zteam: generally you need to consider other values, such as wear_levelling - these values are highly specific to individual make/models too, so you may need to research what they infer [17:45] lordcirth_, I can't find any errors there, it complains about a BIOS bug, however, I have already upgraded BIOS to latest version, no change of Mainboard since upgrading from 18.04 [17:47] TJ, all vaules are OK as far as Ubuntu thinks, I be happy to provide a screenshot or something if you like :) [17:48] lordcirth_, gives me german. but - later, gotta run [17:48] ugh, dd'ing the desktop iso is not working for me. should it? [17:48] https://www.ubuntu.com/download/desktop/ [17:48] gonna try a different sub stick, but just a sanity check from here would help,thanks [17:48] usb* [17:48] jorb, dd'ing it to a usb stick, like /dev/sdd? should work, yes [17:49] jorb, what's the exact command you are trying? [17:49] Zteam: I'm wondering if strace-ing the fstrim might reveal anything [17:49] dd if=ubuntu-18.04.1-desktop-amd64.iso of=/dev/sdb bs=8M [17:49] Zteam: what is the make/model of the SSD? [17:50] jorb, did you run 'sync' before removing it? [17:50] jorb, make sure u try with conv=fsync , to make sure it writes data direct after the command is completed [17:50] jorb: tip: if you want to speed up transfer, increase the block-size. It controls size of the in-memory buffers. I generally use >100M (sometimes 1G) for large sustained transfers [17:51] Last I checked, I didn't get much performance improvement past 1M [17:51] i used cp 2 days ago, was kinda surprised it worked [17:52] jorb: "conv=sync,fdatasync status=progress" also help [17:52] caipiblack, are you sure you need to build GNURADIO from source? [17:53] lordcirth_ Yes I, I already do it many times but not on this computer. [17:53] For some supports I need it [17:54] this is nice to test for the blocksize you wanna use http://blog.tdg5.com/tuning-dd-block-size/ [17:54] caipiblack, are you using pybombs, or build-gnuradio? [17:54] Or manually? [17:56] TJ-, here is output from strace with fstrim https://paste.ubuntu.com/p/bGYRV9Hv9j/ [17:57] TJ-, Kingston V300 (KINGSTON SV300S37A120G) [17:58] Zteam: that strace log doesn't look to be complete; I don't see a final 'exit' - also, I guess it could help to have timestamps! [17:58] lordcirth_ no, I am using this: http://blog.reds.ch/?p=43 I already done this on the past. [17:58] This installation works well with limesdr [17:59] nacc, i would love to use apt, but the versions are obsolete [17:59] Zteam: as in "strace -r fstrim /" [17:59] TJ-, sorry for that my bad [18:00] Zteam: i forgot myself so don't worry :) [18:00] WoC: i don't know what you mean. [18:00] Zteam: so used to reading kernel logs with timestamps the default [18:00] WoC: the versions apt talks about are the versions apt knows about based upon the last time `apt update` was run [18:00] how do I cancel this strace CTRL+C can't cancel it [18:00] WoC: they are not obsolete unless you are running an obsolete version of ubuntu, or if you aren't keeping your package lists current [18:01] les than 15 mins ago [18:01] Zteam: if it's in a syscall got to wait for it [18:01] WoC: ok, so what do you mean by obsolete? [18:01] disco dingo [18:01] WoC: that's not released yet, so you need to be in #ubuntu+1 [18:01] Zteam: also, use a log-file, as in "strace -o /tmp/strace.log -r fstrim /" then you can do "pastebinit /tmp/strace.log" [18:01] WoC: that's an ubuntu release, not an answer to my question, anyways [18:02] k [18:02] Zteam: if you don'T mind revealing this information, maybe you could also post dmesg? [18:02] / journalctl -b [18:03] TJ-, okey, wasn't sure how to automate that, since fstrim doesn't write to stdout, that's why I pasted it manually [18:03] as you can see guys.. do anyone found any issues with ubuntu bionic release with kernel 4.19.9 . https://usercontent.irccloud-cdn.com/file/abNfrdnT/irccloudcapture8407092872226339893.jpg [18:03] WoC: the version in 19.04 prerelease is whatever it is, e.g. for autoconf it's 2.69-11. apt-file finding the pacakge you want has nothing to do with the package's version. [18:04] nisankhindia: that's an invalid kernel for 18.04. [18:04] nisankhindia: if you are running mainline you are not supported [18:04] @mac 4.19.9 is not mainline [18:04] Zteam: notable kernel change between default kernels in 18.04 and 18.10 is much improved sata power management, this could be a side effect. [18:04] nisankhindia: ... uh yes it is. [18:04] nisankhindia: stable tree is a mainline tree. [18:05] TJ, I should tell you that I killed strace once before, releasing it writes to stderr so I killed it by closing the terminal, hope, that don't mess up the output if We run it again? [18:05] nisankhindia: in any case, the point being, it's not an ubuntu kernel. [18:06] nacc: but? [18:07] tarzeau: ? [18:07] tomreyn, yes, it complains about a BIOS bug or something with SATA, but mainboard is the same and BIOS is upgraded, so I don't know what to do with it [18:07] @nacc: ubuntu kernel ??? Ubuntu uses linux kernel ..... and it can be fetched from ubuntu itself [18:08] nisankhindia: the ubuntu kernel in 18.04 is 4.15.0-... [18:08] nisankhindia: there is no other supported kernel in 18.04 currently [18:08] nisankhindia, Ubuntu does make several changes to the linux kernel, so it not the vanilla kernel [18:08] Zteam: please post this bios bug message as well as any ata related messages [18:08] nacc: not just add proposed... [18:09] nacc: just bionic-proposed i'd add to get 4.18 kernels. i didn't understand the HWE page i got the link for (it's not listing 18.04 really) [18:09] tarzeau: 18.04 isn't in hwe yet [18:09] Zteam: also note there is also fstrim -v and read what the man page has to sayon it (though i concur on the fstrim approach in this case) [18:09] Zteam: I think you'll be ok as long as there isn't an fstrim process hanging about. Worth checking with "ps -efly | grep trim' [18:09] so how can i try a newer kernel on ubuntu 18.04? [18:09] tarzeau: bionic-proposed has many packages [18:09] the thing i was suggested to [18:09] nacc: i know about that [18:09] tomreyn, whole dmesg http://paste.ubuntu.com/p/NFxDMchWbJ/ [18:09] i saw 18.04.2-named openoffice packages the other day [18:09] tarzeau: you dont' just want to add the repository, as it will add many packages [18:09] something tells me: soon [18:10] tarzeau: you might just try the mainline PPA if you want to do a one-off test, but it's not supported [18:10] tarzeau: nor is -proposed technically [18:10] many package updates, and they are still security supported (if from main), even if proposed right? [18:10] tarzeau: no. [18:10] https://www.irccloud.com/pastebin/gVQzl6jj [18:10] tarzeau: -proposed has no support [18:10] i tried avoiding PPAs, we run our own repo, so i'll just be back building kernels myself again [18:10] tarzeau: they are for testing [18:10] aha got that [18:11] tarzeau: as is the mainline ppa [18:11] so there is no simple way to run a newere prebuilt kernel for ubuntu 18.04 as of now [18:11] nisankhindia: you don't seem to understand how ubuntu, downstream/ustream, etc work [18:11] tomreyn, dmesg | grep IO http://paste.ubuntu.com/p/7cSvxbJPvg/ [18:11] nisankhindia, notice how "Ubuntu" does not occur in that line, as it does on a normal Ubuntu system. [18:11] tarzeau: well, it depends on what you mean, the ubuntu mainline PPA is not a PPA like other, you download debs from it by hand [18:11] tarzeau: that's quite easy to do to test [18:11] Mine: Linux hostname 4.15.0-39-generic #42-Ubuntu SMP Tue Oct 23 15:48:01 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux [18:12] tomreyn, sorry man, I got confused and belived I already posted those :D [18:12] nacc: ah that sounds good, but linux kernel packages are like a bunch of packages i'll have to download [18:12] tarzeau: no, just two [18:12] tarzeau: or three, as the mainline builds don't have -extra, etc [18:12] nacc: which? linux-image and? [18:12] !mainline | tarzeau [18:12] tarzeau: The kernel team supply continuous mainline kernel builds which can be useful for tracking down issues or testing recent changes in the Linux kernel. More information is available at https://wiki.ubuntu.com/Kernel/MainlineBuilds [18:12] tarzeau: i think that page lists them, just the image and modules [18:12] nacc: I'd rather have a supported kernel package from cosmic rather than a mainline build, since the former is supported. as in as I have on my systems: linux-image-lowlatency-hwe-18.04-edge => linux-image-4.18.0-13-lowlatency [18:12] i'll need headers too since i have nvidia 410 driver plus cuda 10 [18:12] Zteam: please keep working with TJ,, i just meant to add to it for later. i think you were on a good path there. [18:12] tarzeau: ah, i doubt nvidia will work [18:13] TJ-: oh i didn't know -edge had been updated [18:13] * tomreyn reads dmesg [18:13] TJ-: last i checked i didn't see the an 18.04-edge [18:13] TJ-: oh it's in -proposed specifically [18:14] TJ-: yes, that would be ok as well, tarzeau [18:14] nacc: ye [18:14] TJ-: weird there's no non-lowlatency one? [18:14] so i'll just stick with what i have, it works, it's only a problem during the installation with preseeding and pxe, and another with systemd-networkd (which i'll get fixed somehow else) [18:15] why i have asked this coz as linus - linus git launchpad , and mainline crak things ... if people can revert here few more details regarding 4.19.9 use with ubuntu .. like issues and performance that would be helpfull [18:15] Zteam: tomreyn I see "\_SB.PCI0.SAT0.PRT2._GTF.DSSP" which, decoded, means something like South-Bridge -> PCI root bus 0 -> SATA controller 0 -> Power something-or-other and you can make up your own meaning for _GTF.DSSP! but it does point to an ACPI node that could be affecting the SSD [18:15] nisankhindia: no, that's not he purpose of this channel. [18:16] nacc; there is; I just prefer lowlatency on desktop systems [18:16] TJ-: oh forgot the -generic- prefix [18:16] TJ-: you're right, sorry [18:16] * tarzeau still looking for ppa testers (pkg lix), bionic: https://launchpad.net/~gagarin/+archive/ubuntu/bionic [18:17] tarzeau: that's not the purpose of this channel either :) [18:18] nacc: is there a specific place for ppa testing? [18:18] nacc, figured it out, turned out it was because some old program required the 1.15 version which i built from source and installed but was used as default... thanks again [18:18] tarzeau: what is the ppa for? a bugfix? [18:18] WoC: ack [18:19] Zteam: tomreyn "_GTF" apparently returns a bunch of ATA commands the drive requires to reset it, so those errors very much look like a possible cause [18:19] nacc: a package not yet in debian/ubuntu, test the game [18:19] TJ-, here is the log from completee strace of fstrim this time it didn't take long at all, for some reason [18:19] nacc: i just use it to figure out missing b-ds :) [18:20] nacc: prototype is an r-type remake, also nice. and fasttracker2 well ... [18:20] TJ-, http://paste.ubuntu.com/p/6kbx2C5Wf9/ [18:21] Zteam: there are 2 possible causes, and 1 possible fix. Causes: 1) the PC's firmware ACPI tables/code do NOT include the _GTF.DSSP method of 2) the method exists but is not enabled because the firmware is tuned to only enable it for recent Windows versions, not Linux. In the case of (2) then the acpi_osi workaround may work for you, which I've written about here: [18:21] http://iam.tj/prototype/enhancements/Windows-acpi_osi.html [18:22] Zteam: Your strace confirms it is the TRIM itself that is delaying things [18:22] Zteam: in that case 16 seconds not 7 minutes, fortunately [18:23] TJ-: good finding. i'm wondering whether the microcode update is too old. it's "revision 0xc6, date = 2018-04-17" on this "Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz (family: 0x6, model: 0x5e, stepping: 0x3)" [18:24] thanks for the dd help earlier folks, it was just me looking at gparted that messed me up :( [18:25] its just a 2GB usb stick, but it says there is an 8GB partition on there somehow.. probably this error is a clue [18:25] The driver descriptor says the physical block size is 2048 bytes, but Linux says it is 512 bytes. [18:26] tomreyn: Cannot imagine microcode would help; this is an ACPI method missing or masked off [18:26] TJ-, usally it takes much longer than that, there are some settings in BIOS which is about compatiblilty with older Windows versions or something like that, should I try to change that? [18:27] TJ-: ok, maykes sense. [18:27] Zteam: no, they won't solve this partcular issue. [18:27] Zteam: give the acpi_osi workaround a try [18:27] Zteam: did you upgrade from 18.04 to 18.10 or reinstall? i'm wondering whether you switched from klegacy bios to efi booting [18:27] Zteam: if that doesn't help we can grab the ACPI tables out of the firmware and I can analyse them to see if those nodes/methods really are missing [18:28] TJ-, will do, but since this never happenened with Ubuntu 18.04 should this be considered a bug? [18:28] Zteam: in which case it'd be a pretty severe firmware bug, because some other ACPI method is referencing them! [18:29] but also the Type recorded from fdisk is "Empty" [18:29] Zteam: it may be to do with the kernel being stricter on ACPI now, and trying to force more functions to be enabled [18:29] tomreyn, upgraded, and I haven't change any bios related settings, before that showed up [18:30] according to https://www.intel.com/content/dam/www/public/us/en/documents/sa00115-microcode-update-guidance.pdf (page 16) the microcode is current for this cpu. [18:31] Zteam: thanks for confirming [18:32] TJ-, okey, I will try that, thanks for all your help TJ [18:32] :) [18:35] Zteam: looking at some other system's ACPI DSDTs, they define DSSP on controller ports but so far not discovered precisely what 'DSSP' stands for although I suspect something like "D???? Solid State Power" [18:37] Zteam: aha, _GTF == Get Task File, DSSP appear to flags that control whether or not certain non-standard SATA controller features are enabled [18:38] Zteam: there's a firmware update available for your drive from kingston. your version: 506ABBF0, latest version: 60AABBF0. details at https://www.kingston.com/us/support/technical/downloads/90683 [18:39] Zteam: is that an HP PC by any chance? [18:39] (the readme lists 56 intermediate firmware revisions) [18:40] * 6, not 56 [18:41] Zteam: I'm reading in an HP system manual, for its BIOS, it has the option "BIOS Support for _GTF in ACPI Mode" with an explanation "BIOS supports _GTF (Get Task File) ACPI control method while in AHCI mode. However a Registry key is required to enable the port(s)" - this obviously targeted at ensuring it works for MS Windows. [18:45] Zteam: the ACPI spec says, for _GTF: "This optional object returns a buffer containing the ATA commands used to restore the drive to boot up defaults (that is, the state of the drive after POST). " [18:48] Zteam: I suspect it's a bug in the firmware of that PC because I notice it refers in your log to SATA ports 2 & 3 'PRT{2,3}' (\_SB.PCI0.SAT0.PRT2._GTF.DSSP) and I'd take a bet that it doesn't happen for a device connected to SATA Ports 0 or 1 (PRT{0,1}) [18:51] tarzeau: maybe ubuntu-users? not sure [18:51] tarzeau: presumably you are packaging it for a purpose, contact the group interested in the purpose, which is not hte support channel :) [18:54] So my laptop with bionic seems to randomly freeze entirely (no key/mouse response, cannot ssh in from another computer, etc), and when I have to force restart with the power button, there seems to be no hint of the cause in any of the logs (kern.log, syslog, x-session-errors, etc). I'm kind of at a loss here :/ Any suggestions? [18:54] (this only happens with my laptop - I have the same distro running on my desktop with no problems, and an 18.04 headless system running on another tower that doesn't have an issue) [18:55] SlidingHorn: which kernel version? I've been investigating that same issue with the v4.18 series of kernels [18:58] TJ-: 4.15.0-42-generic [18:58] Zteam: if you also wantto try the firmware update (the change log doesn't seem related but sounds like you'd still very much want to apply it) and dont have a windows system available, you could try this: https://gitlab.com/vinibali/kingston_fw_updater [19:01] SlidingHorn: possibly a different cause then; I've test 4.15 through 4.17 so far without being able to reproduce. With no log clues your only option is to try to narrow down which versions are affected. I'd recommend trying the v4.18 kernel from 18.10 ( which is available as a HWE 'backport' ) in 18.04 via bionic-proposed and the linux-image*hwe-18.04-edge packages [19:03] SlidingHorn: how often does this happen and what are you doing prior to the freezes? [19:06] TJ-: stupid question, will apt only pull upgrades from proposed if I specify it using -t? [19:06] <`mist_> hey guys, i've got a corrupt vmware disk with ubuntu 18 on it. When i check the partition table it's only showing 40gb of the lvm, it should be 150 [19:06] <`mist_> and i'm missing some valuble data unfortunately [19:07] What would be a handy method of testing the reliability of DVD drive/disc? I have some old DVD-RW discs and I am thinking whether I should move the data to other media. [19:10] Hello [19:10] Does anyone know how apport installs its Python exception hook? [19:12] stavros: cat /proc/sys/kernel/core_pattern [19:12] (if that's what you mean, not sure) [19:13] tomreyn, the crash handler installs a Python exception that intercepts crashes in Python programs, but I can't figure out exactly how [19:13] Ahh, I found something in the docs, thanks [19:37] <_Sym_> Hi, I'm using Ubuntu 18.04 and with an HP laptop and I just noticed that the airplane mode key does not work in the GUI, but it does work in TTY. Anyone know of a solution? Same as this: https://askubuntu.com/questions/1042333/aiplane-mode-key-not-working-in-gui-but-working-in-tty [19:40] Hello! Maybe simple question but what is the difference between these two versions of the same package? https://gist.github.com/jglenn9k/3a84ab7bb3207f2a8dafbe8b537ff69d [19:40] Was there a change in version naming? [19:42] donkdonk: -1build1 was a no-changes rebuild [19:42] donkdonk: -1ubuntu0.1 had some Ubuntu-specific changes that differed from the Debian synced version [19:43] http://changelogs.ubuntu.com/changelogs/pool/main/l/lxml/lxml_3.5.0-1ubuntu0.1/changelog [19:43] donkdonk: namely, -1ubuntu0.1 fixed CVE-2018-19787. See https://launchpad.net/ubuntu/+source/lxml for details and version comparisons and changelogs under "Xenial" [19:43] or what tomreyn gave you :P [19:44] or what clicking on "ubuntu Changelog" at https://packages.ubuntu.com/xenial/python-lxml gvies you [19:47] teward and tomreyn thanks a ton! [19:49] always happy when nvidia accepts help [20:00] SlidingHorn, does it have an intel gfx? some user in #centos added i915.enable_rc6=0 kernel boot parameter to fix a freeze [20:00] I am having an issue with a sys V init.d script that uses start-stop-daemon http://paste.debian.net/1055898/ [20:01] sometimes when I do a service restart - the init script stops the process, but does not start it again [20:02] how can I debug this [20:06] saraf: Have you ever considered translating that to a systemd unit, given that 16.04 and 18.04 run systemd? [20:06] I'm having a problem in a VM in proxmox with the lxcfs systemd unit. I'm unsure what to do with it. It suddenly says that there was an invalid argument to the fuse mount after a reboot [20:07] just says `failed to mount /var/lib/lxcfs` [20:07] fassl: yes...I could try that. What exactly is it? [20:09] iosecure: I have the following script in /etc/systemd/system/hubot.service - http://paste.debian.net/1055902/ [20:10] doing a systemctl daemon-reload ... seems to not do anything. I cannot do a sudo service hubot start [20:10] iosecure: which makes me think that systemd is not being used on this server, even though the directory exists. [20:12] okay - this server is still 14.04 [20:12] hm [20:14] SlidingHorn, something to do with power management if i read it correctly, not sure though [20:15] saraf: That would be why. Remember that 14.04 goes out of support in a few months. [20:16] hmmm ... yes ... need to upgrade :) [20:16] SlidingHorn, ill just paste you the links https://bugs.freedesktop.org/show_bug.cgi?id=105962 https://bugzilla.redhat.com/show_bug.cgi?id=1573482 https://bugs.centos.org/view.php?id=8630 [20:16] Freedesktop bug 105962 in DRM/Intel "[KBL] "enable_rc6" parameter deprecation brings back freezing" [Normal,New] [20:16] bugzilla.redhat.com bug 1573482 in xorg-x11-drv-intel "i915.enable_rc6=0 is not respected in new kernels" [High,Closed: eol] [20:17] fassl: is that supposed to be enclosed in quotes? Sorry, I don't deal with kenel boot parameters often/at all [20:18] i think the whole parameter line is enclosed in quotes, not parts of it [20:18] before adding it permanent, you could just edit the boot parameters for the next boot and see if it fixes the issue [20:18] *permanently [20:19] Hi, did someone know witch package provide Qt5Qwt6 on ubuntu 18.04 ? [20:19] fassl: it's a very random crash that doesn't happen on every use, so enabling it permanently would really be the only "true" test [20:19] thats true [20:22] caipiblack, does apt-cache search qwt give any hints? [20:23] !info libqwt-qt5-6 [20:23] libqwt-qt5-6 (source: qwt): Qt widgets library for technical applications (runtime, qt5). In component universe, is optional. Version 6.1.3-1 (bionic), package size 446 kB, installed size 1618 kB [20:24] Maybe it'"s not this one [20:24] Because cmake doesn't find the library [20:25] !info libqwt-qt5-6-dev [20:25] Package libqwt-qt5-6-dev does not exist in bionic [20:25] !info libqwt-qt5-dev [20:25] libqwt-qt5-dev (source: qwt): Qt widgets library for technical applications (development, qt5). In component universe, is optional. Version 6.1.3-1 (bionic), package size 35 kB, installed size 177 kB [20:28] if the package doesn't exists, what does it means ? [20:28] I will not be able to get it by apt-get ? [20:29] Should it be added in the future ? [20:29] caipiblack, what are you looking for ? [20:29] ioria -- No package 'Qt5Qwt6' found [20:30] caipiblack, libqwt-qt5-6 [20:30] caipiblack, or libqwt-qt5-dev ,butmif you don't tell what are you doing ... [20:30] I already have this lib [20:30] caipiblack, also the -dev pkg ? [20:31] I am trying to build gnu radio from sources on ubuntu 18.04 [20:31] ha [20:31] The two ones are installed* [20:32] caipiblack, you know it's in the repo, right ? [20:32] no ? [20:32] which repo ? [20:33] oh you speak about gnu radio ? [20:33] !info gnuradio [20:33] gnuradio (source: gnuradio): GNU Radio Software Radio Toolkit. In component universe, is optional. Version 3.7.11-10 (bionic), package size 7550 kB, installed size 81633 kB [20:33] Yeah I know [20:33] ok [20:33] But I need the last version with last libs, for get the LimeSDR working. [20:33] I have already installed this many times on another computers [20:34] But on this ubuntu 18.04 computer, I'm locked because of damn libraries. [20:34] They also don't have the support for 18.04 in the "automatic building script" [20:36] caipiblack, who's they ? [20:36] https://wiki.gnuradio.org/index.php/InstallingGRFromSource [20:37] http://www.sbrac.org/files/build-gnuradio this script didn't have ubuntu 18.04 support, So I can't check the libraries that are installed for 18.04 :D [20:37] wait unregged? [20:38] I already tryed the libraries of 16.XX but some of them doesn't exists on 18.04 [20:38] Now I have to go, i come back tomorrow, if i don't resolve the problem looking for the damn libraries [20:38] Thanks for your help [20:38] caipiblack: you've already been shown the libraries....more than once. [20:41] ah, it wants the python version of qt [20:44] i'm having trouble with installing bionic onto a hp6200 but the boot from either a dvd or thumbdrive hanges with a tpm error message - i disabled the tpm in the bios but it still hangs with a tpm error message - no other info found on a web search but i guess i can blacklist the module on the thumbdrive? [20:51] eelstrebor: could it be that it's an EFI Secure Boot and it needs access to the TPM? [20:53] hey guys, i just opened my laptop to clean out the dust, and found out the CD-ROM drive had been disconnected since the last time i had done that (which is why it wasn't working) but i'm not sure why it won't show up. I think it may have something to do with SecureBoot being on, but i'm kind of stupid when it comes to this. Anyone know something about how to get it identified? [20:53] TJ-, don't know. but i just tried modifying the bootup with nomodeset and got past the tpm error message [20:54] actually, i'm going to go check the bios thing, and try a few other things before i come back in here, if i'm back again that means it still isn't working lol [21:00] eelstrebor: maybe that is just the last thing you see, and is only a warning, not an error [21:08] i have a vmware vm with 17.10 server that i'm trying to upgrade to 18.04. i was able to do the upgrade and end up at a seemingly working 18.04 shell. however when i reboot the server it won't successfully boot, just drops to an initramfs prompt complaining about "mount: mounting /dev/mapper/vg1-root on /root failed: No such device [21:09] does anyone have any ideas for me to look into to try and fix this? [21:12] lazr47: you get a busybox 'initramfs' shell? is /dev/mapper/vg1-root there? if not does it show up doing "vgchange -ay" - if that command isn't available it would seem the initramfs-tools didn't correctly add the lvm2 scripts and tools [21:13] lazr47: it may be the virtual disk naming has changed and has caused an issue if the lvm.conf is excluding it, for example [21:14] pene === ubuntu is now known as Guest77907 [21:15] o/ Does anyone know which building system is used in ubuntu? [21:15] like fedora->koji === Guest77907 is now known as chuyd [21:16] Guest77907, for ubuntu/debian it is DPKG [21:18] Hey Guys, is there a good IDE for C++ on Bionic? [21:18] I tried Code::Blocks but its very old and unstable [21:18] chuyd: https://wiki.ubuntu.com/Launchpad [21:19] de-facto, tons of them [21:19] !ide [21:19] Programming editors/suites: Terminal-based: vi/vim, emacs - KDE: Kate, KDevelop, Quanta+, Umbrello - GNOME: gvim, gedit, anjuta, pida, monodevelop, geany - Others: eclipse, netbeans, qtcreator [21:19] OerHeks: DPKG is the building tool, but what i'm wondering is which building system is used to orchestrate the building for all the ubuntu packages [21:19] chuyd: launchpad [21:19] TJ-, it is an error and it hangs (i waited 5 minutes) unless i use nomodeset [21:20] heh, it turns out it isn't the pc that i had ordered anyway [21:20] chuyd, koji is exactly the same, why do you want to know and why did you not find such easy info? [21:20] * chuyd reading acheronuk ty! [21:23] I'm never imagined that launchpad would handle the package building, i was reading through doccumentation but i found the way to do it manually with dpkg-buildpackage [21:24] chuyd: well, it hands off to buildds hosted with (I think) openstack. [21:24] TJ-: yes it's a BusyBox 'initramfs' shell. /dev/mapper/vg1-root is indeed there according to 'ls /dev/mapper/vg1-root' 'vgchange -ay' gets 'WARNING: Failed to connect to lvmetad. Falling back to device scanning.' (has not returned to prompt after a couple of minutes) [21:24] build farm: https://launchpad.net/builders/ [21:25] lazr47: ignore lvmetad, that is expected - that process doesn't start until the real root file-system and init is started [21:26] acheronuk: thx for pointing me to the right direction [21:27] np [21:27] lazr47: if the node exists you need to check the file-system [21:27] lazr47: as in "fsck /dev/mapper/vg1-root" === daniel is now known as Guest17986 [21:30] hello [21:30] what was the name of a torrent downloading software that was already in Ubuntu right after installation? [21:30] TheWild: transmission [21:31] type torrent in search .. [21:31] Hey TJ- if I install that kernel from Proposed, is there anything I need to do to make sure that it's the only package taken from that repo? [21:31] hmm... :/ considering that I use i3 and thought I had mistyped a command [21:32] okay, thanks SlidingHorn. Transmission doesn't seem to be installed on my computer. [21:32] ktorrent, transmission, maybe some other tool, depending on the desktop [21:33] this is interesting, the pc does have a nvidia pci video card but it didn't show up on lspci or lshw [21:41] SlidingHorn: hmmm, I'm not sure. We do have apt pinning but I'm not sure if we can deploy it that way. [21:48] SlidingHorn: I suspect the apt_preferences "APT::Default-Release "bionic"; would do it combined with "apt-get install -t bionic-proposed " to fetch the package from the (-t)arget distribution 'bionic-proposed' - assuming that is listed in the apt sources.list - you may need to test this hypothesis though! [21:50] Hello, I have a question about a package. [21:51] Whoops, premature enter [21:51] Anyway, the package is 'iucode-tool', which exists in main as a package to install, however the sources are listed in restricted. [21:52] AFAIK, this shouldn't happen, the sources for any given package should exist in the same component as installation package. [21:52] Is this an error, an exception to the rule, or am I misunderstanding the apt ecosystem? [21:54] This specific occurrence is in Bionic, I haven't investigated if this situation exists in other releases [21:55] johnnyfive: it is due to it being an automatic import from Debian [21:58] TJ-, thanks for the response. What i'm reading from your statement is that this is intended (not a bug), so this situation could occur again. [21:59] and more specifically won't be fixed. Is there any other data that explains your statement further? They made this decision so it was copied here, or something else? [22:03] johnnyfive: I'm trying to figure it out; it looks like it was originally (pre 2014) in multiverse around the time of the Vivid release, and then moved through 'main' to restricted' at version 1.1-1 [22:04] TJ-, roger, thanks. [22:05] So.. Anyone here using an Acer Nitro 5? [22:05] johnnyfive: I wonder if in the 'main' > 'restricted' move someone forgot to update the binary pocket description. It's actually stored in the restricted pocket [22:05] having a horrible time to get the discreet graphics to be performant [22:06] 7? The one with the plastic hinge caps but otherwise solid [22:06] Ah, so it's supposed to be in restricted, that would make more sense [22:06] getting about half the performance as the ryzen 2500u integrated graphics with an rx 560x [22:06] Been googling but I've not found anyone post how to get the switchable graphics fps to work [22:07] yours is probably a lot newer. The casei s called that [22:07] you have to be plugged in to use power hungry GPU [22:08] TJ-: 'fsck /dev/mapper/vg1-root' reports that i should use btrfs check as it's a btrfs volume. 'btrfs check /dev/mapper/vg1-root' seems to report that things are fine, no errors found. fwiw i followed a guide at some point a couple weeks ago and was able to mount and browse the disks from an ubuntu live iso boot session [22:08] otherwise intergrated graphics have better performance [22:08] BrickedInTheHead, have. Yup. [22:08] lazr47: right, so this sounds like it may be an issue with btrfs detection in the initramfs [22:10] 2500u raven ridge: 1190 ; radeon 560x: 468 [22:10] on glmark2 [22:11] notbobdole, integraded might be better, DRI_PRIME=1 glmark2 - for dedicated https://askubuntu.com/questions/1091846/how-to-switch-between-the-graphics-on-acer-nitro-5-with-amd-ryzen-and-radeon-560 [22:12] integrated should be about 1/3rd to 1/5th the score of the 560 :/ [22:12] I dont suggest using doing that on ubuntu. You end up with horrible battery life and constant fiddling with configs =/ [22:12] johnnyfive: weird; it's actually in both main and restricted! see http://archive.ubuntu.com/ubuntu/pool/main/i/iucode-tool/ vs http://archive.ubuntu.com/ubuntu/pool/restricted/i/iucode-tool/ [22:13] https://unix.stackexchange.com/questions/470351/rx-560x-slower-than-integrated-vega-gpu-on-fedora-28 [22:14] Led me to here [22:14] BrickedInTheHead, borking battery life is a known issue, as the Dgpu cannot be switched off AFAIK [22:14] Seems that the discrete graphics are in development still but there is a git kernel patch you can use. [22:14] I gotta figure out how to swap to it though. [22:14] OerHeks, thanks [22:15] and look for the Amdgpu-PRO driver.. [22:15] TJ-, hah. That's even weirder. I'm in #debian and am starting a conversation there. Maybe there's a use-case for this behavior. [22:16] Used some tool to switch between gfx hardware, but required constant restarts. I would only consider that if necessary for work [22:23] johnnyfive: 'restricted' maybe due to it containing Intel microcode; 'main' because it needs to be included on the installation CDs [22:25] TJ-, Hrm, so maybe intentional. Thanks for all the investigation. I'm going to dig further until i'm 100% sure this isn't just a bug. It is causing all sorts of headaches in our build pipeline. [22:29] johnnyfive: at a time when the devs are awake, I'd ask in #ubuntu-devel. You may need to wait until next week now [22:31] That's fine, not urgent. Thanks again for your help TJ- === coffeeguy is now known as zenguy [22:38] i think that at least since meltdown / spectre, the kernel packages depend on intel-microcode, and either directly or indirectly on iucode-tool. so at least since then one would want to make sure that any form of distribution which provides a kernel package also provides the others. not sure this explains it, though. [22:38] tomreyn: I wondered that, but it seems like it has been like this since 2014 [22:39] this is a lot earlier indeed. [22:39] tomreyn: weirdest is having it in 2 different distro sections simultaneously [22:39] Client: HexChat 2.14.1 • OS: Ubuntu "bionic" 18.04 • Memory: Physical: 5.0 GiB Total (1.9 GiB Free) Swap: 2.4 GiB Total (1.4 GiB Free) • Storage: 66.7 GB / 174.3 GB (107.7 GB Free) • Uptime: 21h 22m 35s [22:39] tomreyn: Just when I thought I finally understood archive organisation (only taken 14 years!) [22:40] IIRC the intel microcode update mechanism somehow broke around that time and was repaired a while later. [22:40] * tomreyn doesn't claim to understand it yet ;) [22:42] at some (recent) point i thought main would only contain GPL'd software, so filed a bug and then learnt i had misunderstood. [22:43] s/software/packages/ === capella is now known as capella|away [22:48] hi! I upgraded from 16.04 to 18.04 and I'm running into some issues with sshd. i went through and commented out the deprecated options and the service isn't logging those errors on startup anymore, but i'm still having trouble with agent forwarding. whenever i log in i get this error: `ln: failed to access '/home/infogulch/.ssh/ssh_auth_sock': Permission denied` [22:50] which looks like: lrwxrwxrwx 1 root root 33 Dec 6 20:19 ssh_auth_sock -> /tmp/ssh-[stuff]/agent.22300 [22:51] any ideas? [22:53] infogulch: Are you trying to login as root? [22:53] nope [22:54] logging in as my normal user [22:54] i see that now.. .missed the user name in the path [22:54] i don't know, but is it weird that that link is owned by root? [22:56] infogulch: I would restore your back up of the config file and look at the difference... Just because something is deprecated, doesn't mean it's a security vulnerability. what you were seeing were warnings, merely there to inform you that a feature may be removed in the future. [22:57] Something that you removed is relied upon by Ubuntu's default configuration of openssh [22:57] key based auth to ssh [22:57] well i had this issue first, and when i went digging around in the logs for an explanation i saw those warnings so i fixed them, but this key forwarding issue is still here [22:58] so i don't think the warnings are related, not directly at least [22:58] because it had issues before and after fixing them [22:58] infogulch: Again, something that you commented out may have resulted in something else breaking. It is entirely possible. You made a massive number of changes and it this point, it is best to go back to square one and start one-by-one making changes and noting what did and did not change. [22:59] ok [23:24] hi! what's the best way in bionic to manage gnome-terminal settings from the command line/in a shell script? [23:25] I found https://askubuntu.com/questions/967517/backup-gnome-terminal but my dconf setup seems to no longer work and I read in another post it's now gconf again [23:28] oh! seems like dconf does work and I just had a misspelling. [23:31] hi [23:31] does anyone know why, when I create a new profile with the gnome-terminal that shipped on the bionic live ISO, those new profile's settings are saved in "legacy/profiles:…"? shouldn't "legacy" be just for old stuff? [23:32] I can't install ubuntu on asus rog ? many errors. [23:33] UEFI/Legacy mode? blackbird1 [23:33] blackbird1, try to get help with each of the errors one at a time. start with the first one. [23:34] this is the 1st one: [23:34] Couldn't get size: 0x800000000000000e [23:34] MODSIGN: Couldn't get UEFI db list [23:34] Couldn't get size: 0x800000000000000e [23:35] blackbird1, what kind of device did you boot from? [23:36] or rather, try to boot from [23:36] Sven_vB, usb derive [23:36] which version of Ubuntu did you put on that USB drive, and which method did you use to put it there? [23:38] Sven_vB, 18.04 LTS [23:39] Sven_vB, I used Rufus to extract the iso in the usb [23:40] blackbird1, did you follow a tutorial on how to use Rufus? could you post the link? [23:43] I just remembered I had a similar problem with an Acer AOD once. it's probably not the best solution but worked for me: to just rename the "EFI" directory on the USB drive to a bogus name so the BIOS would use the old boot mechanism. [23:44] if it doesn't work you can just rename it back to EFI. [23:46] I am having a really annoying issue, I am trying to burn some WAV files that I had previously ripped some months back. As soon as I click on "Burn" in either k3b or brasero (tried on two separate computers - one running Ubuntu 18.10 and the other Ubuntu MATE 18.04) it immediately fails and ejects the disc? [23:47] I have tried playing the files and had previously successfully burned many discs with these particular files, they play perfectly in whatever media player you point them at [23:48] And I have used multiple different CDR media [23:52] Paddy_NI, maybe the WAV has some strange header that brasero can't understand. you could try importing it into audacity and then export it again as WAV. [23:53] Sven_vB, That sounds interesting, worth a shot I suppose :-) [23:53] Cheers