[00:15] dtux: install ubuntu and find out [00:27] :( ok... sorry to bother you all === dominic35 is now known as dominic34 [00:58] so, the live server installer is being fun. it shows free disk space to create partitions, but can't select it to do so. luckily it has a vt terminal that can create partitions, but can't scan for the changes, so one reboot later, I also found that it can't select empty partitions and instead crashes immediately when the menu passes over them [00:58] this normal stuff? [01:00] Kevin: I ran through the live server installer a dozen times before 20.04 was launched and had no trouble creating partitions [01:01] free space is greyed out and skipped over, and I went through the entire vertical menus until it wrapped [01:02] let me see if I can record it [01:11] sarnold: https://www.youtube.com/watch?v=5bDREcrQbc4 [01:12] Kevin: zounds that was quick [01:12] nice bmc is nice [01:13] *ouch* [01:13] Kevin: There's legacy install media too. [01:14] Kevin: I've used the legacy server install media, and I've installed from the desktop live media using the installer, and using debootstrap. [01:14] I've not grown fond of the new text installer. [01:15] Kevin: If you're installing that way, try something from here maybe: http://cdimage.ubuntu.com/ubuntu-legacy-server/releases/20.04/release/ [01:17] hmm I can't select through the free space now, but maybe someone knows where the option *would* be if it's somewhere on the screen [01:19] sarnold: https://youtu.be/bNMkPYJQRuA unexpected result with free space [01:19] notice how it's greyed out though. previously the menu worked but would skip over it [01:19] mason: will try :) [01:20] Kevin: It's basically the traditional Debian-style installer. [01:22] would be kind of nice if it ran wifi btw, even though that's strange. installing on a desk with one ethernet cable that's being used for the kvm [01:24] Kevin: You should be able to set up wifi for the installer. [01:25] will be nice if so. the live one didn't.. or it ran out of space in the list and it was below the ethernets [01:26] I didn't think to test manually whether a driver was loaded [01:26] but it's an ancient realtek usb thing that would have drivers [01:28] indeed the legacy one did do wifi === halvors1 is now known as halvors [01:31] \o/ [01:32] Kevin: Remember: new things, bad. Old things, good. Except that 20.04 is nice. It gets a pass. [01:32] fingers crossed on that one, heh [01:32] I like it so far. I've been running it on various systems since beta. === zbenjamin_ is now known as zbenjamin [01:40] well, the legacy installer managed to install an instance of linux, win for still existing [01:40] Kevin: good good [01:41] Kevin: xv still builds Just Fine on 20.04, incidentally. [01:45] speaking of which, is there a really good reason netplan exists? yaml seems to be designed for computer editing [01:45] also the installer forgot to escape characters in the wifi key and broke it =p [01:46] Hi! all, so who owns ubuntu now ? [01:46] I have a complaint. Same problem has been in ubuntu since version 12 and hundreds have complained an no fix [01:47] escape characters in the wifi key .. seen better trolling than that [01:47] well it put it there without any quotes or escapes [01:47] transhumanist: can you link your bug report? [01:48] dunno what the proper format is because yaml [01:49] in order to log in through xrdp or remote desktop from windows, one must reboot ubuntu and not log in for remote desktop from windows to work, in other words there can be only one login session , other solution is to install xfce instead which doesnt fix it the right way either ( you can't pick up the same session and have it autolog out if necessary from the local console [01:50] not sure what bug report applies but I am sure with the thousands of complaints someone must have filed a bug report by now [01:51] I will personally pay buttersworth or whoever ownes ubuntu now $100 in bitcoin to fix the issue, I am not rich so cant afford more, its rediculous [01:52] perhaps they should take up a collection with all the pissed off people to try and get the money together to get it fixed === halvors1 is now known as halvors [01:53] this is by design? [01:53] https://websiteforstudents.com/how-to-connect-via-remote-desktop-rdp-to-ubuntu-20-04-18-04/ [01:53] its by design that you have to reboot a remote computer in order to use it a second time , its by design that it doesnt just log out the local session or even better take over the session? [01:54] not so sure about that, a few years back it ran multiple sessions [01:54] it used to work about 10 years ago I am sure of that [01:54] I haven't need to use it lately so haven't followed changes/breakings [01:54] transhumanist: can you link your bug report? [01:54] it connects and you get a blank black screen [01:54] tell me where to go and I will file one [01:55] transhumanist: you've been having an issue since version 12 and never bothered reporting it? [01:55] "ubuntu-bug xrdp' or whatever package you're using [01:55] transhumanist: and you're just guessing that someone else did? what if they all thought the same as you? [01:55] no because hundreds of others have had the problem, I assume by now someone has filed a bug report, or someone reads the ubuntu blogs [01:55] perhaps noone reads the blogs, ok [01:55] transhumanist: maybe everyone else thought "I assume by now someone has filed a bug report", and hence no one did. [01:56] ok I will file a bug report [01:56] transhumanist: it's easy to check whether someone has or not. https://bugs.launchpad.net/ubuntu [01:56] then you can also see whether there are any workarounds or whether it's being fixed [01:56] just use the search at the left top. [01:56] indeed, there are some .. https://bugs.launchpad.net/ubuntu/+source/xrdp [01:59] also a similar problem with samba services, in order to get samba to work you have to spend about 2 days compiling the latest version of samba for shares to work with windows 10, the common solution is to degrade the security settings on windows to an abysmal level, the real solution is to compile a later version of samba [01:59] is lm-sensors still the standard tool for hardware monitoring? [01:59] the compile of samba takes for ever and many many bugs to overcome since there is no late version package [02:00] Kevin: I think so; if you find better, it'd be nice to know :) [02:00] Kevin: mo9st of the grossest stuff from lm-sensors isn't needed any more of course.. [02:01] this makes ubuntu look inferior when you try and integrate it on a windows network. and it is truely better than that. [02:01] prevents adoption in companies, another solution is I think in enterprise to used federated services or something I think [02:03] oh, it is just an continues rant.. [02:04] well sorry for ranting, but the fact is from what was pointed to me people in fact have complained about it and its not fixed. [02:04] If I had the money I would hire someone to fix it [02:05] end of rant... sorry [02:06] transhumanist: step one, report the bug. [02:06] not supposed to rereport an existing bug [02:07] ah, what bug number? [02:07] you found an existing matching bug? [02:07] transhumanist: okay, can you link the existing bug? [02:08] hold on looking for an exact match [02:09] transhumanist: I thought you already determined there was an existing bug [02:10] there is a lot of similar bugs, but it seems that it draws certain parts for some people rather than just show a blank screen [02:10] fix is to reboot, not log in and then remote connect [02:10] transhumanist: so there is no existing bug report? [02:11] transhumanist: ah sorry, I meant the samba issue -- afaik we already raised the minumim protocol level there, and if that's been reverted by accident, I'd like to know about it [02:12] https://bugs.launchpad.net/ubuntu/+source/xrdp/+bug/753733 << this is close to the xrdp issue [02:12] Launchpad bug 753733 in xrdp (Ubuntu) "xrdp installed from the Natty Beta1 libraries only gives a black screen." [Undecided,Invalid] [02:12] from 11.04 [02:12] sarnold what version was it raised to ? [02:13] I think its samba 3.11 or something that is the minimum for windows maybe its 4 or 5 [02:13] I dont remember [02:14] transhumanist: and that one was marked invalid in 2015 because it was reported for an EOL version [02:15] yes I realize that Its been a similar issue in all versions, I just finally blew up, sorry, should have handled it better [02:15] transhumanist: I'm just saying, if you can't find an open bug report about it, it's never gonna get fixed. [02:17] I will look into this and get back to you...I promise and file a report [02:17] for both [02:17] got to run now though [02:17] no worries [02:33] hi I have a sound question can someone help? [02:33] and change the internal mic to headset mic in order to listen to sound [02:33] I currently had to fiddle around with the settings [02:34] is there such thing in budgie? [02:34] on Unity there was a dialogue box asking me when ever I plugged a set of headphones or headset [03:32] Hey all, I'm running an Ubuntu VM in VMWare for a job and I have it set up with two interfaces - one bridged for internet access and one host-only for low-latency inter-VM communication. On Ubuntu, the host-only iface is setup with a static IP and is given a default route in the routing table, which I think is messing up my internet connection as of late. Is there a way to prevent a default route [03:32] from being assigned for this interface? [03:33] a small disclaimer, I don't know that much about networking :P but on my Arch VM (where no default route is generated for the host-only iface) the internet connection is working, so I suspect that's the source of my woes [03:36] Messing up your connection how^ [03:36] I cannot connect to any internet site at all. Strangely, dns resolution still works [03:36] You can ping google.com? [03:37] The name resolves but no packets go through [03:37] I can do so on my Arch VM however [03:38] what does "ip address" output? [03:39] Also, did you setup some exotic things, like dnsmasq, etc? [03:39] https://pastebin.com/9fYRam4c [03:40] I have not set up any nonstandard network stuff in Ubuntu besides setting up the static IP in /etc/network/interfaces [03:41] from the paste, inet addresses are configured as expected. The host-only iface is ens38 [03:41] on which version are you jadelclemens? as ubuntu doesn't use ifupdown anymore [03:41] I'm on 18.04.4, required by my job [03:41] For compatibility with Yocto I think [03:42] Sorry, should've specified [03:43] jadelclemens, as I am not 100% familiar with ubuntu, but with 18.04 they are using cloud-init for setting up networks, maybe this works https://www.serverlab.ca/tutorials/linux/administration-linux/how-to-configure-network-settings-in-ubuntu-18-04-bionic-beaver/ [03:43] wish that URI could be longer [03:44] I'll take a look, thanks quadrathoch2 [03:45] rather this than a url shortener matsaman :p [03:46] https://serverlab.ca/tutorial/linux/admin/ubuntu/18.04/network/ [03:47] although who knows if they even provide things other than tutorials, for OSes other than Linux, for purposes other than administration, for distros other than Ubuntu =P [03:47] it is shorter, but still different content which is not helpful :> [03:48] My suspicions are confirmed by the fact that deleting the defualt route for the host-only IP restores internet connectivity. I'll keep digging. [03:55] Do I need to configure anything to use netplan rules (in /etc/netplan)? [03:56] Not that I know of jadelclemens [03:57] jadelclemens: See: https://blog.ubuntu.com/2017/12/01/ubuntu-bionic-netplan [04:12] Well the route is still there but I got it generating my bridge route first. So problem sorta solved? [04:15] Nevermind, fixed it. Thanks for the help all [04:18] can i install ubuntu for 2.3GHz Intel core i3-6100u dual core? [04:19] futureRich: why wouldn't you be able to [04:19] i3 is too slow for ubuntu os? [04:20] futureRich: That old of hardware - the question is how much ram is onboard ? ubuntu wants 4 gigs for a good experience. There is lubuntu/xubuntu for the less endowed :P [04:21] intel i3 isn't slow or fast. i3-6100u is rather fast. [04:21] 8GB [04:21] futureRich: install away :D [04:21] thanks :) [04:22] I realize 2.3GHz of today isn't the same as 2.5GHz from 2015, but 2.3GHz is still fundamentally 2.3GHz [04:22] as 2.3GHz* from 2015, even [04:23] :) === isolier3 is now known as isolier === amosbird_ is now known as amosbird [06:31] Since the last update my Ubuntu 18.04 can't suspend anymore. The Screen goes black and then comes right back to the lock screen. It worked fine before. Does anybody know what i can do to fix it? [06:34] kab0m: this might help http://iam.tj/prototype/enhancements/Windows-acpi_osi.html [06:37] hi [06:37] How do I enable SSH auth? - The old tricks like `PasswordAuthentication yes` aren't working on Ubuntu 20.04: https://askubuntu.com/q/1249507 [06:40] AlecTaylor: is it root you're trying to use? [06:40] no [06:40] They're not even in sudo [06:41] AlecTaylor: you have to enable them in /etc/ssh/sshd_config not ssh_config [06:41] sshd is for the daemon, ssh is for the client-side [06:41] I did [06:41] ohhh [06:41] yeah [06:42] bit silly to have that disabled by default [06:42] Thanks [06:43] no problem [06:59] EriC^^ That sadly did not fix the problem. It goes to black screen for two times and then straight back to the lockscreen. Any other ideas? i mean it worked before, i dont know what has changed. [07:01] kab0m: Please pastebin the output of "journalctl --boot 0" after a failed suspend. (That will give us all logs, for all services, since you booted the machine). [07:02] kab0m, I think that's a known problem. Can you use hibernate instead? [07:12] Jordan_U Ok, here is the requested log https://bin.disroot.org/?ab661def697dd2a9#221UntjronhPCQYcoQBtUEdLmXbznaaNxJcH7y5gGJ46 [07:12] DragonRiver No, sadly hibernate never worked. But suspend to ram did work nicely before. [07:27] Hi, i'm trying to set up a server & i'm pretty sure my ISP is blocking most common port. Could anyone tell me if i can find out which ports they aren't blocking? [07:28] viktor_, how familiar are you with setting up a server? Did you enable port forwarding on your router? [07:32] DragonRiver: I'm new, but i did succesfully set it up with forwarding & DDNS. I was able to SSh into my server from outside after support of the ISP "fixed something". But now when i check from outside with e.g. ping.eu, my port can't be found anymore. So i think the it got automatically blocked again. [07:34] DragonRiver: so to avoid having to call tech support every couple of days, i'd like to find out if there are other ports they don't block. If that's possible? [07:42] viktor_, I don't know your location or ISP policy. My ISP doesn't block any ports. It's usually your firewall or router that will block ports. Port 80 and 8080 should always be open usually. [07:43] You could also try 21, 25, 110. Telnet is a way to check. As how to bruteforce, you will need to read up on kali linux. I'm not familiar with those tools. [07:46] DragonRiver: I've tried it with firewalls off & the port fwd worked before and is pretty straightforward. Maybe I'll google how to do that with kali. But no idea how to do that with my ISP. [07:47] Ok ive came across the problem i guess: Device 0000:03:00.0 failed to suspend async: error -110 [07:48] This device is a "03:00.0 USB controller: ASMedia Technology Inc. ASM1142 USB 3.1 Host Controller" [07:48] How can stop this usb controller from bugging my suspend? [07:56] If i lost my root password and don't have an sudoers user, i have to physically go to the pc and reset the root password, correct? [07:57] ok nevermind i found the problem and solved it :) [08:00] IsntFunny, yeah. You'll probably have to chroot into you physical pc from a rescue disk also. [08:00] I think i messed up my sshd config damn [08:02] <[Pokey]> IsntFunny: If you have physical access to the machine and can make a rescue disk thertes several online tools which can crack a shadow file and get the password back in most cases [08:02] <[Pokey]> I've had to do it for an old server box before [08:02] i think i would just have to edit my sshd file.. i added "AllowUsers" but only added my gameserver user, not root as a server [08:03] not sure if this is my problem though [08:03] <[Pokey]> You might need to set PermitRootLogin [08:03] IsntFunny, but if you know your root password, you can just su instead of sudo. But if you don't, no need to crack anything. If you have physical access, then just chroot and then passwd, everything will be transparent. [08:03] <[Pokey]> it defaults to prohibit-password [08:04] If i have permitrootlogin, do i also need to add root to "AllowUsers"? [08:04] <[Pokey]> Never used AllowUsers, I just allow all users to sign in and use a large SSH key [08:04] DragonRiver, omg thanks for the tip! that worked [08:04] Never permit root login, that's a huge security risk [08:04] it's just my little local livingroom server [08:05] People forget su, but it just works (if you have know the root password) [08:05] <[Pokey]> DragonRiver: Many decade old argument that one. If you know what you're doing and secure it right theres little problem. I sudo su most the time if I am a sudosuer anyway [08:05] yay it's fixed using su! thanks [08:06] [Pokey], well, I am several decades old, and so are my arguments ;) [08:06] <[Pokey]> ;3 Opinions be opinions [08:11] IsntFunny, when you're done being su, remember to add your user to wheel (or sudoer, whichever you're using), add the permissions with visudo, and exit the su state. === mateen1 is now known as mateen === mateen1 is now known as mateen [08:41] I am trying to bind the USB controller with its ID to the vfio-pci-module instead of xhci_hcd. I added its if to the file /etc/modprobe.d/vfio.conf but it does not work. The system still bind it to xhci_hcd at boot...what have i missed? [08:46] kab0m, what's your output of "dmesg | grep -i -e DMAR -e IOMMU" [08:47] You have to pass the iommu group over, not the controller ID per say. [08:49] DragonRiver Ohh ahm, i gave the wrong info to you sorry. I did indeed pass the iommu group-id to /etc/modprobe.d/vfio.conf. [08:50] DragonRiver here is the output https://bin.disroot.org/?4df0b6b292b83c64#4EAVNR5zqVSgiG8fziAQCVZ4AhvGuijSM2QoWuePMo8e [08:51] DragonRiver my device is in iommu group 11. and i added its id to /etc/modprobe.d/vfio.conf and rebooted. [08:52] Could it be the case that xhci_hcd is picking up the device before it gets picked up by vfio-pci-module? [08:53] It might be. Do you have the module "vfio_pci vfio vfio_iommu_type1 vfio_virqfd" in mkinit? [08:57] DragonRiver what is the path to the mkinit file on ubuntu? [08:58] kab0m, I don't have a ubuntu instance readily available, but I would assume it's /etc/mkinicpio.conf [09:09] DragonRiver hmm, google tells me there is no mkinit on ubuntu, but initramfs. [09:11] DragonRiver lspci tells me that the device is still using the kernel driver instead of the vfio-pci driver :( "Kernel driver in use: xhci_hcd" [09:11] *it is still using the xhci_hcd driver instead of the vfio-pci [09:12] Ok kab0m, that's very possible. Concept is the same. You need to add the kernel parameters. initramfs is virtually the same as mkinitcpio (mkinit generates a initramfs). The vfio needs to be loaded early in the kernel to isolate it. [09:13] If I had access to an instance right now I would test the thing for you, but I just can't at the moment. vfio is a tricky dog, you'll need to read up, and some headaches are going to be involved. [09:14] DragonRiver ok thanks ;) is it a bad idea to just blacklist the xhci_hcd driver completely? [09:14] It's also dependent on your hardware, but your dmesg output shows me most things are up for the framework. [09:15] I'd say try it. It's not a bad idea if it works, but again, no garantees. [09:15] DragonRiver is the xhci_hcd driver need for my usb 3.0 ports? I just want to get rid of the usb 3.1 controller. [09:15] *needed [09:17] From what I can see, yes, and it's needed for all USB devices. I'd search for a usecase similar to yours. [09:17] Which, btw, why are you trying to isolate the USB driver and pass in to vfio? [09:18] You can pass USB devices to qemu instances on demand, don't really need vfio for that. IME, vfio is for pcie cards, when you want to pass a device that doesn't initialize easilly. [09:20] DragonRiver No i just try to isolate the usb 3.1 controller from my ubunu system as its cause my suspend to fail. i would maybe use it in a windows VM but atm i just want to isolate it from my host system. [09:22] im running ubuntu inside virtualbox, i have a shared folder but i can't get ownership of it. the owner stays root and group vboxsf .. how can i change ownership ? i already tried chown command [09:25] kab0m, is there a unique bus and device id when you run lsusb for your usb 3.1 controller? because if so, you could disable it with udev rules. That said, I cannot think of a "good" solution. It's much less than perfect. [09:26] Also, I think maybe your USB 3.1 interface has direct lanes to the CPU? I've read about that before. That would mean it's using some voodoo that's beyond my knowledge. [09:29] ph88, did you try this? https://askubuntu.com/questions/747974/virtualbox-shared-folders-are-owned-by-root-in-lubuntu-guest [09:30] yes i tried .. atm i'm looking at this https://superuser.com/a/640028 [09:30] ph88, or just "sudo adduser ph88 vboxsf" [09:34] ph88, do you have the guest utils (or equivalent) installed? If all else fails, you can share a windows folder with samba and mount that network style to your VM (not sure how VB handles networks, maybe you won't be able to see the host from the VM's network). [09:35] i have the guest additions installed yes [09:37] what does "rw dropbox" mean in https://superuser.com/a/640028 ?? [09:39] rw just means read-write [09:41] what does dropbox mean though ? [09:41] -- read-write of what ? === alazred_ is now known as alazred [09:46] maybe similar question what is "WebApps" in https://superuser.com/a/640028 ... i tried to adjust these examples to my situation but i can't get it right [10:19] hey, is it possible to not set any password for a user and have them set it themselves the first time they log in (with an ssh key) [11:07] How do packages get updated in ubuntu generally? As an example, prometheus node exporter 1.0.0 was released recently, but ubuntu 18.04 still has 0.18. Can I expect this to be updated in ubuntu? Is there a place where I can track it? [11:08] command ‘htop’ not found, but can be installed with ? what’s wrong? [11:10] I see - the answer may be "no you cannot expect it to be updated" since it contains some breaking changes and there's this: https://wiki.ubuntu.com/StableReleaseUpdates [11:10] Would the same go for 20.04 since that's also a LTS release? [11:11] What's the best strategy to get the updated verion on my machine? Use a package provided by upstream? [11:11] futureRich: I don't understand your question. Is that an error message that you copied from your terminal? What did you expect to see instead? [11:12] yeah why i can’t execute ‘htop'? [11:12] You may need to install it first. [11:13] `sudo apt install htop` should work [11:18] thanks codehotter [11:28] why i can’t copy file in usb to Desktop using mouse? [11:50] codehotter: its reccomended to use the package versions for your specific ubuntu release [11:51] codehotter: if you want/need other versions try installing another ubuntu release or try !backports !ppa or snaps [11:53] ubutn vs mac, which is more useful to use for noobs? [11:53] futureRich: join #ubuntu-offtopic for that [11:55] ok [12:05] why can I not add this user to a group? https://bpa.st/3JRA I tried as root too. [12:08] Sup, I am still struggling with this after a few days. On macOS fswatch --exclude ".*" can be toggled to include a few files via --include "xyz". This doesn't seem to be the case in Ubuntu. The --exclude ".*" option can't be countered with an --include "xyz"; it just excludes everything. [12:09] zmagii: maybe the placement of the option matters? [12:09] EriC^^: That was the first thing I tried :'( [12:10] hi friends ive got btnx and btnx-config installed and its pretty cool but somehting is strange - one of my mouse buttons always triggers browser "back" function even when i havent configured anyting for it [12:11] It seems like such a basic thing and yet I have no idea how to proceed other than doing compilicated regexes. [12:11] so when i do configure a command for tha tbutton, that command executes *and also* browser back [12:11] is there any lower level mouse config i can edit? some other program is capturing mouse events and delivering the browser shortcut outside of btnx [12:12] hello. Can someone please help me to speed up my router bitrate? I've tried various commands like iwlist and iwconfig to set it to the 'officially' supported value (150M) to no avail. It is stuck on 1M currently. [12:12] zmagii: did you try the man page? [12:13] also iwlist bitrate says "unknown bit-rate information." [12:14] EriC^^: I did read about the differences in the man page between macOS and Ubuntu. The main point seems that fswatch is actually a macOS thing in the sense that it's the only operating system for which they say it works with all features. [12:14] What I could surmise is that Ubuntu uses inotify and macOS has their own native monitoring system. [12:15] The one limitation that they mention is that on Ubuntu it doesn't listen recursively in folders. But this doesn't seem to be the case, many of my files are not in subfolders. [12:15] zmagii: the manpage makes it sound like if you put --include first it should work *shrug* http://manpages.ubuntu.com/manpages/bionic/man7/fswatch.7.html#filtering%20paths [12:17] Basically, I am using a repo that was written using macOS, and I just want the same functionality, namely, automatically reloading source scripts if they are changed. This is so that I don't have to stop and restart on every change. [12:18] "... can specify multiple filter expression (sic) in any order and the first mateching expression wins." [12:19] So, I agree that --include should then be first... [12:19] zmagii: yeah, maybe try it on one dir with 1 file, maybe it's a case sensitivity issue? [12:28] EriC^^: Yup, still doesn't work. I used simply 'fswatch --include "test.txt" --exclude ".*"' and then did 'echo "test1" > test.txt' [12:28] Sorry, I meant: [12:28] 'fswatch --include "test.txt" [12:28] Butterfingers. [12:29] 'fswatch --include "test.txt" --exclude ".*" ./' [12:29] But if I take out the --exclude then it works and prints "/path/test.txt" [12:30] 'Morning all [12:36] hello. I have some issues with my wifi connection. The most important one is that it disconnects and I can't even access the gateway even though Network Manager says it's connected. [12:38] can you access your router? [12:38] zmagii: maybe try with full paths? sorry cant think of anything else [12:39] TeleGhost, no [12:39] zmagii: fswatch --include "test.txt" --exclude ".*" /path/to/dir [12:39] my router is my gateway [12:40] zmagii: maybe this thread helps https://stackoverflow.com/questions/34713278/fswatch-to-watch-only-a-certain-file-extension [12:46] hello, what's atig used for? === beaver is now known as beaver|SYN [13:01] /24 [13:21] I'm currently doing full image backups for ~50 laptops, and I'd like to transition to a more time- and storage-efficient backup solution. In this case, all that's really needed is to restore the laptop's applications and configuration; Users are instructed store data on a file server. I'm thinking something like enumerating a list of installed packages and rsync the /etc and $HOME/.config directories. [13:21] I know this is highly dependent on user workflow, but are there other directories I should consider? [13:22] g3poandlsl doesn't it make more sense to just backup the user configs, and use something like ansible to configure the rest, so if the OS breaks, it's just a install away to get back the system? [13:23] g3poandlsl: you should look into a deployment orchestration system like ansible [13:23] quadrathoch2, That's the idea here. I'm using salt instead of ansible [13:25] g3poandlsl: you shouldn't just dump all of /etc or .config. You should decide on a set of applications and configs and only worry about reinstalling and restoring their configs. You'll never be able to catch everything. Even by restoring all of etc and ~/.config, which I don't recommend [13:26] EriC^^: It seems even that those examples don't work on Ubuntu. [13:26] I am on Xubuntu 20.04. [13:27] fswatch 1.14.0 [13:28] According to this thread, fswatch was only added to Ubuntu from 18.04+: https://github.com/emcrisostomo/fswatch/issues/170 [13:29] leftyfb, Thanks. The problem here is that we have multiple projects with different application and configuration requirements. Even within those projects, users may need a different set of applications and config than other team members. [13:29] g3poandlsl: so work with those teams to generate the list of applications and their configs [13:30] There's a bit of overhead for our team to identify each application and corresponding configs, which I'm trying to avoid due to manpower constraints === eTexNerd is now known as plarkinjr [13:36] hello, sometimes after clocking on the activities overview button on the top left corner, my laptop freezes. i have to use "ctrl+alt+f5" to restart the laptop to make it funtion again. [13:36] why is this happening? [13:37] sorry it was clicking* [13:37] and function** [13:45] All:No Answer YDay trying again: Tried to upgrade to Kubutnut 18.04 LTS on my file server. Live DVD install seemed OK but reboot stalled at the "KUBUNYU" screen. left it on over night to make sure and still stuch there this morning. What patches are out there as obviously this Live DVD was not tested and errors? [13:45] hello my webcam doesnt work [13:46] All: Wrote this up at: https://askubuntu.com/questions/1249313/live-dvd-problems [13:55] EriC^^: Maybe it doesn't work because I am using a VM [13:56] I don't do any development outside of VMs for the most part, so I tend to forget that it could cause problems. These days Xubuntu on a VM is so error free that I even forget I am in a VM if I am... [13:59] How can I check which key-binding is attached to which application\process? [14:05] Hello everyone [14:06] * [Pokey] waves [14:07] I want to install icecast with ssl support. I came to know(but not very sure) that the package in ubuntu is not compiled with ssl support as debian thinks openssl's apache license is not compatible with gpl2 [14:07] <[Pokey]> Ahaha [14:07] but it seems the icecast maintainers are themselfs providing compiled binaries via ppa [14:07] [Pokey]: hi [14:07] <[Pokey]> One sec murthy [14:08] <[Pokey]> let me pastebin a simple script I built fo build for SSL [14:08] So is it legal to use their package of personal or commercial use [14:08] [Pokey]: meaning openssl? [14:08] will the cloud-init on the ubuntu server ova image I'm using with virtualbox, resize root partition? If so, how best to get this going? or is that a cloud-init question? [14:08] <[Pokey]> murthy: You want IceCast2 with SSL Certificate support, correct? [14:08] [Pokey]: ya [14:08] <[Pokey]> alrighty [14:08] [Pokey]: but I can compile [14:09] [Pokey]: I want to know if its legal [14:09] [Pokey]: I want to install on an amazon aws server instance [14:09] <[Pokey]> Yea yea, you can, they just most likely don't allow it to be included direct. For your hosting services legality please refer to your T&Cs [14:10] <[Pokey]> I'm running an ubuntu backed dradio at the moment with SSL IceCast2 [14:10] [Pokey]: thats nice [14:11] <[Pokey]> murthy: I suggest you use this https://paste.ubuntu.com/p/KxwQv8rxb8/ [14:11] checking [14:11] <[Pokey]> means you don't have to fiddle with compiling it [14:11] ah I saw that one [14:11] <[Pokey]> that script is what I use to deploy to my relays [14:11] [Pokey]: you know find this interesting https://gitlab.xiph.org/xiph/icecast-server/-/issues/2310 [14:12] [Pokey]: hope I want get a license infringement notice [14:12] *I wont [14:12] <[Pokey]> I wouldn't be concerned as thats more of a debian repo issue [14:12] <[Pokey]> as far as I can see [14:13] ya [14:13] [Pokey]: I will try that. Thank you so much [14:13] <[Pokey]> If you have any issues with bundling your certificates and configuring them (took me a few days to work out how as IC2 is a pain with certs) HMU [14:14] oh I figured out the configuration part [14:14] <[Pokey]> hjow about the bundling? [14:15] <[Pokey]> Ah this is a short one, I can paste it here [14:15] <[Pokey]> cat /etc/ssl/mydomain.com/fullchain.pem /etc/ssl/mydomain.com/privkey.pem > /etc/ssl/mydomain.com/bundle.pem [14:15] I tried the package in ubuntu repo assuming it was compiled with ssl support and finished all the config and stuff with certs from letsencrypt [14:15] <[Pokey]> that just concatenates the certs super easy [14:15] [Pokey]: yep thats the one [14:15] <[Pokey]> Aha [14:15] <[Pokey]> yea that will take your LS certs and do it for ya [14:16] <[Pokey]> you don't wanna know how logn it took me to work that out [14:16] :D [14:16] <[Pokey]> also I need to type slower. I'm too used to being able to edit messages on Discord [14:16] <[Pokey]> XD [14:17] [Pokey]: I guess you hangout in #music? also people may thing we are offtopic :) [14:17] think [14:17] <[Pokey]> I'm actually new to IRC so I don't really hang out anywhere other than #freenode, but I will join #music [14:18] [Pokey]: You are most welocme [14:18] welcome [14:18] [Pokey]: I followed this guide https://mediarealm.com.au/articles/icecast-https-ssl-setup-lets-encrypt/ [14:18] <[Pokey]> Couldn't find that when I did it [14:18] <[Pokey]> I had to work it out myself xD [14:19] ya :) [14:20] how do I disable the GUI popup on segfault ? I do this for work (fixing segfaults) so it's pretty annoying to see these popups :) [14:22] Is there a list of "what I should be looking out for" for an Ubuntu 19.10 --> 20.04 upgrade? i.e. breaking changes, core component changes, et cetera? [14:23] electrostrong: I think you should read the release notes for known bugs [14:25] thx murthy - I'll check that out [14:26] Samba has always been garbage for me. From my win 10 box, I can access my share on my 20.04 box. Though, trying to copy a 3GB file, it fails 26% the way through (after restarting both computers to try again, it failed twice at around the same spot) [14:27] Looking closer at my 20.04 box, the wifi icon in the system tray has a question mark on it [14:27] switching wifi off and on again lets me TRY the file copy again [14:27] electrostrong: ty [14:28] I've tried googling for solutions but I'm not getting great search results, nothing I click on is relevant. Just lots of "how to install" "how to configure" samba intro blog posts [14:32] fwiw so far this is the most relevant https://www.linuxquestions.org/questions/linux-server-73/large-file-transfer-over-scp-or-samba-crashes-wireless-connection-786667/ [14:42] whelp. 4th time's the charm so they say [14:59] does ubuntu comes with gnome now? [14:59] 20.04 [15:00] Yeah [15:08] starting to explore spice for remote desktop - considering using it to let kids play minecraft via netbooks or outdated laptops, is this a path worth considering? [15:11] Simonious: Something to consider on a related topic would be Mineclone, which is free and quite a bit lighter than Minecraft. [15:11] Or, rather, Minetest - Mineclone is an add-on for it. [15:15] mason: my intent is to get them on a specific modded MC server [15:16] Ah, then yeah, pushing rendering off would be a valid idea - not unlike SteamPlay. [15:16] yeah [15:16] Just not sure how much work it's going to be for something I'm not sure will work [15:17] Simonious: Other options, VNC, maybe even straight remote X. [15:17] mason: are those better options? [15:18] Simonious: Well, I had to move to VNC from Spice for virt-manager because Spice was too heavy, so at least for libvirt and looking at remote VM consoles, VNC was a better option for me. [15:18] It's probably worth trying both. [15:18] too heavy for server or client? [15:18] Not sure if straight remote X would work, but I'd probably try that first, being the least effort. [15:19] Too heavy for the connection. [15:19] this would be on a local network [15:19] and I can upgrade the wifi if need be (probably) [15:22] Hrm, https://www.nas.nasa.gov/hecc/support/kb/vnc-a-faster-alternative-to-x11_257.html suggests that VNC is a better option than straight remote X. It doesn't speak to a comparison with Spice. If I were you I'd try both Spice and VNC and if they're both acceptable, pick the one with the smaller carbon footprint. [15:23] that is interesting [15:42] lotuspsychje: I see. It looks like I should request a backport [15:42] I wonder to what extent I could personally contribute to a backport - maybe if I do most of the work it is more likely to be accepted [15:42] I am thinking about upgrading my Ubuntu 18.04 to 20.04 but i read that snapd is tightly integrated, how much restrictions would it mean to uninstall and purge snapd in 20.04? [15:42] can SPICE ONLY connect to VMs? Can I use it to connect to a running OS that isn't virtualized? [15:44] Simonious: don't quote me on this but I'd imagine no since it is tightly integrated with the virtual graphics device et al [15:44] Simonious: perhaps you could try vnc? [15:45] spice comes more strongly recommended, but VNC may be a better option if I'm not running VMs [15:47] Simonious: you could look into https://www.spice-space.org/xspice.html [15:47] Maybe that's similar to xvnc [15:47] I'll take a look, thanks [15:48] There's apparently also x11spice: connects a running X server as a Spice server [15:49] Simonious: OK - I now think I was wrong. The answer appears to be: yes, that is possible. [15:54] codehotter: cool, I'll have to dig into that more later, I've got two friends recommending spice to me over VNC [15:56] in order to make a bootable usb stick should dd if of work? [15:57] Elodin: dd if= of= bs=1M status=progress [15:58] oh, forgetting the bs could had impacted? [15:59] Elodin: you could use a bigger blocksize [16:04] hiya guys... i'm on 20.04 LTS and i was wondering how i could go about installing a game... [16:05] it has deb files, but i'm still new to Ubuntu and i don't know how to install the debs directly like i would with an rpm [16:13] Ringtailed_Fox, I would use gdebi [16:13] oh derp, yes, thanks [16:13] sorry just so used to fedora.... still learning ubuntu [16:15] Ringtailed_Fox: if you right click on the deb and use gdebi, it will check/verify all deps are met before installing [16:17] ahh ok [16:18] hmm, yeah, i have to go deb hunting for some old libboost packages... tryign to install secret maryo chronicles (one of my favourite older linux games) [17:16] .bef [17:16] .dedbef [17:21] I just discovered jq and am the happiest developer in the world. this is me sharing it with you. [18:08] I am running nethogs and see lines like: ? root my.server.address:8000-113.11.246.95:12892 - does this mean that MY server has connected to 113.11...? I guess I do not understand the output format of nethogs [18:09] I am running nethogs and see lines like: ? root my.server.address:8000-113.11.246.95:12892 - does this mean that MY server has connected to 113.11...? I guess I do not understand the output format of nethogs === Guest96541 is now known as slowstoop [18:15] hey everyone, is there a good link explaining how I can make my /boot read-only, and conversely how I would make it writeable again when doing a system update? === beaver is now known as beaver|SYN [18:22] slowstoop: mount -oremount,ro /boot [18:22] it'll be a fun surprise when your next update fails :) [18:23] interesting thought, but no security at all [18:24] aardvark__, that line you posted suggest such connection, yes [18:25] slowstoop: What is your reason for wanting to do that? [18:25] oerheks: I have NO idea what on my server could be doing that, the address is in vanawatu - [18:25] oerheks: generally am I getting it right that the format is: source.ip:port-destination.ip:port? [18:26] how to uninstall scripts which is intalled with ./install.sh? [18:26] boktan: read the script -- sometimes there's an uninstall option. sometimes not. if there is an uninstall option, it might work. [18:27] no it not have that option i was able to do it with xarg with making txt output but i forget it :( [18:27] run that script again, with the -d option, dry run? [18:28] aardvark__, not sure what vanatu has to do with your server [18:29] oerheks: my point EXACTLY - I have NO IDEA why my server would be connecting to an island chain the the S Pacific [18:32] oerheks how to use that command? will it remove installed things via ./install.sh ? [18:32] no,the idea of -d dryrun is to do nothing, just give an output, so you can spot the places where your script writes things [18:33] then you can manually delete them, good luck! [18:36] sarnold, will that make it so /boot is automatically mounted as RO everytime the system boots? [18:36] slowstoop: no, it changes it to read-only at that moment [18:36] Jordan_U, just making sure the volume cant be messed with [18:37] sarnold, do I need to edit fstab to make that change permanent? [18:37] slowstoop: wouldn't it be easier to just fix whatever program is writing on your /boot so it won't do that? [18:37] guys, I have my reasons. [18:39] so it would be "mount -oremount,rw /boot" do make it writeable again? [18:40] sarnold, Jordan_U ; the main reason is that /boot cannot be encrypted by LUKS, or anything else, so I need whatever small protection I can get on it. [18:40] slowstoop: You're missing a space between "-o" and "remount", but yes. What stops the process that's trying to write to /boot/ from remounting rw itself? Are you familiar with secure boot protections and signed kernel images? [18:40] slowstoop: what protection does this offer? [18:42] slowstoop: /boot/ can be encrypted by LUKS, as long as you use LUKS1 rather than two. Grub can read from LUKS1 volumes. The downside is that (in most configurations) you need to enter your LUKS password twice, once for grub to read from /boot/ and again for linux to mount its root filestem, even if both are on the same volume linux needs to start from scratch at boot. [18:42] well damnit I've been lied to! [18:43] as far as stopping a process from RW'ing I'm gonna guess the lack of root privileges for one? [18:43] are there any file managers with context menu editors? besides nautilus [18:43] slowstoop: That should also stop writing to /boot/, since that also requires root. [18:44] right, so the idea is, the partition is set as RO, cant be tampered with while the system is up, and cant be tampered with if the disk is removed? [18:46] i mean... can I just not mount /boot at all? [18:46] its a seperate partition [18:47] although i'd like to be able to do system updates or edit GRUB occasionally [18:47] slowstoop: There's not really any such thing as a partition being "set" as read only. If someone has the hard drive, and the contents aren't encrypted, then anyone can write whatever they want to it. [18:49] ok well setting up LUKS1 is gonna be a bridge too far for me I think [18:50] slowstoop: don't forget that processes on the system with cap_sys_admin can just mount it again; processes with cap_sys_module can load new kernel modules to do whatever they want; cap_mknod can make their own device nodes, etc etc .. [18:50] im a bit of a rube, just trying to make a system I can carry around in my pocket without fear of physical or data loss [18:51] (i have twin sticks I just dd back and forth for longevity) [18:51] slowstoop: There's a lot to think about there. Encryption does an extremely good job of preventing someone who steals your laptop from learning your secrets. Preventing tampering is another question, encryption alone helps there, but doesn't protect you entirely the way that it protects secrets. For example, if someone tampers with your laptop while you're out of the room, they can add a physical [18:51] keylogger to it that will send them your LUKS passphrase and then they have all of your data, and access to do anything else with your machine, with you not knowing that anything has happened. The same can be done even if all they have access to is the ability to change bits on your hard drive. They can setup a fake linux kernel that asks for your passphrase, then passes it on to your real kernel and [18:51] again they have full access and you're none the wiser. [18:52] its a USB stick, so they're gonna have to take it from me first :) if they just find it in the street, they are gonna have a hard time, and like I said, I keep a twin stick in a vault [18:53] slowstoop: Then I'd say don't worry about /boot/ at all. There's nothing there that's secret, and anyone that has had access enough to modify it has generally had access enough to do other things as well. [18:54] slowstoop: Making sure that secure boot is enabled on the computers you use is a good idea though, and you can make sha256sums of all of the files in /boot/ and compare them to your backup stick if you ever want to check if they've been modified. [18:56] Is there a reason why grub can't read LUKS2 partitions? or was it that till now nobody was interested? [18:57] latest; https://www.phoronix.com/scan.php?page=news_item&px=GRUB-Boots-LUKS2-Disk-Encrypt [19:00] quadrathoch2: Support has been added to grub master, but it's not quite ready for prime time yet from what I've heard from those that have tried to use it. [19:00] ahh nice to hear, thanks oerheks Jordan_U [19:00] quadrathoch2: You're welcome. [19:02] have fun! [19:03] ya really :) [19:25] Hi [19:25] Is there any tool like `pv` but in graphical mode? [19:58] any recommendations for a well supported thunderbolt 3 dock primarily for external monitors? mostly so i can entirely turn off my nvidia card in my lenovo p1 gen 2 laptop. [20:01] having an external monitor means you don't have to use your nvidia card? [20:01] matsaman: having a graphics... thing with an hdmi or display port means i can turn off my nvidia. it's also got intel inside for the laptop display [20:01] but the existing hdmi port and the thunderbolt graphics are off the nvidia [20:02] oh so you don't mind using just the intel, but the external monitor ports are all through your nvidia? [20:02] biggest bother is the nvidia drivers crashing maybe averaging once a day forcing a reboot. second i suppose would be power [20:02] power would bother me, but not if I bought a laptop with a dedicated gpu on purpose [20:02] this is nvidia's drivers and nouveau both? [20:03] matsaman: i think i'd be ok with just the intel when away from my desk. i should probably test that out a bit i suppose. in case it presently is running ok on my lap because of leveraging graphics compute whatever from the nvidia [20:03] matsaman: nvidia drivers. i got the impression the nouveau weren't great [20:04] well they're much more stable [20:04] they lag behind in terms of hardware accelerated support, AIUI [20:04] matsaman: hmm, more stable but... not as fast? or... [20:04] CUDA, and that stuff [20:04] mm [20:04] matsaman: welp, maybe that's a direction worth trying i guess. [20:04] if you're really crunching numbers hard with that GPU, you could probably do it better with the binary drivers, assuming they work right [20:04] if you're just using it the way most people do, nouveau is probably going to be better [20:05] matsaman: i have piles of chrome windows and tabs. some konsoles, a customized eclipse, pycharm, and my own pyqt programs. so yeah, not gaming etc. [20:07] yeah that's gonna be almost all your proc [20:07] and modern intel can do the small graphics things most people use these days just fine, too [20:07] translucent windows, animated desktop effects, etc. [20:07] although can't say I find them super useful =P [20:08] https://help.ubuntu.com/community/BinaryDriverHowto/Nvidia#:~:text=By%20default%20Ubuntu%20will%20use,which%20are%20developed%20by%20NVIDIA. "By default Ubuntu will use the open source video driver Nouveau for your NVIDIA graphics card. This driver lacks support for 3D acceleration and may not work with the very latest video cards or technologies from NVIDIA." outdated? [20:08] seems to be based on https://nouveau.freedesktop.org/wiki/ "2D/3D acceleration supported on all GPUs" [20:10] Hey all. For work I have two VMs set up - Ubuntu 18.04.4 as a Yocto build server and Arch for development/other use. I'm looking to set up a fileshare so that I can access firmware source (hosted on Ubuntu) with emacs running on Arch. The two machines are connected with static IPs on a VMWare host-only network adapter, so I would expect ping to be low and speeds to be high. However, with an NFS [20:10] share, some operations (opening files, `git status`) are much slower on Arch than Ubuntu. any idea how I might speed this up a bit? [20:10] Read/write speeds seem totally fine, I think it has more to do with the latency of accessing many individual files [20:11] jadelclemens: NFS is a PITA [20:11] jadelclemens: just use sshfs [20:12] I want to use iptables to redirect traffic to a port to my proxy unless if it's traffic from my proxy. I added rules for -j ACCEPT if it's the packet originating uid matches proxy uid and redirect to proxy port if not. But it seems the match sometimes fails so the proxy connects to itself. Can someone help I have details here http://ix.io/2p0O [20:12] hello [20:12] hi pep [20:12] featherlessbiped: I'd ask #netfilter [20:12] matsaman: sshfs is performing much worse for me [20:12] They're all afk :( [20:12] jadelclemens: heh, I heard a very similar lament a few hours ago in another channel. I'm not sure there's much you can do if you want to keep using nfs.. *maybe* iscsi or similar to get some readahead behaviour on the client? dunno.. comparing disk-at-a-distance vs disk-locally, it might take a LOT of network before the two are competitive.. [20:13] jadelclemens: you'd want hpn with encryption low or disabled [20:13] I mean you could try samba, but meh [20:14] Hmm, I'll try just disabling encryption/compression before I look into hpn. Thanks matsaman, I'll check it out. [20:14] I'm not sure if this is the right place to ask but I have two Odroid-HC2s and I'm looking for a solution to sync the 2 drives ("backup") what would you guys recommend as a solution given what is out there? Was thinking glusterfs but not sure. Planning on using open media vault. Anyhow what are everyone's thoughts on that? Am I crazy? I'd use Rsync but I believe it doesn't work on locked files? [20:15] Peppi: if you want backup, zfs with send and receive may be useful [20:15] matsaman: hmm, i've got a T2000 mobile/max-q per 'software sources' > 'additional drivers' labeled as a tu117glm which https://nouveau.freedesktop.org/wiki/CodeNames/ maps to an nv167 and https://nouveau.freedesktop.org/wiki/FeatureMatrix/ lists every feature but kernel mode setting as 'todo'. still worth trying? :] [20:17] altendky: maybe not =) [20:17] sarnold, yes was looking at that. How does zfs handle locked files and if you have 2 nodes does zfs work ok? What does it do to the bandwidth? Will it essentially cut everything in half? [20:17] matsaman: welp, thanks for trying :] [20:18] there's always #hardware [20:19] matsaman: for my original question about well supported docks for ubuntu? [20:19] yeah [20:19] support will be the same for any GNU/Linux [20:19] matsaman: invite only, you have one handy? [20:20] Snap question: on every boot, I have /dev/loop* mounts from snaps I use maybe once a week. Is there some way to make that optional, and only mount them when I'm planning to use that app? [20:20] altendky: mmm, should redirect to ##hardware [20:20] wonder if irccloud doesn't support that [20:20] * matsaman grumbles about stupid freenode ######### insanity [20:21] matsaman: i thought it forwarded but... ## worked. thanks [20:21] akk: Are they causing any practical problems? [20:22] Peppi: zfs isn't a cluster-filesystem; if you want *backup* then zfs is probably good for that. if you want HA, then you probably do want glusterfs or ceph or lustre or ocfs2 or similar.. [20:22] Jordan_U: Only minor ones: they clutter up mount listings and clutter up systemctl list-units (the latter being why I'm annoyed with them right now). [20:24] Peppi: btrfs-send and btrfs-receive also exist, and they are modeled after zfs. The nice thing about copy on write filesystems is that you can just send whatever the current state is, and know that if it changes the filesystem knows exactly what changed since the last send. That doesn't mean that you automatically get consistency, but it's often enough for practical purposes. You can also have snapshots [20:24] and keep the state of the last N syncs around in perpetuity. [20:26] Jordan_U, does it play nicely with locked files? For example data bases and VM machines on a drive? [20:26] I have the entire disk (sda) on LVM, after adding another disk (sdb) to the same volume group and rebootin I get grub error: disk 'lvid/...' not found [20:27] Peppi: Locked in what way? Are you worried about a backup failing because a lock would prevent it, or are you worried about your backups not being consistent / having a snaphot of a database that was taken mid transaction? [20:28] Jordan_U, backup failing because a lock would prevent it [20:28] spoonsearch: If you run "ls" at the grub rescue shell, does it list more than one drive? It's possible that your boot firmware can't read from the new drive. === beaver|SYN is now known as beaver [20:32] Jordan_U, why it needs to read from the new drive? Using LVM is added the entire new drive to the Volume Group, and created a single Logical Volume from 2 Physical Disks [20:32] Peppi: I wouldn't expect that to be a problem. I would expect that zfs/btrfs send/receive just wouldn't care about locks. Don't bet money on that though. Also, if you do switch to a copy-on-write filesystem you may end up not wanting to have files like VM images and databases be copy-on-write, as that tends to lead to major fragmentation over time. You can still use send / receive with files that are [20:32] nocow, because you can still snapshot files that are nocow, but then you are introducing a certain amount of copy-on-write even though they're nocow. That is probably a lot to parse through if you aren't used to copy-on-write filesystems. [20:34] spoonsearch: If you have one Logical Volume that spans two physical volumes, how could you possibly read from that logical volume without having access to both physical volumes? [20:35] Jordan_U Sorry, I forgot to mention, both the hard drives are visible to GRUB [20:36] Peppi: But also understand that for the most part, *NIX people and software tends to do read locks even though Windows does them by default. Have you encountered a piece of software that prevents you from reading from a file while it's being written to? [20:37] spoonsearch: Is this machine booting via UEFI or BIOS? Is the lvid that you're seeing in the error message the correct lvid? [20:39] Jordan_U, most VM machines and databases like postgres will lock files [20:42] Peppi: This seems to imply that postresql doesn't: https://www.postgresql.org/docs/9.1/backup-file.html Note that while *-send has many of the problems they list here, it *does* make an atomic snapshot of the state of the filesystem. [20:44] Jordan_U It is UEFI, I don't know whether the livid is correct [20:44] Jordan_U The first time I added the new drive and rebooted, GRUB failed, but after a clean reinstall and adding it again this time I did `grub-install /dev/ssda` and `grub-update`and it rebooted fine. Is it necessary to run these two grub commands after adding, removing or resizing using LVM [20:45] Peppi: I would not be surprised by a write-lock, but in my experience read locks are rare. [20:48] spoonsearch: Where is your EFI System Partition? On UEFI systems a device argument doesn't make sense, and if one is provided it is ignored. So "sudo grub-install /dev/sda" is exactly equivalent to "sudo grub-install" when booted via UEFI. === beaver is now known as beaver|idle [20:53] Jordan_U: Sorry for the confusion, I checked the server if it was UEFI using `efibootmgr`, my system is BIOS === beaver|idle is now known as i_need_you_baby [21:17] exit [21:40] Hello [22:09] Hi guys [22:10] Hi, i know i am a cannonball [22:10] I interrupted a do-release-upgrade [22:11] and when I restarted it said to me "Please Install all vailable upgrade for your release before upgrading" [22:11] I had interrupted it during the download of the 2000 packages [22:11] so I did an apt upgrade [22:11] and it offers me to install and update the 2000 packages that wanted to install do-release-upgrade [22:11] But how will you react do-release-upgrade when I launch it after? [22:11] or else I don't know how to do it? Advices ? thank you [22:11] sudo apt install -f [22:12] it tells me that I have 5 old kernels [22:12] only that [22:12] is it good or bad ? [22:12] not bad. [22:13] StatelessCat: Also what shows ' sudo dpkg --configure -a ' ? [22:13] no [22:16] can I try a apt full-upgrade ? [22:17] as you started do-release-upgrade, no i do not advise that, try apt install -f to continue/fix [22:21] ok [22:21] good evening [22:41] I am seeing traffic from MY server to various ip addresses, I am running an "out of the box" ubuntu 18.4 - how do I associate those acesses to processes on my machine [22:45] aardvark__; what ports are they using [22:45] kk4ewt: the port numbers are all high, non-specific ones [22:49] aardvark__: sudo ss -tnlp [22:51] sarnold: that ss shows me 6 jobs that are listening and NOTHING else [22:52] and they are all valid - being DNS, and ssh daemon [22:53] aardvark__: hmm, take off the -t for udp, etc too? [22:54] sarnold: wow - there is a lot more now - not sure yet how to read it [22:55] t [22:55] what services do you run on your server? [22:56] sarnold: what does unconn mean here? [22:56] oerheks: if that question was addressed to me - NOTHING yet - it will be having apache - but that is not yet on [22:57] sarnold: the entries from the ss other than the UNCONN are just as expected - DNS and SSHD only [22:58] er. oops. I have a sligtly older Dell Precision 15 laptop. I have a problem in Firefox where, when I scroll up with 2 fingers on the trackpad, once it reaches the top or bottom, it continues to send the scroll signal for a few seconds. As a result, if I hit ctrl within those few seconds (which I often do), it starts re-scaling the size of the web render. Very annoying. Can I perhaps get it to stop trying to [22:58] scroll once it's reached the edge of the scroll view? Not sure if it would be a bug/config in firefox, the X server, or what. [22:59] or trackpad driver [23:07] sarnold: so, 1) does that UNCONN mean unconnected, 2) any ideas as to what I should look at next. I would add that it is nethogs that is showing these connections [23:12] aardvark__: UNCONN means "unconnected" -- UDP sockets can be connected or unconnected [23:19] sarnold: in the case of those UNCONN how do I know if they are the culprit then? [23:20] aardvark__: I'm not sure -- you'll need to get more details from your logging system to figure out what kind of traffic you've got [23:21] not sure what u mean [23:27] since upgradinf to 20.04 vlc and firefox, if startes from a terminal with 'firefox -p' start very slow: I need to wait 5 seconds on each program to have a working gui. what could cause this? [23:27] geez [23:28] since upgrading to 20.04 if I start vlc from the gui and firefox from a terminal with 'firefox -p', they start very slow: I need to wait 5 seconds on each program to have a working gui. what could cause this? [23:35] frad: lack of ram causing swapping? [23:35] tomreyn, I have 16 GiB of RAM [23:36] tomreyn, after upgrading, I copied my .firefox from the old ssd into a new unit, a nand drive. Is this relevant? [23:36] frad: or maybe a storage which is loosing cohesion? [23:36] tomreyn, you mean the nand drive is not phyisically correctly installed? [23:37] frad: shouldn't be relevant unless the nand drive is slower than the ssd. [23:37] tomreyn, if that were the case, I'd have been ripped off [23:38] i assume you compared the specifications? [23:38] yes, but I forgot :D tomreyn, is there any command to check the speed of both the ssd and the nand drive? [23:40] yes, there is: gogo gadgetto google [23:40] i mean, just look up their specifications [23:41] if you don't know the models, you can check those at /dev/disk/by-id/ [23:42] or, e.g. for SATA devices: journalctl -b | grep ' ATA-' [23:56] I found sudo hdparm -tT /dev/sdb tomreyn [23:56] and yes, the new unit is faster [23:56] frad: is it all new or had you used it previously? [23:56] tomreyn, used [23:57] maybe fstrim its file systems then [23:57] from me, not second hand bought [23:57] tomreyn, do I need to backup the info before fstrim-ming it? [23:58] it's always good to have recent and complete and proven restorable backups. this said, for fstrim you shouldn't need them. [23:58] unless something goes wrong, of course. ;-) [23:59] use fstrim -v /path/to/mountpoint and run it twice for each file system supporting trim. [23:59] just so you'll see it actually did something