[01:08] <Deihmos> is there a custom install for the server? the install is 4GB
[01:08] <Deihmos> while debian is just 800MB
[01:10] <sarnold> you could try http://ports.ubuntu.com/ubuntu-ports/dists/bionic-updates/main/installer-arm64/current/images/netboot/
[01:10] <sarnold> oh sorry that's arm64
[01:10] <sarnold> http://cdimage.ubuntu.com/netboot/18.04/
[01:12] <Deihmos> the mini iso?
[01:13] <sarnold> yeah, chances are good that'll leave you a system with almost nothing installed. should be tiny.
[01:17] <Deihmos> lol
[01:17] <Deihmos> i am going to try it in vm now
[01:30] <teward> sarnold: rbasak: and anyone else who cares: nginx 1.15.8 now uploaded to the repos.
[01:31] <sarnold> thanks teward :)
[01:34] <teward> yep.  prepping the PPAs for the update too
[01:36] <teward> cc zzlatev_ ^ tomorrow the PPA will have 1.15.8
[01:36] <teward> in case you intend to do updates as well :p
[01:54] <Deihmos> trying to decide between ubuntu server or debian as a plex server
[02:08] <Deihmos> the mini still ended up with 4GB
[02:08] <sarnold> Deihmos: lol
[02:08] <sarnold> thanks for reporting back :)
[02:08] <sarnold> I sure expected it to be 800~1000 megs..
[02:09] <Deihmos> maybe i need to do expert install
[02:09] <sarnold> you can apt-get purge whatever you don't need, of course
[02:23] <zzlatev_> teward: thanx\
[02:35] <teward> sarnold: ubuntu-standard pulls in a lot of stuff ;P
[02:35] <teward> might be where the bloat is coming from
[02:40] <sarnold> nano, ureadahead.. deadweight :)
[02:41] <teward> :P
[02:41] <teward> debian bare is *very* tiny because it includes pretty much nil in it by default
[02:41] <teward> Deihmos: ^
[02:41] <teward> just wanting to make that clear
[02:42] <teward> but 4GB does sound excessive
[02:42] <teward> Deihmos: did you choose anything in the installer or...?
[02:43] <teward> or did you only go for the bare minimjm
[02:43] <Deihmos> 3.5GB
[02:43] <Deihmos> openssh was selected
[02:43] <teward> 4GB usually means you selected other things to install beyond the standard bits.  any GUI?
[02:43] <teward> pretty sure "no" but
[02:43] <Deihmos> nope just openssh
[02:45] <teward> that... doesn't sound right.
[02:45] <teward> *pulls down an ISO and does a VM test*
[02:45] <sarnold> can you pastebin your dpkg -l ?
[02:45] <teward> ^ that
[02:46] <teward> Deihmos: can you also confirm *how* you're determining 3.5GB
[02:46] <Deihmos> df -h
[02:47] <teward> can you include that output as well?
[02:57] <Deihmos> https://www.irccloud.com/pastebin/LnQ2oS6x/
[02:58] <sarnold> that looks like a partial list
[02:58] <sarnold> e.g. there's no bash
[02:58] <sarnold> but you probably has bash installed :)
[02:59] <teward> sarnold: 1.5GB with just SSH-server, 1.8GB with OpenSSH + Ubuntu Server standard
[02:59] <teward> so it sounds something else is going on possibly
[02:59] <sarnold> teward: that's way more like it. still a bit bigger than I expected but at least it's not 4gigs :)
[03:00] <teward> sarnold: and there's a 20GB partition in this VM so
[03:00] <sarnold> teward: thanks for giving it a try, it's nice to know I'm no insane, hehe
[03:00] <teward> Deihmos: can you provide the `df -h` output you're seeing as well?
[03:00] <teward> also `dpkg --get-selections`
[03:00] <teward> which should look like what sarnold is expecting to see :p
[03:01] <Deihmos> https://www.irccloud.com/pastebin/xWHMf4rf/
[03:02] <Deihmos> https://www.irccloud.com/pastebin/gL4uGgye/
[03:03] <Deihmos> https://www.irccloud.com/pastebin/Cz9qjjRj/
[03:03] <Deihmos> error
[03:03]  * teward headscratches
[03:04] <teward> that looks like only a partial scrollback copy
[03:04] <Deihmos> what are you referring to ?
[03:05] <sarnold> Deihmos: run dpkg -l | wc -l
[03:05] <sarnold> note how it's a lot more than 60
[03:05] <sarnold> but we're only seeing sixty packages in your pastes..
[03:06] <teward> http://paste.ubuntu.com/p/6xnnjbXHTP/
[03:06] <teward> should look more like that
[03:06] <sarnold> if copy-and-paste is hard on that machine, install the pastebinit package, and run dpkg -l | pastebinit
[03:06] <teward> ^ that
[03:06] <teward> install pastebinit, then it'll give you links on paste.u.c, and you give us those links
[03:06] <sarnold> teward: hah, I look at that list and immediately think "ethtool probably should be installed"
[03:07] <teward> sarnold: well I stripped away the ubuntu server standard set in this :p
[03:07] <teward> but ethtool is in my install :P
[03:07] <teward> unless you mean their list
[03:09] <cryptodan_mobile> You could pipe it to nc termbin.com 9999 and itll generate a pastebin link
[03:10] <teward> there's a dozen different ways to do it :P
[03:13] <Deihmos> it appears this ssh software i am using have imitations
[03:14] <Deihmos> or just buggy
[03:14] <sarnold> cryptodan_mobile: yeah, very nice when pastebinit isn't available
[05:39] <zzlatev_> teward:
[07:24] <pangolino> Hi, I've got a fresh install of Ubuntu 18.10 on an intel NUC (8i) and have a question about chaning the MTU for the NIC, if this is more of a #ubuntu question feel free to send me there
[07:25] <lotuspsychje> pangolino: are you installing server non-lts on your nuc?
[07:25] <pangolino> I'm just trying to set the MTU to 9000, however via google/stack overflow etc I'm heading deep down a rabbit hole of netplan and networkd
[07:25] <pangolino> yes, non-lts
[07:25] <lotuspsychje> pangolino: for testing or production?
[07:25] <pangolino> better to go with lts and keep life simple?
[07:26] <lotuspsychje> pangolino: the users choice
[07:26] <pangolino> home use, non production - for learning but equally I don't want to be too unstable
[07:27] <lotuspsychje> pangolino: did you upgrade bios firmware to latest, those are important on a nuc
[07:27] <pangolino> yes, latest bios
[07:27] <lotuspsychje> great
[07:27] <lotuspsychje> i run 18.04 on a nuc, but not sure about the mtu myself
[07:28] <lotuspsychje> you got network speed issues or so?
[07:28] <pangolino> default mtu of 1500 makes for slow transfers on the 1gig lan, rest of network runs with an mtu of 9000 so was aiming to keep everything in sync
[07:28] <pangolino> (slow multi-gigabyte transfers)
[07:30] <pangolino> I have netplan and generated networkd configs in pastebin if of interest - stack overflow/google seemed to suggest that the DHCP MTU "UseMTU" could not be configured with netplan, and that was overriding my settings
[07:30] <pangolino> https://paste.ubuntu.com/p/P6bVFvwJND/
[07:30] <pangolino> I confess I was unaware that DHCP had any influence over mtu so you live, you learn
[07:38] <lotuspsychje> pangolino: can this help? https://askubuntu.com/questions/1099243/netplan-bond-bridge-mtu-setting-not-being-honored-on-bionic-system
[07:47] <pangolino> lotuspsychje: I think that's one of the ones I've read - I added a match by macaddress but still UseMTU gets set
[07:48] <pangolino> one solution i think is to go for purely static config, remove DHCP from the equation and get on with life
[07:49] <pangolino> I have static leases so it wouldn't be the end of the world but it feels like giving up over something that on the face of it should be pretty easy
[07:53] <lotuspsychje> pangolino: here another guy set 9000 https://ppc64el.wordpress.com/2018/03/22/ubuntu-18-04-netplan/
[08:00] <pangolino> lotuspsychje: that one appears to be for multiple nics, and is going the route of turning of dhcp and statically configuring
[11:28] <azidhaka> there is a proxy, similar to haproxy, but meant for load ballancing VMs, can anyone remind me its name
[11:30] <JanC> pangolino: is there a netplan bug report about DHCP MTU?
[11:33] <JanC> cyphermox might be interested in that
[18:06] <teward> zzlatev_: i wasn't online 13 hours ago - was 1AM sorry
[20:04] <lunaphyte> after upgrading, my serial port is not longer working.  i've collected the following clues so far: http://dpaste.com/2EADK2H.txt
[20:05] <lunaphyte> it seems like there may now be an irq conflict between the serial port uart and the i8042 device, for some reason?
[20:06] <lunaphyte> how can i correct this?
[20:06] <lunaphyte> i've done a bit of research, and tried booting with the i8042.nokbd kernel command line option, but this didn't seem to make a difference
[20:06] <lunaphyte> i'm on 18.10
[20:52] <TJ-> lunaphyte: broken it again!?
[20:54] <lunaphyte> still recovering :)
[20:54] <lunaphyte> although it may be working better
[20:54] <lunaphyte> i've used setserial and now i seem to be able to at least open the port
[20:55] <lunaphyte> and i think it's working right, but haven't validated that yet
[20:55] <lunaphyte> i wonder why it somehow ended up trying to use the same irq as this i8042 device
[20:55] <TJ-> lunaphyte: is this a VM or bare metal?
[20:55] <lunaphyte> and i wonder if i'll have to do something to change this every time it boots
[20:55] <lunaphyte> it's a physical server
[20:55] <lunaphyte> a dell pe1950
[20:56] <TJ-> lunaphyte: are there any options on the kernel command-line (/proc/cmdline) ?
[20:56] <lunaphyte> BOOT_IMAGE=/boot/vmlinuz-4.18.0-13-generic root=/dev/mapper/vg_1-root ro nomdmonddf nomdmonisw i8042.nokbd
[20:56] <lunaphyte> i'd added i8042.nokbd for troubleshooting
[20:57] <TJ-> lunaphyte: generally, unless the BIOS is misconfigured, the way this usually happens is if the ACPI tables are buggy (which tell the OS where things are)
[20:57] <lunaphyte> nothing physical or bios has changed
[20:58] <lunaphyte> i'd been using the serial port for ages, and it broke now that i've upgraded
[21:00] <TJ-> hmmm, can you show the complete dmesg from boot-time?
[21:01] <TJ-> you've (obviously) got 2 devices sharing an IRQ which generally should be fine if the drivers support sharing
[21:04] <lunaphyte> TJ-: http://dpaste.com/2TSJ12W.txt
[21:06] <TJ-> the error you're seeing in dmesg is because the drivers are not agreeing on the per-CPU flags, here's the code: https://paste.ubuntu.com/p/T3zGK2gJrZ/
[21:09] <TJ-> the flag is #define IRQF_SHARED     0x00000080  so ttyS0 doesn't want to share but i8042 does
[21:12] <lunaphyte> why do they end up sharing if ttyS0 doesn't want to?
[21:17] <TJ-> You must be causing something to change, because: "[    0.759125] Serial: 8250/16550 driver, 32 ports, IRQ sharing enabled"
[21:19] <lunaphyte> hmm.  i'm not sure what i could have done
[21:24] <TJ-> what do you get from "cat /sys/class/tty/ttyS0/{flags,irq}"
[21:25] <lunaphyte> 0x100000C0
[21:25] <lunaphyte> 0
[21:29] <TJ-> that seems strange, the C0 is 80 | 40 and we seem to have found 80 is IRQF_SHARED. I have not been able to determine what 40 is
[21:31] <lunaphyte> just fyi, right now the irq is "0" [autoirq]
[21:33] <lunaphyte> http://dpaste.com/1ZSTRTC.txt
[21:36] <maddawg2> command line question...  what's a good way to copy a directory from one location to another AND see the progress of that copy?
[21:37] <lunaphyte> rsync?
[21:37] <maddawg2> will rsync display the progress in the shell?
[21:38] <mybalzitch> --progress
[21:38] <lunaphyte> that was the criteria you specified, right?
[21:38] <lunaphyte> "see the progress of that copy"
[21:38] <maddawg2> ah yea i see that switch
[22:42] <zzlatev_> teward: no problem!
[22:42] <zzlatev_> thank you for everything